Loading...
HomeMy WebLinkAboutAPA1203----------l> r --l> -Ul === " c.J -----l> o=== .!>.!!!!!!!!!!!!!!! ;o Ul=== m c.n= Ul o= 0 o-c o= ;o c.n = () -.l= m c.n=== Ul .!>.!!!!!!!!!!!!!!! m= r .1>. ..... (D ;o -----l> === ;o -< -----ELECTR IC PO\-lER CONS~TION FOR THE RAILBELT: A PROJECTION OF REQUIRE~lliNTS TECHNICAL APPENDICES by Scott Goldsmith Lee Huskey Institute of Social and Economic Research Anchorage ~~ Fairbanks :< Juneau prepared jointly for State of Alaska House Pm,;rer Alternatives Study Co mm ittee and Alaska PoHer Authority May 23, 1980 -- I PORil.AND, Oti:{;fJ:l ( ELECTRIC POWER CONSUMPTION FOR THE RAILBELT: A PROJECTION OF REQUIREMENTS· TECHNICAL APPENDICES by Scott Goldsmith Lee Huskey Institute of Social and Economic Research Anchorage * Fairbanks * Juneau prepared jointly for State of Alaska House Power Alternatives Study Committee · and Alaska Power Authority May 23, 1980 l-ID ~es.'65""" U.lo C::J{.ps- vI fJ_ c.-z... ELECTRIC POWER REQUIREHENTS FOR THE RAILBELT TECHNICAL APPENDICES TABLE OF CONTENTS A. METHODOLOGICAL REVIEW Type of Forecasting Models. • • • • ••••••• Forecasting Alaskan Economic Activity • • • • • • • • Forecasting Alaskan Electricity Requirements .• Endnot.es. • • • • • B. THE ECONOMIC PROJECTION HETHODOLOGY Introduction. • • • • • • • The MAP Economic Model. • • • • • Household Formation Hodel • • • • • Regional Allocation Model • Housing Stock Model • • • • • • • • C. ECONOMIC PROJECTIONS What Are Projections. The Approach of the Current The Scenarios 1980-2000 • • Post 2000 • • • • • • • • • Study • Projections of State Growth 1980-2000 • • Regional Projections. • •••• Housing Stock Projections • • • • • • ••••• D. COMPONENTS OF THE END USE MODEL A-1 A-6 A-10 A-16 B-1 B-2 B-6 B-13 B-19 C-1 C-3 C-5 C-36 C-37 C-45 C-54 Households and Housing Stock. D-1 Residential Electric Space Heating. • D-20 Major Appliance Saturation Rates. • D-51 Major Appliance Fuel Type Mode Split. • • • • D-60 Electrical Appliance Average Annual Consumption • D-74 Small Appliance Electrical Use. • • • • • • D-81 Commercial-Industrial Requirements. • • • • D-82 Endnotes. . . . . . . . . . . . ·. . . . . . . . D-99 E. ELECTRIC UTILITY SALES PROJECTION METHODOLOGY Residential Nonspace Heating Electricity Requirements • • • • • • • ·• . • • • • • • • E-1 Residential Space Heating Electricity Requirements. E-7 Commercial-Industrial-Government Electricity Requirements • • • • • • • • • • • • • • • • • • E-ll TABLE OF CONTENTS (cont.) F. ELECTRIC UTILITY SALES PROJECTION MODEL PARAMETERS Projection Model .Parameters for Base Case • • Assumptions for the Price Induced Shift Toward Electricity Consumption Case • • • • . G. PROJECTING HILITARY AND SELF-SUPPLIED INDUSTRIAL ELECTRICITY NET GENERATION Military Requirements • • • . • • • • • Self-Supplied Industrial Requirements . H. HISTORICAL ELECTRICITY SALES AND ECONOHIC DATA I. A REVIEW AND COHPARISON OF RAILBELT ELECTRIC POWER REQUIREMENT PROJECTIONS J. BIBLIOGRAPHY . . .. F-1 F-49 G-1 G-3 APPENDIX A. METHODOLOGICAL REVIEW A.l. Types of Forecasting Models There are three general types of forecasting techniques or models. Time series or trend models include all mathematical techniques in which the forecast is a function of time. The justification for this approach is that past behavior is the best guide to forecasting the future behavior of a variable. Past trends form the basis for forecasting~ and the various techniques used in this form of modeling are concerned with identifying the most significant past movements of the variables being forecast. A simple example of a trend technique is the classical (past, univer- sally used) method of forecasting electricity consumption growth. During every decade since the turn of the century until 1970, electricity con- sumption in the United States has'doubled. This corresponds to an annual growth rate of 7 percent. This information is enough to formulate a trend model for projecting future electricity consumption growth. Con- sumption next year is 7 percent greater than this year. An obvious advantage of this type of model is that it is easy and cheap to construct and use (although some time series models can be very complicated and expensive to develop). If the actual process by which the forecast variable is determined is stable over time, then this technique will work well. In general, the shorter the forecast interval, the more likely it is that the system will be stable. The historical growth of electricity consumption could be interpreted as a reflection of a very stable process. For many decades, a simple trend model worked well to forecast future consumption. Basically, a time series or trend "model is appropriate when 1) cost and time minimization involved in making the forecast is important, 2) the process determining the forecast variable is stable, 3) the forecast interval is short, and/or 4) no information is available on what factors really account for the level of the forecast variable. Causal models include all mathematical techniques where a specific set of factors is assumed to "cause11 or determine the value of the fore- cast variable and the causal relationship is quantified. The idea behind these techniques is that the identification of cause and effect relation- ships facilitate our understanding of and forecasting of future events. These modeling techniques thus try to pick out the most important causal relationships and to measure them quantitatively. An example of a simple causal model used to forecast electricity consumption would be based on the idea that the level of population and income determines or "causes11 the level of electricity consumption. Thus, as population grows or income grows, electricity consumption is forecast to increase, and vice versa. A-2 The obvious advantage of this type model is that the forecast is based upon the notion of cause and effect~ rather than the identification of patterns in the past movement of the forecast variable (as in· the time series technique). If there are past or forecasted changes in the process by which the level of the forecast variable is determined, this forecasting technique can accommodate them. Thus, a time series model for forecasting electricity consumption would not be able to respond to a sudden drop in population which would~ in all likelihood, "cause" a reduction in consumption. A causal model with population as a causal factor would be able to accommodate this structural change. Another advantage of causal models for forecasting is that they allow one to do "what if" experiments. Using the causal model of elec- tricity consumption, one could determine what would happen to consumption if population doubled, fell by half, etc.. Time series models do not have this capability. Obvious problems with causal models are the time and cost generally associated with their construction and the requirement that the causal variables must be forecast in order to use the model. For example, forecasts of population and income are necessary to use the causal model for the electric~ty consumption forecasting discussed here. The causal model is generally appropriate when 1) the process by which the forecast variable is deter- mined is not constant over time, A-3 2) it is possible to quantify links between causal variables and the forecast variable, 3) the forecasting model will be used repeatedly using varying assumptions regarding the causal variables, 4) the time interval for the forecast is long, and/or 5) sufficient time and money are available to develop the model. Judgmental models are all nonquantitative or nonmathematical tech- niques for forecasting. The simplest example is the informed opinion of an expert. A more complex and structured technique is the Delphi Technique, in which a group of experts interact in a formal way to develop a consensus forecast. The rationale for judgmental models is that individuals with an understanding of the process by which the fore- cast variable. is determined can directly interpret a large amount of information relevant to projecting the future level of the forecast variable and on that basis develop a forecast. No specific mathematical o:i: formal techniques are used. An example of judgmental techniques would be in the forecasting of technological change affecting electricity consump,tion. It is not possible to forecast by quantitative methods the types and timing of technological change which would either increase or decrease electricity consumption. This technique is particularly appropriate where the number of potential causal variables is large and they cannot be systematically analyzed for lack of information, resources, etc. Obviously, the longer A-4 r the forecast interval, the more general are the variables which need to be considered and, thus, the more appropriate judgmental forecasting becomes. A possible dra'tvback of this method is that system interactions (the effect of one variable upon another) may be overlooked in complex systems. This technique is appropriate when 1) resources for forecasting are limited, 2) the process determining the forecast variable is not amenable to quantification, and/or 3) data are not available to develop a quantitative model. Each of these general techniques is appropriate in certain situations depending upon the process being.analyzed, the desired accuracy of the forecast, the resources available to develop the forecasting tool, the uses to which the forecasts will be put, and an evaluation of what method actually forecasts best. In any complex forecasting effort, elements of all three methods will likely be present. The forecaster must determine what technique is appropriate ~t each step in the process of developing the overall fore- cast. ·For example, the forecaster may decide that a causal model is the appropriate general approach. He may use time series or trend analysis to forecast the value for an explanatory variable. Finally, he may adjust the forecast derived from the causal model on the basis of infor- mation he has which is nonquantifiable in the sense that it cannot be incorporated into the causal model in a formal way. A-5 A.2. Forecasting Alaskan Economic Activity The important considerations in choosing a general technique for forecasting Alaskan economic activity for the purpose of projecting electricity requirements are as follows: 1) a thirty-year forecast is required; 2) the forecasting model will be used repeatedly and will be utilized to do "what if" experiments; 3) the Alaskan economy is a complex system with many factors interacting in important but non-obvious ways; and 4) the future growth and development of the Alaskan economy may be considerably different from the past because of a qualitative and quantitative change in the factors determining the state of the economy. These considerations suggest a quantitative causal model which has the advantages of 1) identifying structural relationships within the economic system, quantifying them, and automatically tracing the effects of these relationships through the system; 2) allowing "what if" experiments to exci.mine alternate futures under alternate assumptions about important future events and variables; and 3) providing a framework within which all forecasting assumptions are explicit. Arguments against the causal approach are 1) the complexity of the model makes its development and utilization relatively expensive; 2) the quality of the data may not warrant sophis- ticated causal models; A-6 3) the causal relationships may not be quantifiable; and 4) the causal relationships may not be understood. In choosing a quantitative causal model, we feel that, although the causal relationships in the Alaskan economy may be only partially under- stood and the data of poor quality, the advantages of formalizing the relationships that are considered relevant outweigh these problems. Large elements of trend analysis and judgmental models enter into · the causal modeling approach in the development of the projections of some of the most important variables in the economy, and these techniques become more important the further into the future the forecasts are pushed. Currently, twenty years is the limit for the quantitative causal model. Beyond that time, we revert to,trend and judgment. Having decided upon a quantitative causal modeling approach to economic forecasting, there remains the question of choosing among the many types of models available. In fact, there have been at least seven different economic models of the Alas~an economy developed over the last fifteen years, although only three econometric models are in current use. (These are the Man in the Arctic Program [MAP]" model, the Alaska Department of Labor model LABMOD, and the Alaska Department of Commerce and Economic Development model AEIRS.) In choosing a type of causal model, we would want one which is best able to represent the important factors in the workings of the economy A-7 itself. Also, since the projection is over a long-time period, we would want a model capable of adequately forecasting the level of economic activity in the long run. These criteria eliminate both input-output and short-run forecasting models. The first is rejected because it best represents interindustry flows of intermediate goods in manufacturing. There is almost no manu- facturing in Alaska and, thus, an interindustry transaction table would be mostly zeros. Short-run forecasting models are designed specifically for that purpose and as such can ignore many important long-run phenomena. The choice is thus narrowed down to some type of long-run, econo- metric or simulation model. One advantage of both these model types is their flexibility in terms of the situations they can model and the theories they can represent. The difference between them is that the econometric model uses statistical techniques (such as regression analysis) to identify.the quantitative relationship's among variables, while the simulation model uses nonstatistical methods (averages, point estimates, judgment, etc.). Econometric determination of quantitative relationships has the advantage of being able to interpret formally all historical information concerning those relationships although the techniques used can be expensive, time consuming, and on occasion too sophisticated for the Alaskan data. Because of repeated model use, however, we feel the econometric approach is appropriate. A-8 Finally, we want a model that specifically incorporates several important features of the Alaskan economy. These include: 1) the importance of state government activity in determining the level of economic activity; 2) the importance of and potential variability in development of Alaskan natural resources (particularly petroleum) on the level of economic activity; ·3) the process of maturation of the economy as it grows in size; and 4) the links which exist between the Alaskan and national economy. The MAP econometric model includes features which address all of t~ese important relationships but is criticized in the treatment of some of them. Specifically, it is suggested that the "multiplier" is too large and that the method used to determine population is inappro- priate. The "multiplier" is a quantitative measure of the support sector response to changes in basic ~ector activity. The support sector includes such industries as retail and wholesale trade and services, while the basic sector is made up of those industries which "drive" the economy, such as petroleum, portions of government and construe- tion, etc. In a developing economy such as Alaska, the support sector is growing rapidly, and this is reflected in a large "multiplier" value on increments to basic sector activity. A-9 It is also argued that the "multiplier" is too larg~ because it includes a state government response to increases in economic activity .in the private sector. We argue that this type of response has indeed occurred historically (the level of the state government budget has grown partially because private sector growth has increased the demand for public goods and services) and, absent specific state policies to severe the connection, will continue to occur in the future. The method used in the MAP model to determine the level of net migration to the state, and thus ultimately the level of population, uses the relative Alaskan wage rate and the change in the employment level as explanatory variables. These variables are generally accepted by econo- mists and demographers as being important in the determination of migra- tion patterns. Clearly, other factors are also important, although they are not quantifiable. Neither is it clear how, taken together, they would effect the results obtained using the two economic variables. In the absence of such analysis, the present technique appears appro- priate. A.3. Forecasting Alaskan Electricity Requirements As previously noted, the traditional method of forecasting elec- tricity requirements was the use of trend analyses. In the aggregate, it worked well until the early 1970s. At that time, however, there occurred a sharp break with the past in terms of behavior in all energy markets. Growth in consumption of electricity began to fall A-10 short of growth projected by trend analysis. It was clear that the structure of electricity markets was changing in ways that trend analysis was unable t·o anticipate and adjust to. These changes included a re- versal in the long-term trend of declining real prices for electricity, the attainment of high saturation rates for many appliances, the appearance of the conservation ethic, demographic changes, and other factors. Figure A.l dramatically demonstrates this inadequacy of the traditional forecasting method as applied to California. It may be that a new long-term trend in growth of electricity con~ sumption may emerge after the electricity market again stabilizes, but it is unlikely that this will occur for some time. In the interim, causal forecasting techniques are necessary which explicitly attempt to identify those factors which are causing change in the basic electricity con- sumption patterns. In a9dition, causal models explicitly allow utilities and policy analysts to examine the effects on electricity consumption of policies which effect electricity prices and other causal variables. Judgemental models have a role in this forecasting task for two reasons. First, the data is often not available with which to develop a quantitative model. Second, there are some relationships which may be difficult or impossible to quantify. Future changes in technology which will alter consumption patterns cannot be forecast quantitatively, for example. Emerging important factors effecting load growth in a particular market for which no historical information is available is another. In all these cases, however, the judgement must come down to A-ll FIGURE A.l. CALIFORNIA UTILITIES STATEHIDE ELECTRICITY SALES FORECASTS ----_Utility Forecast ·(]) -----CEC Adoptc1 1977 Forec_?st 235 :X: 220 ;: ~ z c _, _, a3 1980 1985 (D,{DG.O. 131 -Utility submissions to the Public Utilities Commission G),@ ,@Utility s-ubmissions to the CEC ~CEC adopted forecast (1977) -:::s--.-,... --------ource: c l"f 76 BI.LUO J(Wtf · 32 BILLION KWH a ~ ornia Energy Commission "Califor . E A Preliminary Assessment," August 1979~~~. n{l.gy Demand 1978-2000: l 10~ I I ~ 10: a quantitative value which can be used in the analysis because the electricity requirements forecasts are quantitative. The primary forecasting tool is the quantitative causal model of which there are two basic types: econometric and end use models. Econometric models of electricity consumption use statistical tech- niques to quantify relationships between electricity consumption or electrical appliance ownership and price~ income~ and other "explanatory" variables. They are generally quite aggregate in the sense that all residential consumption for all purposes may be lumped together and "explained" by the same price and income variables. These models have been developed by economists who have primarily been interested in identifying the exact strength of the relationship between electricity consumption and the important economic variables price and income. (They often refer to the concept of elasticity~ which is the percentage change in electricity consumption in response to~a 1 percent change in price or income. Thus~ a-price elasticity of -.5 would mean that when the electricity price increased 1 percent electricity consumption declined .5 percent.) End use models disaggreate electricity consumption not only by class of user but also by use. Total consumption is then the sum of all electricity consuming appliances and the average amount of elec- tricity consumed in each. The stock of electrical appliances will be determined by demographic and economic variables and average consump- tion by those same variables~ as well as technology~ use patterns~ and other variables. A-13 Because of the great diversity observable in electricity consump- tion, it is necessary to collect large amounts of data in order to con- struct an end use model. Its advantage is a capability of identifying technological, regulatory, and other policy-related factors which effect demand. Having done this, the model can be used to forecast consump- tion pattern changes in response to changes in these f~ctors. Each o~ these modeling approaches has advantages and drawbacks. The econometric approach uses accepted st-atistical procedures to develop quantitative relationships for· certain variables, but data limitations restrict its ability to analyze specific electricity uses and non- economic-demographic factors effecting consumption. There is concern within the economics profession that the many "demand functions" which have been estimated by researchers over the years really are measuring a relationship between electricity consumption and price and income. We feel there is substantial truth to these arguments. In addition, there is suspicion that relationships estimated using historical elec- tricity prices may no longer be appropriate because of recent and· rapid price increases in contrast to the previous long-run secular decline in· electricity price. Even if there was agreement that the approach of the econometricians was valid, the substantial variation in results re- ported by various studies invites caution in the utilization of ~ny particular estimate. One review article cited long-run price elastici- ties from different studies of between -1 and -2 and long-run income elasticities of between .2 and 2.1 A-14 End use models are disaggregated and lend themselves to policy analysis but generally are lacking in a strong statistical foundation for the parameters (quantitative relationships) used in the modeling. This is largely because of a lack of adequate data. In many instances, a single data point may be all the researcher has available to· work with, and this does not provide a firm basis for quantification of a rela- tionship between two variables. The best solution, suggested by a recent article in Public Utilities Fortnightly, is to utilize the better features of both modeling approaches to make the modeling sensitive to both the economic factors of price and income as well as to regulatory and technological factors.2 In the model developed for this study, this was also the preferred approach with a heavy emphasis on end use because of the anticipated / use of the model not only for forecasting but also policy analysis. Relatively little emphasis was placed on econometrics becau~e of the absence of data, the theoretical problems alluded to.previously, and earlier unsuccessful attempts by one of the authors to develop econo- metric models of electricity demand for Alaska.3 A-15 ENDNdTES: APPENDIX A 1. Lester Taylor, "The Demand. for Electricity: A Survey," Bell Journal of Economics, Spring 1975,. Vol. 6, No. 1, p. 10~ 2. Robert Shaw, Jr., "New Factors in Utility Load Forecasting," Public Utilities Fortnightly, July 19, 1979, p. 19. 3. Scott Goldsmith, "Future Electricity Requirements in Alaska," paper presented at Western Economics Association Annual Meetings, San Francisco, California, June 1976. APPENDIX B. THE ECONOHIC PROJECTION METHODOLOGY Introduction The projections of end-use energy requirements are based on pro- jections of economic activity in the state and its railbelt region. This eco?omic projection provides estimates of employment, population, households, and housing stock for the state and the major regions in the railbelt. These projections are provided for five-year periods through- out the projection period; the projection period is through 2010. The main component of the economic projection methodology is the Man-in-the-Arctic Program (MAP) statewide econometric model, which is used to project the employment, population, and fiscal variables. In addition to this major component, three ne-.:..r components have been deve- loped for this study: a household formation model, a regional allocation model, and a housing stock model. The household formation model estimates the number of households in the state based on the MAP model population projection. The regional allocation model allocates the major variables-- population, employment, and households--to the study regions. Finally, · the housing stock model projects the distribution of housing by type for each region of interest. This appendix provides a detailed description of each of these components. The MAP Economic Hodel The statewide econometric model developed by the Han-in-the-Arctic Program (MAP) at the University of Alaska's Institute of Social and Economic Research is the principal model used in the projection of economic activity for the end-use forecast. The model consists of three interrelated components: an economic model, a fiscal model, and a demo- graphic model. The basic structure of the model is shown in Figure B.l. The economic model divides the economy into exogenous or basic sectors and endogenous or nonbasic sectors. The level of output in the exogenous sectors is determined outside th~ state's economy. The level of output in the nonbasic sectors of the economy is determined within the Alaska economy since their primary purpose is to serve local Alaska markets. The basic industries in the model are mining, agriculture-forestry- fisheries, manufacturing, federal government, and the export component of construction and transportation. The nonbasic industries are whole- sale and retail trade, finance-insurance-real estate, services, communi- cations, utilities,·and the remainder of construction and transportation. Incomes, output, and employment are simultaneously determined in the model. The demand for local economic production is determined by the level of real disposable income in the economy. The level of indus- trial production determines the demand for labor; employment is that level which is needed to produce the required output. The labor demand is always satisfied since the Alaska labor market is assumed to be open to B-2 .. ' I.~UTS~DE ·L ~. PERSONAL TAXES IND'{JStz'lUAL ?RODUC'l'ION . . 1 ,. · DIS1?0SN3LE "' . ·' .. '· I' •• . . . .. .. . i PERSON~L ~~~-------·~ .... . ... , . . .. . ; \v~GB . R."a.TES It t'll\GES AND Sl\LARIES J?U:RSONAL. .. - . . ·, .. ·--.. - - - . ., li' . ~.-________ _,.,...f'-_.,.J-~--· .., . w ~. . .. .------'--.........__...K-t ·I FU~lO STATE R!Wt:::NOES ~O>A' ..... -•• • I '1-~CCUXULATION ...... . . &.-:------J\ /[ .............. OUTS ~DE FORCES ww \!t .. 'I! CONSU!I.ER PRICES . . . . . u.s. ?RICES ,~; . . . IN CO HE >%:1 H' ~ t:.d . 1-' . migration from the rest of the United States. Because of this, both the supply and price of labor (wage rate) are linked to outside activity; wage rates in Alaska are determined in part by wages outside the state. Once wage rates and employment are determined, aggregate wages and salaries are known. The level of disposable income is estimated by ·adding an estimate of nonwage income to wages and salaries and reducing this by the level of income taxes. Real disposable income is found by deflating the level of disposable income by a relative price index; the major determinants of Alaska prices are U.S. prices, the size of the economy, and the growth rate of the economy. The level of population projected is based on a projection of each of its components: births, deaths, and migration. The model uses age- sex-race specific survival rates and fertility rates to project the births and deaths of the civilian population;·· the number of births net of deaths determines the natural increase in the populatio~. Total civilian population is found by adding net civilian migration to the natural increase. Net migration is determined by the relative economic opportunities in the state. Economic opportunities are described in the model by the change in employment and the Alaska real per capita income relative to the U.S. average. Total population is determined by adding an exogenous estimate of military population. The final component of the MAP model is the fiscal model. This model describes the activity of the state and local governments. The fiscal model calculates the level of personal tax payments which are B-4 necessary for projecting disposable income. The fiscal model, based on an assumed state spending rule, also calculates personnel expenditures, the level of state government employment, and the amount spent on capital improvements. The amount the state government spends on capital improve- ments partly determines the level of employment in the construction industry. All three model components are linked by their requirement for information produced by the other compo?ents. A description of the model can be found in Goldsmith, Man in the Arctic Program Hodel Docu- mentation (1979). The model has been updated in connection with the current study. This updating included the reestimation of important economic component equations with data series which ·included the most recent information. In most cases, this consisted of including the 1978 data in the series; however, some data series wer~ recalculated based on recent information. These data changes had only marginal effects on the equations; however, several equations were respecified. The primary reason for these speci- fication changes was to capture the buoyancy of the economy in the post- pipeline period. Equations describing wage rates in the exogenous sector and output in the endogenous sector were respecified in an attempt to improve the model's description of post-boom periods. An additional change was mad0 to the fiscal model. The fiscal. · model was changed to incorporate recent changes in state tax laws which eliminated the state income tax for residents with over three years in the state. The recent permanent fund distribution program was also reflected in model changes. B-5 Household Formation Model MODEL DESCRIPTION The primary unit on which projections of residential energy con- sumption are based is the household.· Households are living units. They are distinct from families and can contain more than one family as well as unrelated individuals. This section describes the model developed for this study to project the number of households in the state. The population projections determine the number of households in the state. The number of households is a function of both the level of population and its age-sex distribution. The age-sex distribution of the population is important because the rate at which people form house- holds differs across age-sex cohorts. The model described below accounts for both of these influences of population on household formation. The household formation model is an accounting model which depends on a set of assumptions about the cohort specific rate of household forma- tion and changes in those rates. The model is based on the assumption that the social, economic, and life cycle factors which determine the formation of households can be described by ~ set of household formation rates. Household formation rates describe the probability that a person in a particular cohort is a household head. The model requires input from the MAP population model in the form of the projected size and age-sex distribution of the population. The B-6 total number of households in the state (HH)· is equal to the number of households summed across age and sex cohorts. HH = L~ HH •. ~J ij (B.l) The total number of households in sex cohort i and age cohort j (HH .. ) describes the number of households with household hea4 or primary ~J individual in the ith sex and jth age cohort. The total number of house- holds in cohort ij equals the number of civilian non-Native households in cohort ij (CHH .. ) plus the number of Native households in cohort ij ~J (NHH .. ) plus the number of military households in cohort ij (MHH .. ) • ~J ~J HH .. ~J CHH .. + NHH •• + MHH .. ~J ~J ~J The number of civilian and Native households in each cohort is a (B.2) function of the population and household formation rate for the cohort. The number of households in any cohort equals the cohort specific house- \ . hold formation rate (HHR .. for civilian non-Natives and NHHR .. for Natives) ~J ~J multiplied by the total population (CNNP .. for civilian non-Natives and ~J NATP .. for Natives) net of the population in group quarters (CPGQ .. for ~J ~J civilian non-Natives and NPGQ .. for Natives). ~J CHH. ~ = (CNNP .. -CPGQ .. ) * HHR .. ~J ~J ~J ~J (B.3) NHH .. = (NATP .. -NPGQ .. ) * NHHR.'. ~J ~J ~J ~J (B.4) B-7 Household formation rates describe the probability that someone in a particular cohort is the head or primary individual of a: household. · These rates change through the projection period. The initial rates are those found in 1970 (HHR:~70 for civilian non-Natives and NHHR:~70 1] 1] for Natives). (B.S) ·(B. 6) Household formation rates are assumed to change at a constant yearly rate (CHHR .. for civilian non-Natives and NCHHR .. for Natives). So the 1] 1] household formationrate inany year equals the ratein the previous year times the rate of change. HHR .. = HHR .. (-1) * CHHR .. 1] 1] 1] (B. 7) NHHR .. = NHHR .. (-1) * NCHHR .. 1] 1] 1] Finally, the cohort distribution of military households is assumed to rema.in constant throughout the projection period. The number of military households (MHH .. ) equals the number in 1970 (MHH:~70 ) times 1] 1] the percentage of 1970 military population in the state (MILPCT). MHH .. = MHH~~70 * MILPCT 1] 1] B-8 PARAMETER ASSUHPTIONS The most important source of variation in the model results is the assumed rates of household formation. These rates have been subject to dramatic changes in recent history. The change in these rates can be judged by examining recent changes in the average household size. Between 1960 and 1970, the average household size in the United States fell by 6 percent from 3.33 pe?ple per household to 3.14 people per household. Between 1970 and 197&, the rate of decrease was almost twice as fast as in the previous period; the average household size fell by more than. 10 percent from 1970 to 2.81 (Bureau of the Census, Projections of the Number of Households and Families: 1979 to 1995, 1979). Alaska is experiencing a similar trend, the average household size fell by almost 7 percent between 1970 and 1976 (Bureau of the Census, Demographic, Social, and Economic.Profile of States: Spring 1976, 1979). The important recent changes in household formation rates, illus- trated by the recent changes in household size, and the lack of agreement by population experts on future changes in these rates make the selection of any particular set of parameters probabilistic. Two sets of para- meter assumptions are required by the model, household formation rates arid yearly changes _in those rates. Table B.l. presents the initial set ·of household formation rates. Thes.e rates are derived from the 1970 census and equals the number of household heads per population in households in the cohort. These rates were adjusted in some cohorts to insure consistency with U.S. rates in 1970. B-9 TABLE B.l. 1970 ALASKA CIVILIAN POPULATION HOUSEHOLD FORMATION RATES (HHR .. ) . l.J NON-NATIVE NATIVE Male Female Male Female 0 - 1 0 0 0 0 1 -5 0 0 0 0 5 -9 0 0 0 0 10 -14 .001 .001 .003 0 15 -19 .040 .018 .017 .006 20 -24 .583 .107 .238 .069 25 -29 .900 .109 .576 .082 30 -34 .933 .117 .746 .095 35 -39 .955 .126 .881 .119 40 -44 .962 .133 .894 .120 45 -49 .963 .148 .907 .139 50 -54 .964 .164 .922 .149 55 -59 .956 .207. .947 .296 60 -64 .956 .245 .926 .313 65 + .885 .320 .816 .385 SOURCE: Bureau of the Census, 1970 Census of Population Detailed Characteristics: Alaska, 1972, Table 153. B-10 Table B.2. illustrates the second set of required parameter assump- tions, the assumed yearly change in the household formation rates. These changes are based on the changes implicit in the Series B projections of households by age and sex cohorts prepared by the Bureau of the Census (Bureau of the Census, Projections of the Number of Households and Families: 1979 to 1995, 1979). ·The average yearly change in household formation rates by age-sex cohort for the nation were assumed to hold for the civilian non-Native population. It was assumed that Native household fordation rates would not change as rapidly; Native household formation rates were assumed to change at rates which would provide half the change in average household size projected for the nation. B-11 TABLE B.2. YEARLY PERCENT CHANGE IN HOUSEHOLD FORMATION RATE (CHHR .. ) ~J NON-NATIVE NATIVE Male Female Male Female 0 - 1 0 0 0 0 1 -5 0 0 0 0 5 -9 0 0 0 0 10 -14 1.002 1.045 1.001 1.028 15 -19 1.002 1.045 1.001 1.028 20 -24 1.002 1.045 1.001 1.028 25 -29 1.000 1.045 1.002 1.028 30 -34 1.001 1.040 1.001 1.024 . 35 -39 1.000 1.027 1.000 1.016 40 -44 1.000 1.027 1.000 1.016 45 -49 1.001 1.012 1.000 1.006 50 -54 1.001 1.012 1.000 1.006 55 -59 1.001 1.000. 1.000 1.000 60 -64 1.001 1.000 1.000 1.000 65 + 1.001 1.000 1.000 1.000 SOURCE: Bureau of the Census, Current Population Reports Series P-25, No. 805, Projections of the Number of Households and Families, 1979 to 1995, May 1979. B-12 Regional Allocation Model MODEL DESCRIPTION A method for making substate regional projections was required by this study. The economic and household projections described above are made at the state level. These projections serve as the basis for the energy demand projections; however, the Susitna project will provide energy for only a portion of the state, the railbelt region. This section describes the regional allocation model used in this study to allocate statewide projections to the region of interest. Methods of projecting substate regional economic activity range from the simple to the complex. The simplest methodology is to allocate state economic activity to the region based on its historical share or to allocate nonbasic activity based on the regional share of basic sector activity. The most complex approach involves the estimation of complete regional models. The first approac~ suffers from its simplicity; it fails to recognize the importance of changes in the structure of the regional economy over time as the economy grows. The·more complex approach requires massive commitments of time and resources to develop. It may also suffer from a lack of consistent data in the regions, par- ticularly in areas like Alaska where many of the regions have small, undeveloped economies. In choosing a regional projection technique, we were interested in three things. First~ we wanted a model which was simple and efficient B-13 ... to use and could be used to project activity in a number of regions. Secondly, we wanted a model which made maximum advantage of the short·. data series available in the regions. Finally, we wanted a model which provided results consistent with the state projections. The method used in this study is a regional shares model. In this model, the regional shares of st~te support sector employment, state and local government employment, and population are estimated econometrically as a function of basic sector activity as well as proxies for comparative advantage and scale of the regional economy. A pooled time series cross- section approach is used to estimate the model. This econometric approach has two purposes. First, it allows us to make us.e of the short data ·Series in the census divisions. Secondly, it allows us to capture the major variability in the regional shares of economic activity which is across regions rather than over time. Traditional explanations of regional economic growth explain growth as a function of growth in the region's basic sector. The regional allocation model recognizes that the local support sector response depends not only on basic sector growth but also on the position of the region in economic space •. Larger economies will provide a greater support sector response since they offer economies of scale and produce more goods and services locally. Regions may also respond differently if they provide support sector services to. regions other than their own. These trade centers have a comparative advantage in producing these goods and services. The scale and comparative advantage effects B-14 . . are accounted for in this model by the use of lagged population and regional dummies in each equation. The model consists of four equations which estimate the regional share of population, support sector employment (in two categories), and state and local government employment. State and local government employ- ment is assumed to be allocated across regions in the state primarily to serve the population, so the share of state and local government employ- · ment in region i (REGSL.) is primarily a function of lagged share of ~ population in the region (LRPOP.). Special characteristics such as the ~ capital in Juneau and administrative centers in larger regions are accounted for through the use of regional dummy variables (D.). ~ REGSL. = F1 (LRPOP.,D.) ~ ~ ~ (B.lO) Support sector economic activity is disaggregated into two distinct sectors: direct support, which includes construction and transportation employment, and other support which includes trade, services, finance, utilities, and communication. This distinction is a function of the assumed relation between these sectors and basic activity and assumed differences in causes of growth in each sector. Direct support sector employment depends not just on the size of the community or its basic sector, but on the growth of the community. This is a result of con- struction employment serving mainly an investment function. Because of this, the share of direct support sector employment in region i (RESA.) ~ is a function not only of the size of the community (LRPOP.) but also ~ B-15 the change in the economy; this change is described by the lagged share of the change in total employment in region i (L298.). Uniqueness of ~ regional economies is captured by the inclusion of a regional dummy (D.). ~ RESA. = F2 (LRPOP.,L298.,D.) ~ ~ ~ ~ (B.ll) Employment in the other support sector is assumed to be a function of the level of basic sector activity in the region. For our purpose, basic sector is defined broadly to include traditional basic sector in- dustries as well as local and state government, and direct support sector employment in region i (RESB.) may differ from a direct relation to the ~ regional share of basic sector employment (RR3EB.) for two reasons. ~ First, the support sector may not expand immediately because of lags or because the basic sector growth is temporary. Secondly, the region may provide support sector services to a large region and be related to the basic sector activity in those regions. To account for these effects, both the lagged share of population (LRPOP.) and regional dummies (D.) ~ ~ were included. RESB. = F3 (RR3EB. ,LRPOP. ,D.) ~ ~ ~ ~ Finally, the regional share of population (RPOP.) was assumed to be . . ~ a function of employment in the region. In Alaska, workers often travel to jobs in other regions; this is most important in basic sector activi Because of this, the effects of basic and other support sector employment or population w.ere separ.ated. · Population may also differ from employment B-16 since a region may house the families of workers who work in other regions, such as Anchorage providing the homes of families of Prudhoe Bay workers. The regional share of population (RPOP.) was assumed to be a function of ~ the share of support sector employment (RESB.), the share of basic sector ~ employment (RR3EB.), and a regional dummy (D.) which reflects the fact ~ ~ that a region can serve as home to families of workers employed in other regions. Regional totals are found by multiplying state totals by the shares estimated by the model. PARAMETER DESCRIPTIONS These equations were estimated using a pooled time series cross- sectional technique. Data onAlaska labor divisions (similar inmost cases to census divisions) from 1965 to 1976 were used in the estimation. A linear form of the equations was estimated; the primary reason for this choice of functional form was the ability to use the set of equations to ~project the growth of any region defined as an aggregate of census divisions. Conopsak (1978) identifies both the advantages and problems with using this technique. First, the use of pooled data increases the number of degrees of freedom compared to either cross-section or time-series regressions. Second, pooling techniques limit structural change biases which may occur in time ·series. Thirdly, by using a time series of cross-sections, it is possible to measure the effect of time and struc- ture on the relationships. This is particularly important when estimating B-17 a shares model since examining time series of pooled cross-section provides for ample variation in the data. Two types of problems may exist when using this approach. There may be systematic bias of the disturbance term because of cross-sectional influences, or the source of bias may be autocorrelation of the residuals. The first problem may be reduced with the inclusion of the regional dum- mies in the equation. The second problem may be reduced by adjusting for autocorrelation in the regression; both of these corrections were made. Equations B.l3 to B.l6 are the equations used in the model. Dummies were included for all regions; however, only those dummies for the regions in this study are shown (DA-Anchorage, DK-Kenai, DS-Seward, DM-Matanuska, DF-Fairbanks, and DV-Valdez). The high R2 (uncorrected) result from two factors, the inclusion of the regional dummies and the relatively small variability of regional shares in the historical period. REGSL = .246 * LRPOP + .224 * DA + .124 * DF + .025 * DK (3.48)1 (6.99)1 (10.07)1 (3.56)1 + .021 * DM + .009 * DS + .018 * DV (3.15)1 (1.41) (2.78)1 RESA = 1.357 * L298 + .744 * LRPOP + .148 * DA + .045 * DF (4.73)1 (3.44)1 (1.67)1 (1.27) + .025 * DK (2.36)1 .008 * DM (-1.07) .003 * DS + .003 * DV (-.45) (.45) R2 = 987 . R2 = 987 . 1 t statistic in parentheses significant at greater than 95 percent. B-18 RESB = .269 * LRPOP + .086 * RR3EB + .409 * DA + .111 * DF (3.26)1 (1.31) (11.46)1 (7.71)1 (B.l5) + .017 * DK + .007 * DM + .004 * DS + .006 * DV .997 (j.74)1 (1.95)1 (1.19) (1.94)1 RPOP = .290 * RESB + .157 * RR3EB + .213 * DA + .073 * DF (4.70)1 (2.93)1 (5.78)1 (4.58)1 (B.l6) + .029 * DK + .022 * DM + .005 * DS + .012 f DV (7.38)1 (6.81)1 (1.59) (3.45) R2 = .995 Housing Stock Model MODEL DESCRIPTION Regional projections of households and housing stock are the basic components of the residential energy demand projections. This model uses the output of the three components described above to project both the number of households by region and the distribution of.those households by housing type. The housing types projected by this model include single- family, duplex, multifamily, and mobile homes. The total housing stock by type is found by adjusting for vacan·t housing. Housing stock projections are influenced by three factors. First, the number of households determines the aggregate demand for housing. Secondly, the distribution of households by income, family size, and tenure determines the effective demand for different types of housing. Finally, housing has a long life; once a type of housing is built, it 1 t statistic in parentheses significant at greater than 95 percent. B-19 exists for a long time and influences the actual distribution of hous- ing by type. The model described in this section attempts to account for each of these factors. The number.of households in region i (THH.) is found by dividing l. the regional population (POP.) [from the regional allocation model] by l. ; a regional population per occupied dwelling units parameter (PPODU.). l. This provides an estimate of the total number of households in the region. On-base households (BHH.), which are assumed to remain constant l. throughout the period, are subtracted from total households to find total off-base households (HH.). l. THH. = POP./PPODU. l. l. l. HH. = THH. -BHH. l. l. l. (B.l7) Once total off-base households (from hereon, total households) are found, the demands for various housing types are projected through the use of housing type demand coefficients. The demand for housing type T (H:) equals the total housing units (HH.) times the demand coefficient l. ~ for type T (HDT). The demand coefficients describe the distribution of households by "preferred" housing type. HH. * HD: (B. l. l. B-20 The initial stock of housing of type i in any period is equal to last period's housing stock of that type cs:(-1)) minus the removals ~ . from the stock since the previous period. Removals are due to demoli- tions, accidental loss (fire, flood, etc.), and conversions to other types of units or uses. The model finds the initial stock of housing of type T (s:) by multiplying the stock from the previous period times ~ one minus the removal rate (rT), where the removal rate equals the proportion of the previous period housing lost between the periods. (B.20) · Construction of new housing of each type is determined by the net T demand for that type (ND.). The net demand equals the demand for housing ~ T AT of type T (H.) minus the initial supply of that type (S.). ~ ~ T AT =H. -S. ~ ~ (B.21) If the net demand for all types of housing is positive, new construction (ND:) equals net demand plus the equilibrium amount of vac?nt housing. ~ . In this case, new construction equals net demand plus the vacancy rate (VT) times the net demand plus the initial supply of housing type T. (B.22) If the net demand for a particular housing type is negative, an adjustment is required. The adjustment recognizes implicitly the effect B-21 of prices on demand. When net demand for a housing type is negative, this excess supply is assumed to drive down the price of this type of housing relative to others and switch demand. For this adjustment, single-family and mobile homes and multifamily and duplexes are assumed to be close substitutes; when one type has excess supply, it is filled by households with the other type demand. If excess supply continues to exist and the vacancy rate is greater than an assumed maximum rate, the units are filled proportionally from the other types with excess demand. Once these adjust- ments are made, new construction occurs to meet the excess demand. PARAMETER DESCRIPTION There are four sets of parameters which determine the housing stock projections. The assum~d people per occupied dwelling unit (PPODU.) ~ determines the number of households in a given regional population. The removal rates (rT) determine the proportion of last period's housing which has been removed from the housing stock. The vacancy rates (VT) determine the supply of vacant housing in any period. Finally, the housing demand coefficients (HD:) determine the initial distribution ~ for demand by housing type. The initial people per occupied dwelling rates were taken from the most recent information found for e~ch region. Table B.3 shows the initial rates used in each region. These rates were adjusted each period to reflect projected changes in the population per household rate on the state level; this adjustment assumed the changes in the state rate were reflected proportionally in each region. B-22· TABLE B.3. INITIAL PEOPLE PER OCCUPIED DWELLING UNITS 1 Greater Anchorage 3.03 Fairbanks2 3.01 3 Valdez 3.1 4 Rest of State 3.5 ~eighted average of rates found in: Anchorage Municipality, 1978 Population Profile, 1978 (for Anchorage); Kenai Borough, Profile of Five Kenai Peninsula Towns, 1977 (for Kenai and Seward); and Rivkin Associates, Workbook on the Economic and Social Impacts of the Capital Move on Juneau and the Mat-Su Borough, 1977 (for Matanuska-Susitna). 2Assumes Fairbanks people per occupied dwelling decreased at same rate as the U.S. average between 1970 and 1977·; the 1977 rate was .91 less than in 1970 for the United States. 3 . M. Baring-Gould, Valdez City· Census, 1978. 4weighted average for the nonrailbelt area of the state in the 1970 census (3. 7) assumed to d.ecline at one-half the rate of the decline for the United States. An average removal rate is assumed for this report and applied to all types of housing in all regions. Table B.4 shows the."rates assumed in this study. TABLE B.4. ASSUMED HOUSING REMOVAL RATES 1975-1980 1980-1985 1985-1990 1990-1995 1995-2000 1.0% 1.25% 1.50% 1.75% 2.0% B-23 Removal rates are a function of the age of the housing stock and the growth of the region. Older housing may be subject to filtering, and areas with more rapid growth may remove more older housing to make room for new construction. Comparisons of dwelling unit estimates in 1970 and 1979 with building permits da,ta on units constructed during that period show an approximate removal rate of 1 percent per five-year period in Anchorage and Fairbanks. It was assumed that, as the existing stock ages, the removal rate will grow toward the U.S. average which has been estimated to be between two and four percent for a five-year period (deLeeuw, 1974). We assumed the removal rate would reach the lower bound of 2 percent by 2000. Vacancy rate assumptions are based on U.S. averages and recent Alaska -experience. Table B.5 shows the assumed normal and maximum vacancy rates used in the housing stock projection. The normal vacancy rates are the ten-year U.S. averages for owner and renter units (Bureau of the Census, Housing Vacancies: Fourth Quarter 1979, 1980). Single-family and mobile homes have the owner rate; multifamily, the rental rate; and duplexes, the TABLE B. 5. VACANCY RATE ASSUMPTIONS Single Family Multifamily Duplex Mobile Home Normal 1.1 5.4 3.3 1.1 B-24 Maximum 3.3 16.0 10.0 3.3 average of owner and renter. Maximum rates are based on recent Anchorage . multifamily experience (Anchorage Real Estate Research Report, 1979); single-family, mobile home, and duplexes are assumed to maintain the normal relationship to multifamily vacancies. The final parameter assumptions concern the housing demand coeffi- cients; these are the most important parameters for determining the housing stock distributions. The assumed housing type demand distribu- tions used in this study are based on the analysis of existing survey information from Anchorage. A 1978 survey of the Anchorage population conducted by the Urban Observatory .of the University of Alaska -Anchorage (Ender, 1979) provided information on housing type choice and demographic variables. Regression analysis was used to analyze this information. A linear probability model was estimated for each of three housing types: single-family, multifamily, and mobile home. These regressions estimate the probability that a person would occupy a particular housing type as a function of the age of the household head and family size. Family size and incomes have often been isolated as the major determinants of housing type choice. .In reality, current income and wealth influence a household's ability to purchase a home; the age of the household head is a proxy for both wealth and income. Table B.6 shows the results of the Anchorage regressions. These equations were tested by estimating the 1970 housing type distribution in the Anchorage, Fairbanks, and Valdez regions and with more recent data in Fairbanks and Anchorage. The performance of the B-25 TABLE B.6. HOUSING CHOICE REGRESSIONS Single Family SF = .461 -.303 * Sl -.175 * S2 + .08 * S4 + .182 * A2 (70.36)1 (20.52)1 (1.87) (12.24)1 + .317 * A3 + .380 * A4 (47.33)1 (43.85)~ MultifamiU:_ MF = .383 + .225 * Sl + .086 * S2 -~09 * S4 -.203 * A2 (50.75)1 (6.46)1 (3.07) (19.84)1 -.280 * A3 -.352 * A4 (47.96)1 (49.02)1 Mobile Home MH = .097 + .068 * Sl + .039 * S2 + .014 * S4 + .008 * A2 (7.01)1 (1.98) (.121) (.043) .020 * A3 -.016 * A4 (.366) (.151) Family Size Age of Household Head Sl <2 S2 3 S4 5< A2 25-30 A3 30-55 A4 55< -2 R = 1F statistic in parentheses significant at greater than 95 percent. B-26 models in these cases suggested that no specific regional adjustmen·t of the equations was required in Anchorage and Fairbanks, and they were used to project housing demand coefficients in each region. In Valdez, comparison with 1978 housing stock distribution showed a major difference. This was assumed to be a result of the recent rapid growth in the region connected with construction of the Trans-Alaska Pipeline Service (TAPS) pipeline and associated port facility. In Valdez, demand coefficients were assumed to change linearly from the 1978 housing type distribution to the projected 2000 housing demand distribution. Use of these equations assumes that the existing relationships .between housing type choice and non-included variables remains the same. Most important in the case is the effect of housing price. We are _implicitly assuming that the relation between housing prices and income will remain constant throughout the period. The importance of housing means there may be some adjustments such as two-income families. The importance of the government in the housing and mortgage markets makes any change impossible to forecast. This approach also ignores the existence of land-use constraints which may prevent actual construction of these units. The housing type parameters were projected over the period based on assumed changes in household head age distribution in each region and the family size distribution. The state distribution of households B-27. by age of household head was used to estimate the regional distribution; each region was assumed to mainta·in the same relation to the state dis- tribution as in 1970. The family size distribution in each region was assumed to follow a pattern of change which approached the 1977 Western Regional distribution (Bureau of Census, Annual Housing Survey: 1977, 1979). The level of household size in the Western region of the United States was similar to that projected for our regions in 2000. The change from the 1970 distribution to this 1977 distribution was assumed to be at the same rate as the rate of the change in population per household projected by the regional allocation model. B-28 APPENDIX C. ECONOHIC PROJECTIONS The projections of electricity consumption in the Susitna Hydro- electric Power Project service area are based on a set of projections of economic activity in the state and the railbelt region. This appendix describes those projections. The appendix is divided into three parts. First will be a short general discussion of projections, the reasons for doing them, and general problems with projections. A discussion of the specific sets of assumptions which were used in conjunction with the models described in Appendix B follows the first section. Finally, the projections themselves will be reviewed. What are Projections? Projections provide a description of a future level of activity; the economic projections described in tpis appendix describe possible future levels of important economic variables. Projections cannot be assumed to be an accurate description of what will happen but rather a description of what could happen if the assumptions which determine the projections come true. This means that projections are probabilistic. The uncertainty of the .future, though it may increase the problems associated with making projections, increases the importance of projec- tions. Decision makers in both public and private sectors need informa- .tion about the future in order to plan their actions since they must make decisions which both are affected by and affect future events. The more uncertain are the future events, the more important are some projections of them in decision making. All methods for making projections of future economic activity re~ quire assumptions about the future. The simplest projection technique is simply to assume a certain growth for each of the major variables of interest. More complex methodologies employ some form of model to translate assumptions about specific events into projections. Models describe the relationship between variables about which assumptions are made (exogenous variables) and those of which projections are made (endo- genous variables); an important assumption when a model is used is that the relationship described by the model remains constant. The use of models makes explicit the assumptions implicit with simpler projection techniques, and it provides consistency between sets of projections. The major problems with projections is the uncertainty attached to the projections. The uncertainty results because of uncertainty about the future levels of exogenous variables and uncertainty that the rela- tionships will continue to hold as described by the model. There are two major ways to limit the importance of this problem, although un- certainty can never be eliminated from projections. The first measure is to provide a clear, complete description of the assumptions on which the projections are based; this allo"tvs users to know exactly what is behind the projections. The second measure involves producing many alternative projections instead of just one; these alternative projec- tions provide an indication of effect of altering major assumptions. C-2 These measures in themselves do not limit the uncertainty of any par- ticular projection, but they allow the establishment of a range of possible outcomes which the researcher expects to have a very high probability of occurrence. The Approach of the Current Study In the present study, we present a series of projections which, although of limited number, reflect the range of probable future levels of economic activity. Because these projections may be used in the actual design of the pr0ject, it is important to provide a range of futures with a relatively high probability of occurrence. We do not assume that the actual growth in future economic activity will match any of our scenarios in level of activity or timing and magnitude of events. What we assume is that the general level of economic activity described by-these scenarios will occur with a high probability. The projections in this study are for a thirty-year period from 1980 to 2010. Because of the long projection period, there is a large potential for error in utilizing any single technique in developing the projections. Two general techniques have been employed in these projec- tions. The first is-scenario building in which the aggregate values are developed by constructing scenarios composed of specific events. For example, the level of total employment in the petroleum industry would be estimated by .assuming the development of specific reserves with ' - associated manpower requirements and timing. The scenario building C-3 technique reduces the potential for error in the projection by dividing the development of the assumptions into a series of small decisions. This technique allows specific information about future developments to be built into.the projection. The most important reason for using the scenario building technique is that it allows for consistency in the forecast. By assuming specific events, we can make sure the growth rate projected is at least possible. The major disadvantage of the scenario building technique occurs as the projection period is extended in time. Because many of the possible future events .are unknown, they may be ignored when developing the sce- narios. This myopia may result in a downward bias in the projection as the projection period lengthens. To attempt to overcome this problem in these projections, a second technique was used in the post-2000 period when information about speci- fic events is undefined. This second technique is a judgmental approach The judgmental technique projects directly the aggregate variables. The judgmental projection is based on an analysis of both the historic period growth and the growth in other regions. This technique fails to provide reality or consistency checks on the assumptions but provides a method of projection when the scenario building technique is impossible to use. The study combines the scenario building and judgmental techniques to produce the required projections. Each technique is used to project C-4 economic activity in the period when it is most appropriate; the scenario building technique is used in the period between 1980 and 2000 when reasonable information about possible economic events is available and the judgmental approach is used after 2000 when there is only limited information on the possibilities. The Scenarios: 1980-2000 For the period between 1980 and 2000, specific scenarios are developed. These scenarios consist of two major components, an economic scenario and a state government fiscal scenario. The economic scenarios consist of a set of assumptions which describe_the special projects and industrial growth in the period. The state government fiscal scenario describes the assumed level of state expenditures; these expenditures result in the creation of jobs in both state government and the construction industry. Each economic scenario describes the growth in the exogenous indus- · tries: mining, manuf a·c turing, agricul ture-fores try-fisheries, federal t government; and in the.exogenous components of construction and transpor- tation. Each state government scenario describes the growth of state operations and capital expenditure which affects the level of employment in state government and construction. The next two sections describe the economic and state government scenarios. C-5 THE ECONOMIC SCENARIOS The economic scenarios consist of time series on employment and output in certain export base or exogenous industries. These assump- tions are organized into three separate scenarios which describe a high, moderate, and low series of economic events which describe what we feel to be a reasonable range of economic events. This does not mean that we are predicting that all or any of these events will occur since there is a highly variable degree of uncertainty with respect to the levels and timing of a lrtain I I the events in these scenarios. What it does mean is that with degree of probability, we expect the general level of economic abtivity to follow these scenarios. We assume that there is a very high \_ probability that the level of activity will be at least as great as that described by the low scenario, a medium probability that the level of activity will be at least as great as that described by the moderate scenario, and a low probability that activity levels higher than those described in the moderate case will.occur. Primarily as a result of the uncertainty attached to the occurrence, magnitude, and timing of any particular event, agreement about particular scenarios is hard to achieve even among those most knowledgeable about the Alaska economy. Emphasizing our concern mainly with general levels of activity and the probabilistic nature of any specific scenario should reduce the disagreement. In an attempt to reduce even further the dis- agreement, the scenarios were developed based upon existing scenarios which have attained some measure of consensus. The most important source for these scenarios were the scenarios developed in the level B Southcentral Water Study (Scott, 1979). C-6 The individual scenarios are described in Tables C.l through C.3. The assumptions are described below; these discussions are organized by industry. Mining Currently, the mining sector in Alaska is dominated both in terms of employment and output by the petroleum industry. This is assumed to continue in the future in all scenarios. All three scenarios include production at Prudhoe Bay and in the Upper Cook Inlet. Production from the Sadlerochet formation at Prudhoe is assumed to include both primary recovery and secondary recovery using water flooding. The Kuparak formation is also assumed to be developed with production rising to 120,000 barrels per day by 1984. Employment associated with these developments peaks in the early 1980s "tvith the development of Kuparak and the water flooding project. Upper Cook Inlet employment is assumed to remain at its existing ievel throughout the projection period. This assumes a rising level of exploration, develop- ment, and production of gas in the Kenai fields which would replace employment lost because of declining oil production~ The major new source of petroleum production assumed in these scenarios is Alaska's Outer Continental Shelf (OCS). Alaska is the area of primary importance to future OCS activity. Nearly 60 percent of oil and 40 percent of gas resources which are expected to be found in the United States OCS are expected to be in Alaska waters (Bureau of Land C-7 (") I 00 Special Projects Trans-Alaska Pipeline Northwest Gasline Prudhoe B.ay Petroleum Production Upper Cook Inlet Petroleum Pro- duction TABLE C.l. HIGH SCENARIO ECONOMIC ASSUMPTIONS Description The construction of the TAPS was com- pleted in 1977. Additional construc- tion of four pump stations is assumed as well as pipeline operations. Construction of natural gas pipeline from Prudhoe Bay which in- cludes construction of an associated gas conditioning facility o~ the North Slope. Primary recovery from Sadlerochit formation, secondary recovery using water flooding of that formation and development of the Kuparuk formation. Employment associated with declining oil production is assumed to be replaced by employment associated with rising gas pro- duction maintaining current levels of Dates & Employment 1979-1982 -Pump station construction of 90/year 1977-2000 -Operations employment of 1000/yr. 1981-1985 -Construc- tion peak employment of 7,823 (1983) 1986-2000 -Operations begin employing 400 petroleum and 200 trans- port workers 1982-1984 -Construction of water flooding pro- ject peak employment of 2,917 (1983) Railbelt Location Operations employ- ment allocated: 1/3 to Valdez 1/3 to Fairbanks 1/2 of construc- tion and trans- portation employ- ment in Fairbanks 1980-2000 -Mining employ- ment long-run average of 1,802/year 1980-2000 -Mining em- ployment of 705/year All in Anchorage region Source E. Porter, Bering-Norton Statewide-Regional Economic and Demographic Systems, Impact Analysis, Alaska OCS Socioeconomic Studies Program, Bureau of Land Management, 1980. E. Porter, 1980. E. Porter, 1980. E. Porter, 1980 n I \0 Special Projects National Petro- leum Reserve in Alaska State Capital Move Beluga Coal Production TABLE C.l. HIGH SCENARIO ECONOMlC ASSUMP~IONS (cont.) · Description Petroleum production in NPRA. Production in five fields with a total reserve of 2.5 billion bbls equiva- lents of oil and gas. Construction of 525 miles of pipeline. Movement of the state capital from Juneau to Willow begins in 1983. A full move involving 2,750 state employees. Major development of Beluga coal reserves for export. Dates & Employment Leases held between 1983-1990. Develop~ ment and .exploration begins in 1985. Average mining employ- ment of 460/year. ' . ' 1983-1996 -Construe- Railbelt Location · All in Anchorage tion -peak employment region in 1990 of 1,560/year. · Move completed in 1996~ · 1985-1990; 1994 -Con-Located in struction with peak Anchorage region employment of 400 (1987) 1988-2000 -production employment of· 370/yr. for long-run average Source Based on mean scenar~o under Management plan 2 in Office of Minerals Policy and Research Analysis, U.S. Department of Interior, Final Report of the 105(b) Economic and Policy Analysis, 1979. High Scenario in M. Scott, Southcentral Alaska's Economy and Population, 1965-2025: A Base Study and Projections, Economics Task Force, Alaska Water Resources Study (Level B), 1979. Pacific Northwest Labora- tory, Beluga Coal Field Development: Social Effects and Management Alternatives, 1979. ., {") I 1-' 0 TABLE C.l. HIGH SCENARIO ECONOMIC ASSUMPTIONS (cont.) ·special Projects Outer Continen- tal Shelf Petroleum Pro- duction Description Production in eleven OCS lease sale areas: Beaufort 1 (1979) Lower Cook (1981) . Bering-Norton (1982) St. George 1 (1982) North Aleutian (1983) Beaufort 2 (1983) Navarian Basin (1984) Hope Basin (1985) Chukchi Basin 1 (1985) Navarian Basin 2 (1989) Chukchi Basin 2 (1994) U.S. Borax Mining Development of mining operation. Alpetco Project Major petrochemical facility developed as originally pro- posed by Alpetco. Dates & Employment Peak OCS employment -mining -9,066/year (2000) -construction -5,300 /year (1992) Exploration and devel- opment begins in 1980. Long-run mining em-· ployment of 440/year begins in 1993. · Railbelt Location Lower lease· sale (68) is in Anchorage region. Headquarters employment averag- ing 12 percent of mining is in Anchorage 1982-1986 -Construe-Located in Valdez tion -peak employment region of 3,500/year (1984- 1986) 1987-2000 -Operations employment of 1,925/ year Source E. Porter, 1980 {for Lower Cook and Bering- Norton lease sales). Employment scenarios for remainder of sales esti- mated based on N. Gulf (Sale 55) high case ad- justed to include LNG plant (Huskey and Nebesky, Northern Gulf Petroleum Scenarios: Economic and Demographic Systems Impacts, Socioeconomic Studies Program, Alaska OCS Office, 1979). Northern Gulf Scenario was apjusted by difference in resource estimates to produce scenarios for specific areas. U.S.D.A. Forest Service, E.I.S.: U.S. Borax Mining Access Road for Quartz Hill Proposal, 1977. S. Goldsmith and L. Huskey, The Alpetco Petrochemical Proposal: An Economic Impact Analysis, Institute of Social and Economic Research, 1978. I Special Projects Pacific LNG Project Forestry/Pulp and Paper Manufacturing n Other Manu-1 1-' facturing .... Federal Govern- ment Fairbanks Petro- chemical TABLE C.l. HIGH SCENARIO ECONOMIC ASSUMPTIONS (cont.) Description .Const~uction of cur- rent proposal by Pacific LNG Employment in these industries expands to accommodate an annual cut of approximately 1.3 million board feet by 2000. Expansion of existing manufacturing as well as new local manu- facturing of locally~ consumed goods. A doubling of the re- cent growth rate of civilian federal government. Military government employment assumed to remain con- stant at 1978 level. Moderate.petrochemical facility using the state's royalty gas as feedstock. Dates & Employment 1982-1985 -Construc- tion peak employment of 1,323/year (1984) 1986-2000 -Operations employment of 100/yr. Growth of output at 3% a year Civilian federal government employ- ment grows at 1% per year 1984-1986 -construc- tion of 1,500/year 1987-2000 -operation employment of 600/yr. Railbelt Location Located in Anchorage region Source E. Porter, 1980. Approximately 11% M. Scott, 1979 of this activity occurs in th~ Fairbanks region 81% in Anchorage region, 15% in Fairbanks region and .4% in Valdez region • 56% of civilian employment in Anchorage, 15% in Fairbanks, .3% Regional distribution based on existing distribution of employment M. Scott, 1979. in Valdez. Military employment as in 1978. :Located in Fairbanks region J. Kruse, Fairbanks Petrochemical Study, 1978. (") I .... N 7 Industry Assumptions Other Mining Agriculture Fisheries/ Food Processing TABLlf C .1. JIIGH SCENARIO ECONOMIC ASSUMPTIONS (cont.) Description Assumed expansion of hardrock and other mining opportunities in the state. Major development of agriculture in Alaska. Reflects favorable state and federal policy and favorable markets. Continued level of em- ployment in existing fisheries. Major de- velopment of bottom- fishing, with 100% replacement of foreign fishing effort in 200 mile limit by 2000. Dates & Employment Growth of employment at 1% a year. Start- ing at existing level. Employment in agri- culture reaches about 4,600 by 2000. Employment in fish- eries increases to 1,350 by 2000. Fiah hatchery and proces- sing plant construc- tion employment averaging 150/year. Appropriate expansion of food processing industry. Railbelt Location Growth is assumed to be distributed across regions as existing mining employment. (67% in Anchorage and 2% in Fairbanks) Source Major emphasis in M. Scott, 1979. Tanana Valley. 71% of growth in Fairbanks and 18% in Anchorage. .14% of existing fisheries and 8% of bottom- fish development in Anchorage; .1% of exist- ing fishery in Valdez. M. Scott, 1979. M. Scott, "Prospects for a Bottom- fish Industry in Alaska, 1' Alaska Review of Social and Economic Conditions, 1980. ("') I ..... w Special Projects Trans-Alaska Pipeline Northwest Gasline Prudhoe Bay Petroleum Production Upper Cook Inlet Petroleum Pro- duction ,1'_ .- TAaLE C.2. MODERATE SCENARIO ECONOMIC ASSUMPTIONS Description The construction of the TAPS was com- pleted in 1977. Additional construc- tion of four pump stations is assumed as well as pipeline operations. Construction of natural gas pipeline from Prudhoe Bay which in- clude!') construction of an associated gas conditioning facility on the North Slope. Primary recovery from Sadlerochit formation, secondary recovery using water flooding of that formation and development of the Kuparuk formation. Employment associated with declining oil production is assumed to be replaced by employment associated with rising gas pro- duction maintaining current levels of employment. Dates & Employment 1979-1982 -Pump station construction of 90/year 1977-2000 -Operations employment of 1000/yr. 1981-1985 -Construc- tion peak employment of 7,823 (1983) 1986-2000 -Operations begin employing 400 petroleum and 200 trans- port workers 1982-1984 -Construction of water flooding pro- ject peak employment of 2,917 (1983) Railbelt Location Operations employ- ment allocated: 1/3 to Vq.ldez 1/3 to Fairbanks 1/2 of construc- .tion and trans- portation employ- ment in Fairbanks 1980-2000 -Mining employ- ment long-run average of. 1,802/year 1980-2000 -Mining em- ployment of 705/year All in Anchorage region Source E. Porter, Bering-Norton Statewide-Regional Economic and Demographic Systems, Impact Analysis, Alaska OCS Socioeconomic Studies Program, Bureau of Land Management, 1980. E. Porter, 1980. E. Porter, 1980. E. Porter, 1980 Special Projects National Petro- leum Reserve in Alaska Petroleum Production Outer Continental Shelf Pe troleum Production Beluga Coal Pro- duction TABLE C.2. MODERATE SCENARIO ECONOMIC ASSUMPTIONS (cont.) Description Petroleum production in NPRA. Production in two fields with total reserves of 1.2 billion barrels equi- valents of oil and gas. Construction of 266 miles of pipeline. Production in six OCS lease sale areas: Beaufort 1 (1979) Lower Cook (1981) Beaufort 2 (1983) Navarian Basin 1 (1984) Hope Basin (1985) Chukchi Basin (1994) Moderate development of Beluga coal re- source for export. Dates & Employment Leased between 1995 and 2013. Exploration and development begins in 1998. Average mining employment of 286 (between 1998-2000). Peak OCS employment -mining -4,900 (1996) -construction -3,300 (1992) 1985-1990 -construc- tion -peak employment of 400 (1987) 1988-2000 -operations employm e nt of 210/ye ar 1on R -run av era~e Railbelt Location Lower Cook lease sale in Anchorage region. Head- quart e rs employ- ment averaging 12% of OCS mining employment Located in Anchorage region Source Based on mean scenario under Ma nagement Plan 4 in Office of Miner a ls Policy and Research Analysis, U.S. Dept. of In t erior, Final Re port of the 105(b) Econo- mic and Policy Analysis, 1979. E. Porter, 1980 (for Lower Cook and Bering- Norton lease sales). Employment sc e narios for remainder of s a l e s esti- mated based on N. Gulf (Sale 55) high case ad- justed to includ e LNG plant (Huskey and Nebesky, Northern Gulf Petroleum Sc e narios: Economic and Demographic Systems Impacts, Socioeconomic Studies Program, Alaska OCS Office, 1979). Northern Gulf Scenario was adjusted by difference in resource estimates to produce scenarios for specific areas. Pacific Northwest Labora- tory, Beluga Coal Field . Development: Social Effects and Management Alternatives, ' 1979. n I ...... 1.11 Special Projects Alpetco Project Pacific LNG Project Industry Assumptions Other Mining Agriculture Fisheries/ Food Processing / TABLE C. 2. MODERATE SCENARIO ECONOMIC ASSUMPTIONS (cont.) Description Development of modi- fied Alpetco proposal; configuration is pri- marily as a refinery rather than petro- chemical operation. Construction of cur- rent proposal by Pacific LNG No expansion of exist- ing non-special pro- ject mining. Assumes that a rela- tively low priority is given to agriculture development because of priorities for recreation and wilder- ness or the lack of markets. Maintenance of current levels of employment ·in existing fishery. Expansion of bottom- fishery to replace one-half of foreign fishery in the 200 mile limit. Dates & Employment 1982-1984 -Construc- tion employment of 900/year 1985-2000 -operations employment of 518/yr. 1982-1985 -Construc- tion peak employment of 1,323/year (1984) 1986-2000 -Operations employment of 100/yr. Employment constant at 1979 level, 2,350/yr. Employment grqws to 1,037 by 2000. Employment in .fisheries increases to 1,228 by 2000. Construction of hatchery and processing facilities employs 75/ year. Approriate ex- pansion of food pro- cessing industry •. Railbelt Location ~ocated in Valdez region Located in Anchorage region Regional allocation ~onl!>tant (67% in Anchorage and 2% in Fairbanks) 71% located in Fairbanks region and 18% in Anchorage region 14% of existing fisheries and 8% of bottom- fishery in Anchorage; .1% of existing . fishery in.Valdez Source E. Porter, 1980. E. Porter, 1980. M •. Scott, 1979. ~· Scott, 1979. M. Scott "Prospects for a Bottom- fish Industry in Alaska," Alaska Review of Social and Economic Conditions, 1980 •. Industry Assumptions Forestry/Pulp and Paper Manufacturing Otner Manu- facturing Federal Govern- ment TABLE C. 2. MODERATE SCE~ARIO ECONOMIC ASSUMPTIONS (con t • ) Description Employment expands to accommodate 960 mil- lion board feet of lumber. Expansion of existing manufacturing of locally consumed goods. Civilian employment assumed to grow at recent historical rate. Military constant at current level. Dates & Employment Growth of output at 2% per year. Civilian employment grows at .05%/year Railbelt Location Source Approximately 11% M. Scott, 1979. of activity in Fairbanks region. 81% in Anchorage, 15% in Fairbanks, .4% in Valdez 56% of civilian employment in Anchorage, 15% in Fairbanks, .3% in Valdez Regional distribution based on existing di$tribution of employment. M. Scott, 1979. Special Projects Trans-Alaska Pipeline Northwest Gasline · Prudhoe Bay Petroleum Production· Upper Cook Inlet Petroleum Pro- duction -~ ' ·~ .,. ': .I • TABLE C.3 •. LOW SCENARIO ECONOMIC A~~UMPTIONS Description· The construction of the TAPS was com- pleted in 1977. Additional construc- tion of four pump stations is assumed as well as pipeline operations. Construction of natural gas pipeli~e from Prudhoe Bay which in- cludes construction of an associated gas conditioning facility on the North Slope. Primary recovery from Sadlerochit formation, secondary recovery using water flooding of that formation and development of the Kuparuk formation. Employment associated with declining oil production is assumed to be replaced by employment associated with rising gas pro- duction maintaining current levels of em- ployment. Dates & Employment 1979-1982 -Pump station construction of 90/year 1977-2000 -Operations employment of 1000/yr. 1981-1985 -Construc- tion peak employment of 7,823 (1983) · . 1986-2000 ~ Operations begin employing 400 petroleum and 200 trans- port workers 1982-1984 -Construction of water flooding pro- ject peak employment of 2,917 (1983) Railbelt Location Operations employ- ment allocated: 1/3 to Valdez 1/3 to Fairbanks 1/2 of construc- tion and trans- portation employ- ment in Fairbanks 1980-2000 -Mining employ- ment long-run average of 1,802/year 1980-2000 -Mining em- ployment of 705/year All in Anchor~ge region Source E. Porter, Bering-Norton Statewide-Regional Economic and Demographic Systems, Impact Analysis, Alaska OCS Socioeconomic Studies Program, Bureau of Land Management, 1980. E. Porter, 1980. E. Porter, 1980. E. Porter, 1980 (') I 1-' 00 Industry Assumptions Other Mining Agriculture Fisheries/Food Processing Forestry/Pulp and Paper Manufacturing Other Manu- facturing Federal Govern- ment TABLE C.3. LOW SCENARIO ECONOMIC ASSUMPTIONS (cont.) Description Assumed reduction in mining employment in the state as a result of land policy or world market conditions. Unfavorable conditions for agricultural de- velopment. These in- clude land policies as well as lack of markets. Agriculture disappears in Alaska. Existing fishery is maintained but no bottomfish develop- ment occurs. Employment expands to accommodate 960 mil- lion board feet of lumber. Expansion of existing production for local markets. Civilian employment assumed to grow at recent historical rate. Military constant at current level. Dates & Employment Mining employment de- clines at 1% per year from existing levels. Employment in agricul- ture declines to zero by 1992. Employment remains at 1000. Moderate growth in food processing to accommodate expanding catch. Growth in output at 1% per year. Civilian employment grows at .05%/year Railbelt Location Decline distributed across regions as existing mining employment. 14% in Anchorage and .1% in Valdez Source M. Scott,· 1979. M. Scott, 1979. Approximately 11% M. Scott, 1979. of activity in Fairbanks region. 81% in Anchorage, 15% in Fairbanks, • 4% in Valdez. 56% of civilian employment in Anchorage, 15% in Fairbanks, .3% in Valdez. Regional distribution based on the existing employ- ment distribution • M. Scott, 1979. ; _._ ., - Management, 1980). The present scenarios are constructed around the present lease schedule, although the projected probability of finding oil in each area is considered. For areas with large reserves; we assumed more than one sale would be held. Table C.4.1 describes the lease sales, their assumed lease date, the assumed level of resources developed, the probability of finding oil or gas, and the scenarios in which they are included. The low scenario assumes no OCS development in the period prior to 2000; this is a result of assumed environmental and legal challenges in the Beaufort Sea, lack of technology and market conditions for the major resource areas in l-lestern Alaska, and only limited resource finds in the Gulf of Alaska. In the moderate scenario, only the most probable areas are developed. The high scenario includes both more areas and a second round of sales in some areas in the moderate scenario. Although five sales are scheduled for the Southcentral region of the state, the probability of finding resources in all of them is extremely low; only the second Lower Cook sale is assumed to be developed in the high and moderate scenarios. Both the moderate and high scenarios also include petroleum develop- ment in the National Petroleum Reserve in Alaska. In the high scenario, five fields are developed beginning in 1983 and extending through the period. These fields contain reserves of 2.5 billion barrels of oil equivalents in oil and gas. Pipelines are constructed to bring the resources to the Trans-Alaska Pipeline Service (TAPS) pipeline and to the Northwest gas pipeline. In the intermediate case, two fields with approximately half the reserves of the high case are developed. This development does not occur until near the end of the period in 1998. C-19 TABLE C.4.1. FUTURE OCS ACTIVITY Lease Sale Area Reserves 1 Risk Factor 1 Scenario Oil Gas (Probability (H = high, (billion (trillion of finding M = moderat barrels) cu. ft.) no resources) L = low) Beaufort Sea .04 Sale 1 (1979) .75 1.6 M,H Sale 2 (1983) .75 1.6 M,H Northern Gulf (1980) .s 1.3 .95 Lower Cook (1981) .2 .5 .95 M,H Bering-Norton (1982) 1.4 2.3 .60 H St. George .40 Sale 1 (1982) 1.4 5.2 H Kodiak (1983) .2 5.4 .92 North Aleutian Shelf (1983) .7 2.7 .29 H Navarian Basin .33 Sale 1 (1984) 2.8 9.8 M,H Sale 2 (1989) 2.8 9.8 H Chuckchi Sea .30 Sale 1 (1985) 2.1 5.2 H Sale 2 (1994) 2.1 5.2 M,H Hope Basin (1985) .43 1.72 .35 M,H 1Alaska OCS-Office, BLM C-20 In addition to the petroleum development, .some other mining is assumed to take place. Development of the U.S. Borax mining operation at Quartz Hill in Southeast Alaska is assumed to occur in the high scenario. In addition, development of the Beluga coal resources is assumed in both the moderate and high scenarios. In both scenarios, coal is assumed to be produced for export. The special projects described above do not exhaust the mining employment in the state. Additional employment occurs in the explora- tion, development, ·and production of nonpetroleum minerals, as well as a major component of headquarters employment in Anchorage. Market forces and governmental policies are assumed to be such that this component of mining declines in the low case, remains constant in the moderate case, and grows in~the high case. Table C.4 describes the three separate mining scenarios used in this study. In the low scenario, mining rises in connection with devel- opment at Prudhoe, but falls after 1983. By 2000, mining employment is almost 275 less than in 1980. Growth occurs in both the moderate and high scenarios throughout the period. By 2000, mining employment is 9,500 greater than in 1980 in the high scenario and 2,900 greater in the moderate. Agriculture-Forestry-Fisheries This industry is, in reality, three distinct subindustries which represent Alaska's renewable resource industries. Of the three, the C-21 TABLE C.4. MINING EMPLOYMENT (thousands of employees) Low Moderate High Scenario Scenario Scenario :1.980 ~::; .. '\ ""1f::· r.:-1. 3'7' r.:-:1.63 .,. t .. -l .... J .... J {o .. } • :L!f81 ~:; + :1. 6~5 r.:-. ., ... ... 5 36-4 ,; • .. :a .L / • 1~->82 I • ~12~~ 7 -;·':i.3 ""? 784 • , . • :1. 9~33 ~\ () ~~\ (~) !""!-+ ····-~ 8 6'?!3 t> • Ci • .\ // • :I.9B-4 ..... i!):2:4 8 06:1 . B 410 I .. .. .,. 19B!5 1::· ,; • :i ::} l~. ~) • .4;~~5 6 .. O!:):i. :1. (_:;t~ C) ·-r.~· 097 1:!' ;=·''?£{. 6 ~j~i9 ,J .;. ·..J • .. 1 '7lfj~7 !5 • ()~7~5 \.1 • 3<S6 7 .. .. -,r.::o ., ...... .; ... 1 ,-, .. ..,. ..... .. ,. (j(".) ;3 • ()5·4 6 .. .:>55 7 • 9Cl•:; :1.989 ~=.i o) 05() I 6 .. , 4? 8 40:L + I • 19{1() 3 0:1. :L 6 , ........ -.. ·-:; 885 • • -~!.:>I • 1. ·:j 9 :f. . -:'t 9'?0 . l.i t::-.. ~.· C'l :1. () 6C1 :L ·> ·> .... J .. :i'-) • ;LSHJ::!. lt • {;i .:S 9 6 • 8-d.C.: ' ' :1. 1 • ,., ~( .. .:..• .. ) .. !. 1"?93 4 ; ~-~ /.{. s:· 7 • 2 \·:~'J. 1 1 .. 91 :1. 19';'4 ··4 • 9:::.~~3 ,., 0 • ()89 :l =~ + 7-4rl 19";;.'5 l+ ·~/()8 •'''I 1. 51. 13 .. 2·43 • (.") ·> 1996 4 • t;:ti~3 8 .. 2·4·;:t :1.3 • 808 :i. 9'7~7 4 • .··-. .. ;·-:::"' Oz .. ) I...; 8 • :i. ~s ~:5 1 ., .;> • 966 J. '7 s=· tl ·4 ::;.<"J.a ~, O(i? :i. 4 2~)9 • 0 • • :i.99(? 4 • 829 8 • ().4;2 14 + 650 :~~0()0 -4 ,. ......... 8 0~)-4 14 639 • (:).J.'.,..t • • -.. ... -.;-.._·_, . fishing industry is currently the largest in terms of both.employment and value of product •. Agriculture is currently only a marginal industry employing few people statewide (Scott, 1979). Current state efforts to develop agriculture may lead to its increased importance in the future. Forestry consists of only a small component; the future of forestry is most appropriately discussed with the future of lumber and wood products manufacturing. The future of agricultural development in·the state depends impor- tantly on governmental policies and actions. State and Federal land policies, infrastructure development and loan programs, and marketing programs will determine the future of this industry. In the low sce- nario, it is assumed that government policies do not favor agriculture. New land is not opened up; ·old agricultural areas suffer from competition With other land uses (recreation, residential) and from competition for markets from outside producers. In the low scenario, agriculture dis- appears in Alaska. The high scenario ass~m~s a major positive government effort in support of agriculture with a fifty-fold increase in land in agricultural production by 2000. In the intermediate case, agriculture is assumed to rise only slightly from its current levels of employment. This assumes, as in the low case, that agriculture receives low pri- orities from government. Fisheries also hold promise for the future. The major determinant of future increases in· fisheries employment will be the expansion of the Alaska bottomfish industry. The creation of the 200 mile limit may C-23 support increased Alaska bottomfish activity. In all cases, employment in the existing fisheries is assumed to remain at its current level. Increases in production are assumed to have no effect on employment because of limited entry and labor-saving improvements in the fleet. Employment increases occur in both the high and moderate cases as a result of the development of an Alaska bottomfish industry. In the high case, the Alaska fishing industry is assumed to replace all of the existing foreign fishing·effort inside the 200 mile limit; while the moderate case assumes only 50 percent replacement. No bottomfish industry is. assumed to be developed in the low case. Table C.S illustrates the three agriculture-forestry-fisheries sce- narios used in this. study. In the low case, employment decreases by about 170 over the period due to the reduction in agricultural employ;.. ment. The high case shows employment rising by almost 4,700 during the projection period. In the moderate case, employment rises by almost 1,000 between 1980 and 2000. Federal Government Federal government employment has always been an important component of Alaska's economy. In recent years, federal government employment has been growing very little; increases in civilian employment have been offset by decreases in military employment. Low rates of growth in federal government employment are assumed to occur in all three scenarios. In all scenarios, federal military employment is assumed to remain constant at existing levels. In the low and moderate cases, federal C-24 TABLE C.S. AGRICULTURE-FORESTRY-FISHERIES EMPLOYMENT (thousands of employee~) Low Moderate High Scenario Scenario Scenario .. 1'n10 1 + ;~03 1. + ~)0:3 j_ v304 1.981. l .204 1 • :3:1.-4 1 + 32~5 ., r''"'"'·~ :i. ·' r"-J :i. .::n2 :i. ........ OJ'"' J. "To..: • .L r .•. o} ..:}~-<t 1.983 l • :L83 1 (·· ~3 ~::. :L . ----·---1. v -4()3 :1.984 1. • J. 8:3 :i. • ~} <;= ~5 1. .,480 198!"5 1 • :1.55 1. A_.., .. \ + ·'t Jt,·:. i' :l + !5-41 1.986 :L • :1.54 j_ • o\t~)9 :J. (p ~)8\) 1. <?G/' J l ::2_.-=~;. :1. + -4~38 :l ll'"\1"\ • t-C>L-7 1.9f38 ·r 1 (J!:j 1. + ~:5 ~:5 J. 1 / .. 32 -~ • • 19G9 -:l .087 :L • 612 ·! +863 .. 1. 9'70 :l (1..., ,. • ,}/ 0 1 ~ 6·:;'~:; 2.069 1991 :1 .• 06~! :1. _,,_., ""'r +I ,':J/ 2.309 1 (;)•;:) •") , J -~ :1. +031 1 + 7i.Y·7 ;2-l-620 J q(">"Z ~ .1' 'l w :1. ·:-()32 :l. (""t ·" -'"'\ +C) C)~·-3.029 I 19S.'4 :L o··:t ... 't 1 s~~2~l -.r 450" ' • ...:> .. .:.: • ..... ' 1995 :l + 03~) 1 .. 7'86 3 + 74-4 1996 :1. • o~}::s :~+04~1 4.054 1997 :l + (;~)3 2. llO 4.462 1.998 ·I .033 ~, :L 7:1. 4.896 .~ ~v :l 9(?·~~) :L (pC•33 .·: .23:3 5.396 .... ~ 2000 :1. • 0~5:~ ~~-):~95 5+996 C-25 civilian employment is assumed to continue to grow at its historical rate of about .05 percent per year. In the high case, this rate of growth is assumed to double to one percent per year. Table C.6 illus- trates the three alternative federal government employment scenarios. Manufacturing The manufacturing industry in Alaska has four important components: seafood processing, lumber-wood products-pulp, petrochemicals, and manu- facturing for the local economy. (Assumptions are discussed in terms of industry product since this is their form of input in the MAP model.) Production of seafood processing is expected to continue to dominate the food processing industry in Alaska. Growth of this industry is dependent on the growth of the fisheries catch by Alaskans, so these scenarios reflect the fisheries scenarios. In all scenarios, the output of the food processing industry is affected by growth in the catch in existing Alaska fisheries and growth in the bottomfishery. In the high case, output in the food processing industry is assumed to expand by 100 percent between 1980 and 2000 due to increases in the catch of the existing fishery and by an additional 57 percent because of the develop- ment of a bottomfish industry. In the moderate case, output expands ~y 149 percent in existing fisheries and an additional 49 percent because of the bottomfish development. In the low case, no bottomfish industry is assumed to develop, so output expands only because of increased catch in existing fisheries, and a 22 percent increase is assumed. C-26 J TABLE C. 6 • FEDERAL GOVERNMENT EMPLOYMENT (thousands of employees) Low Moderate High Scenario Scenario Scenario 1 ~)8() 4~3 ·':) :'i ·"" A,.,_. . -· .-, r.,... • ~ ....... ~ ~)· .. :. ·> ... ~·: \}f.j i~ "=· + .-:.'J\J 19~31 "'':." "'t,;) • ·<10 () A1.:1 • ... ,., .... .-;~. •,,/\,/ 43 • ·40(> :1. '7"c1;.~ 4~i;. 600 4:::-> . , 600 4::> • l>0() 1 9FJ:3 ·43 • 7(){' " .... .._1 --:13 • ?0() 43~ 80() l'l'E::..:} -4:3 • ~~(){) A1 ::~: • ~:i 0() 4~) • 'iOO 198!5 43 + ·t~-)!J 4~5 ·> '-;: t.,} ~) 44 ·> lOG ·f 9B6 ·4.-;~ ()()() .\{ .• -:t (' " .. , 44 -.... .:•. '"'\ J. • • ,li\.}t.} • ~\}t,,: 1 ~y'B:;/ 4-4 t ()() .• .... :l. ........ lf4 ~)()() • ~t .!.~ ·> t..)' .. J • 1.S>ti Ej ,.~~4 ... , .... '" -44 ~:~o ()· 44 + :70() .,:.t,}'J ·> :L 9f~ s:· -4~+ .300 4A·+~500 44 ., 900 :1.9':>'0 44 400 44 1: •••• , •• ·4~5 1.00 <· ·> ··f \) ... , • 1.99:1. 44 • ::'iOO ·4·4 • 5()0 ..ct5 + 300 :1.9?2 ··:f.~~} • <SO() ··J-<{. • ~~· ~) () 4!:) • 500 :l9<7'3 -4i~ (o -1' ()(). 44. '7()0 45 • BOO :1.994 44 .. 80() 4·4 • f~ () () 46 + ooc, l99~'i 44 900 4-4 9()() _, ·' 200 • • -4t0 • :i.996 45 + 000 -4!:5? 000 46 .400 :l ~1 r:;· /' 4!5 • J. ()~ ·' t·--'·{ .::..• • :L 0() ·4C) • 60•) j ,.,,..,, .. , • -;~ 7 i...1 -4~j • 20() .-:t~.) ·:r-"2(;() 46 • 800 :i. 99S·) .l~5 ~ 3()() -45+ 300 47'. 00() 2000 4!5 • 400 -45 • -40() ·4i"' + 300 C-27 The growth of the lumber-wood-paper-pulp sector of man~facturing in the state is determined primarily by two factors. These are the Forest Service allowable annual cut and the Japanese market conditions. In the high case, these industries' growth reflect almost a doubling (over its 1970 level) of the annual cut by 2000. In the low and moderate cases~ growth in annual cut is only one-half this amount. The petrochemical industry in Alaska currently consists of the developments in Kenai. In the low case, there is no expansion of this industry. In the moderate case, the petrochemical industry .expands with the construction of the Pacific LNG facility as currently planned, the development of LNG facilities associated in the OCS activity in Western Alaska~ and the development of a fuels refinery as the ALPETCO project. The high case contains two additions to these projects. A petrochemical complex is assumed to be established in Fairbanks, using the state's royalty gas, to produce ethylene or fuel-grade methanol. The Alpetco project is assumed to be developed as a major petrochemical facility as originally proposed. The final component of the manufacturing industry consists of those industries producing for local consumption and other diverse specialized production. It was assumed that this sector would grow in all scenarios because of increased market size, allowing scale economies which make local production viable. This sector was assumed to grow at 1 percent per year in the low case, 2 percent in the moderate, and 5 percent in the high. C-28 Table C.7 shows the three alternative manufacturing scenarios. Manufacturing employment increases continually through all scenarios. It increases by SO percent over the period in the low case, .83 percent in the moderate, and 137 percent in the high case. Transportation I The exogenous portion of the transportation industry is that which serves special projects. In all scenarios, this industry includes the .operations employment for TAPS and the Northwest gasline. The other major source of transportation employment is.the· OCS petroleum development. This employment is associated with both SUPPfY ships and helicopters used in the OCS development. The difference in transportation employment reflects the difference in the OCS lease sale areas assumed to be developed. Table C.8 iliustrates the three transportation employment scenarios. Construction The final exogenous industry for which scenarios are required is that portion of the construction industry where the level is determined outside the economy. This sector includes construction employment associated with the special projects described above. This sector does not include capital improvement projects of any level of government or construction activity which supports the local economy;. the remainder of construction activity is determined endogenously in the MAP model. In all scenarios, the major development of special projects occurs in the early part of the projection period. The most important project during this period is the construction of the Northwest gasline which is assumed C-29 :1.'700 1'7B1 1982 :J.<Yt53 1. 98-4 1 9:3!.5 1. ~"'8<.1 :i.9B7 1 ';> 8f~ 1913S) 1.99:l 1·9S)2 :J.l.?93 1.994 1995 17"?6 19?7 1998 :woo TABLE C. 7 • MANUFACTURING EMPLOYMENT (thousands of employees) Low Scenario :1.2. ~):i. f:l 12 .,. s-~6f:3 l :3 i :3-.:f.-4 :L :7) f .s ~) .·ff 14~()0() 14+:~0~~ 1. . .;:} , ... ~:; ';> '? ,1 ;::• ,••, .r '"' .l ,J "' t.j .L ~::. :J. !5 ~-2 .(~. :L 1!5 , .. 48? l !5 + /' .\l!) :1.6. 0:1.2) •f ·' ·-)(;\r''\ .i.O -o-..-..\.)7 1.6.570 16. a5a - l/ + j_~)-4 1 j! (o -4!:5.:S :l?.,. 771 :1. B ~ 09/' :1.13. 4:34 C-30 Moderate Scenario :t~o):'?-41. _, 1~:-··:rr.:·~ 1 •... J ·:0 -.:> -...i ... .r 1 ~:; T ·::;. :~2 4 l (:) v ~~8~5 j, '7 + !~j-4:f. ·t ,., -~ -:-· r:~ .L C) • .1. ,J •.• J l f~-o) ~)~~5 '"1E)-=-::;'-4-4 :i.'~i. :35S\ 2:1. ~ 3!:)-4 21 + 8/'() _.,,.., .. ,..'lC" .,:....;:,y ~ ... :.0 High Scenario l~J + ~:)48 14.2/'!..:t :!.5.1.76 l ~:; ~ 9C)~) •f ·' ·-t I' ... , .L<:>+/07 :1.? ., 54J. :i.El.317 ~~1 + -46i7 :~2to :L4l ;.~3. 610 :23.341 24~350 2~5.461 ::!<S •. :582 2:/' t 555 28.,.601 29.370 ~}().139 31.180 • -:i 980 1981. 1982 1'7'83 :t 9<':)·4 .. ,.,, ... , ... ·.L ;' \:>\:> :i.988 :L 99<) :L993 1994 ·I •'""•"":I!':" J. '17 .... 1 :i. ~~'?6 199·7 1'1'98 :i. c;~i\)-' :2000 TABLE C.8. EXOG~NOUS TRANSPORTATION ENPLOYMENT (thousands of employees) Low Scenario ..__ --· ----·. .t , ... ~ .... .1. • ;".)~.v :L • ;:)00 :!..500 1.500 1.500 1 + ?0\) :t f. ·?c~o :L.?OO :t-.;. ::.:-' () 0- :l + 70(j :I.+ 7()0 l (, '70CJ 1.700 1.700 :L .. /'00 :1..700 1 + 7<J·() :! .• 700 1.700 :l <· 700 C-31 Moderate Scenario :L .,. ~50{) .t l::>r,t::• .1. + .. J~~o-: ... _1 ~~ •. ~5lti 2.,3:tOi 2 ~ ~549 2t6~~:L 2<-t.)6t3 '2-)'?0~5 .• ., •'''1''7 " •• -<-•0<0 ~-~ + t)():~ :~~~596 High Scenario :!. • 500 1. ·> !500 :i. t 52~5 1 + !:f82 :L .... 85~5 2~~4f~:l 2 ~ !:558 ~~ y 92~) ,..,_ , ·" ,.,. ,. .. •• <-t:ic .. ~·. ,. ... -_,"" 4 .-:.~ + 1·"1-.. :..~ 2 t 9'72 3.291 3.:397 ~5 + 722 ~s~;;.-76 3.901 3 .; 986 3.913 3.751 to begin in 1981. This is the only special activity assumed in the low case. The high and moderate cases reflect completion of other projects. Both cases assume Pacific LNG and Alpetco projects will begin in 1982, although a more massive-scale Alpetco development is assumed in the high case. Construction employment is also required in the development of the OCS fields, NPRA, and Beluga. Additional sources of construction employment in the high case are the construction of a new capital at Willow and a petrochemical complex in Fairbanks. Table C.9 illustrates the three exogenous construction scenarios. In all cases, employment peaks in the early 1980s. This peak is pri- marily a result of the construction of the Northwest gasline·which is a major one-time project. The bunching of other large projects, as well as the beginning of OCS development at this time, also leads to this early peaking. "THE STATE GOVERNMENT SCENARIOS Past studies of the Alaska economy have indicated the key role state government plays in the Alaska economy. State fiscal policy has been a major determinant of state economic growth. State expenditures determine not only direct government employment, but also through expen- ditures on goods and services and capital improvements, they will affect all endogenous sectors of the economy. The state government scenarios described in this section attempt to define the most likely range of state government activity. C-32 TABLE C.9. EXOGENOUS CONSTRUCTION EMPLOYMENT (thousands of employees) Low Moderate High Scenario Scenario Scenario --.. I I 1~18() • O'iOj .<>s:·o~-.090 .J .f (''of"":·! J. 7 (;) .1. () ~ ~)S:'C' o. 73.":4 _., .... n.-, \) ~ /07 1982 2.,8a5 ·4 + 3:3~5 3t 960 198:3 I' 8,., .. ., + ,:;. . .:) ,., ·•n····, 7*C>07 U. "I(~ C) .; ~t..\.J, :t 98-4 '7 ~ ()38 9 r.--···-· + .. :)\j..;} l4v488 . J. 98~:.:; 1 r·-, ·-+ ~(j..:} 2 ~ 35!.=5 8. 963 :L 9B.:; o .. ooo :L ~ (j ~:.~ '? (~ ~ :;:?2 1 ;;i<:5:? o.ooo 1 :i. ~:)·:;;-•;.· ·"' t::"' , .. , .. ..:>.,; "t,JO :l9E~3 0 ·> 000 1 •I ''1'1-,. ~-. M''Y .. ,. • .I.~C) .:..; 'l..:i..:> 198S-~ o.ooo 1. • 08~) 2+7..;fS:t 1. 99(j o.ooo 0 + ·4("32 3 ., 88-4· 199:!. ()v_0()0 :i. + (j()g> 6o293 199;.~=: () y ()()() ,..) r::· ,, C) ' 7()1 ~--~ ... J.L.~ Ot •! !'~ ..... ··~·· • 1. 'l ·;-· ... ::a 0 ~ ()0() :'"\ •··.• ,, __ , ... ...:. y / ~} .1 . , Jj--.,1~ b+.:-17 199-•l· 0;. ()()() 1. ,:.0\::i-4 ·4 + 158 :l9i_;t5 Ot•JC'O o. ~5() ~:-· 3. 193 1.99r.~ 0 ->-()0() 0 + :2!)2 3 + ·<+36 1. ~19/' o.ooo o .. -r r::·.-, ..,. 101 .L..J . ..::. ,j + 19~tt~; i.).? 00() Ov3()~,j .-., .•, ,, .t::. t 00~) 1999 o.ooo () ·> 40() o. 714 ...-_ ............. ~\,'i .. }v o .. 00() C•,t54~:; 0 ~ 7~18 C-33 Two factors affect our ability to project the future course of state expenditures. First, since the beginning of production at Prudhoe Bay, state revenues have overtaken expenditures; revenues from this pro- duction will continue to increase in the projection period. Secondly, the establishment of the Permanent Fund and recent tax reduction and ' wealth-sharing prog;.am? place constraints on the use of certain petro- leum revenues. These recent changes in the structure of state spending constraints limit the usefulness of past fiscal policy for determining projected future policy. For this study, we will assume thr'ee separate directions for state fiscal policy, each of which will be defined by the growth of real per capita expenditures. Real per capita expenditures measure the effect of increases in prices and population on state expenditures. Between 1970 and 1972, real per capita expenditures grew at almost 24 percent per year; this was primarily a response to the lease sale bonus of $900 million from Prudhoe Bay in 1969. After 1972, the rate of growth dropped to .5 percent per year. We will describe the growth of real per capita state expenditures in terms of its relation to real per capita incomes. The relationship between income and state expenditures will be described in terms of the income elasticity of state government expenditures; this elasticity equals the assumed proportionate increase in real per capita expenditures which would result from a one percent increase in real per capita income. The historical pattern of state expenditure growth shows real per capita C-34 expenditures as an increasing proportion of ~eal per capita income through most of the period., The proportion increased through 1971 with a rapid expansion between 1969 and 1971 as a result of the Prudhoe lease sale bonus. Between 1971 and 1977, the ratio of real per capita expen- ditures to real per capita income remained constant (Goldsmith, 1977) • . The state's present revenue situation makes it hard to forecast how this ratio will change in the future. Our three scenarios assume that real per capita expenditures consume a growing, constant, and declining portion of real per capita income. The low case assumes that the level of real per capita state expenditures stays constant through the projection period and real per capita state expenditures decline over the projection period as a proportion of real per capita income. The moderate case assumes the real per capita state expenditures proportion of personal income stays constant with real per capita state expenditures increasing at the rate of real per capita income~ Finally, in the high scenario, real per capita expenditures increase at one and one-half the rate of real per capita income and increase as a portion of real per capita income. In combination, these three state expenditure scenarios and the three economic scenarios produce nine growth scenarios for the period between. 1980 and 2000. C-35 POST-2000 For the period between 2000 and 2010, a judgmental approach to projecting the level of economic activity was used. The approach used for the post-2000 period was to assume a rate of growth which described the possible continuation of the high, moderate, and low scenarios. In each case, a similar rate of growth was assumed for all three scenarios for the major variables--employment, population, and households. This implicitly assumes changes in household formation and labor force par- ticipation assumed between 1980 and 2000 do not continue after 2000. The assumed growth rates describe three possible post-2000 growth paths which are based on examination of growth in other similar areas as well as the historical growth of the Alaska economy. The high case assumes a continued expansion of the Alaska economy as a result of increasing resource development, although a reduced role of state government. The major economic variables are assumed to grow at 3.3 percent per year, which is approximately the rate in the high economic-moderate government scenario in the last part of the period. The moderate scenario assumes slightly slower growth at 2 percent per year, which is slightly less than in the moderate economic-low government scenario. This growth is assumed to result from more moderate resource development and reduced government activity. The low scenario provides only minimal growth at one percent per year, which reflects a self-generated growth 'from govern- ment expenditures. C-36 Projections of State Growth 1980-2000 This section presents the statewide projections of future economic and demographic activity. These projections are the basis for the energy end-use projections. The projections presented in this section are projections of the MAP model and the economic and government scenarios described above. The combination of three economic and three government scenarios produced the nine alternative projections presented here. Table C.lO describes the projected growth of total employment in each scenario. As would_be expected, the combination of high economic and high government scenarios (HH) produces the greatest growth~ and the low economic-low government scenario (LL) produces the lowest. In scenario HH, total employment .grows by over 300,000 between 1980 and 2000, an average annual rate of growth of 4.5 percent per year. Total employment grows by only 78~000 in the ~cenario LL, which is an average annual rate of 1.6 percent per year. In all the scenarios, the bunching of major construction projects in the early 1980s results in the most rapid growth occurring in this period. The effects of the alternative economic scenarios can be seen by comparing three economic scenarios with the same government expenditure assumption. We will examine those scenarios with a moderate level of government expenditure. Total ~mployment grows at an annual average rate of 2.3 percent per year in the low growth case, a growth in employ- ment of 122,150. In the moderate case, total employment grows by 161~420 C-37 1 =:_;.:·=.;.:() :1. =:? ·:) :.~~; ·' , •••••.••• ,!'". l. '7 Ci iv' ·: ::::==::)!:-: ·'· J ,_,,...; TABLE C.lO. TOTAL EMPLOYMENT, 1980-2000 (thousands of employees) i_EE T CJL. SCENARIO NAMES: :.~:30 ~ '7' 16 :~::)B + 1. :?<) 210 .. 0 '·?';.~ 304.(·:1."7 445 + 032:: / • -. I , ·-.• ,• r •. •••t .~.-:: ·'·~ ,:) + (':) ··;' / -2 '/ :i-~-\).I ~5 ::~9A \\ 96::) LES.GL -Low Economic/Low Government LES.GM -Low Economic/Moderate Government LES.GH -Low Economic/High Government MES.GL -Moderate Economic/Low Government MES.GM -Moderate Economic/Moderate Government MES.GH -Moderate Economic/High Government HES.GL -High Economic/Low Government HES.GM -High Economic/Hoderate Government HES.GH -High Economic/High Government 26<_;: ... ~):;B ::29~::; ~ j·'f3~5 3-49 ·) B;~~5 t·iES. GL -,.,-1-'~ . -0· -;_;--9.-, ··-: ·_V: ~-~'- 2-45 ;098 . ::~5B. l 02 292.023 319+9'72 'I •---, ,..,,~ t··; 1:. :J + L"JI ;I 2:to~o~~r; 29:1..414 3 30 + 47"7 Note: Values in 1980 adjusted to be the same in all cases; adjusts for minor differences in exogenous series. C-38 between 1980 and 2000, which is 32 percent greater than in the low case and an average annual growth of 2.9 percent per year. Total employment in the high scenario grows by 244,550, which is an annual average rate of 3.9 percent per year. The effects of the alternative government expenditure scenarios on economic growth can be examined by comparing three alternative projec- tions 'tvith the same economic scenario. Examining the projections with the moderate economic scenario and low, moderate, and high exp~nditure scenarios shows that the effect of ·varying state expenditure scenarios is similar to altering the economic scenarios. Under the moderate economic growth scenario, total employment increases at an annual average rate of 2 .1. percent per year·-Total employment increases at. an annual average rate of 2.9 percent per year in the moderate expenditure case. This is 38 percent faster than in the low case; when the government ex- penditure assumptions are held constant at the moderate l~vel, the growth rate in the moderate economic scenario is 2.6 percent greater than in the low. The average annual rate of growth in the high government scenario is 3.4 percent per year, which is 17 percent faster than in the moderate scenario. This compares with the 34 percent difference in growth rates between the moderate and high economic scenarios. Examining the effects of altering the government expenditure scenarios shows that in all cases state government expenditure is expected to play an important role in projected future growth. 1 State government employ- ment assumes a different role under each scenario, which reflects the C-39 alternate assumption about state government expenditures as a proportion of personal income. State government employment as a proportion of total employment falls in the low scenario, increases slightly in the moderate scenario, and increases in the high scenario. In 1980, state government employment is 21 percent of the total. By 2000, this proportion has fallen to 19 percent in the low scenario, risen to 23 percent in the moderatescenario, and risen to 26 percent in the high scenario. The importance of state government spending to the projections of total state activity makes it necessary to examine the consistency of these projections. It is necessary to ask whether the state can make .1 ~1 .] ·~ :,J '1 . ;.!J i ·~ -~ .J ~~ 41 -~ -~ this level of expenditures without running out of money or requiring {~ .~ large increa.ses in taxes. One consi~tency check is to examine the state' sJ fund balance in 2000. The fund balance is where the state accumulates · ::.1 excess revenues; it includes both the Permanent and General Funds. The most important source of revenue for the state during the projection period will be petroleum revenues. The revenue projections used in this study are based on the most recent projections of the Alaska Department of Revenue (Alaska Department of Revenue, Petroleum Production Revenue Forecast: Quarterly Report, March 1980). Based on this assumed gro~~h in revenues, the fund balance is positive and large in all scenarios in 2000. Only.in the high economic-high government expenditure scenario has the fund blanace peaked. This scenario has the lowest level of fund balance in 2000, $48.9 billion (in current dollars). Given the petroleum revenue assumptions, all three of the assumed government expenditure scenarios are possible. C-40 Table C.ll describes the growth of the population in each scenario. In the high economic-high government scenario, population more than doubles over the period, growing by 487,000. In the low economic-low government scenario, population is projected to increase by only 36 per- cent~ In all scenarios, population growth follows the pattern of employ- ment growth. Examining the moderate government expenditure scenarios illustrates the effect of the different economic scenarios on population growth. In the low economic scenario (LM), population grows at an average annual rate of 2.1 percent per year, reaching 635,578 by 2000. In the moderate economic scenario, population grows slightly faster (a rate of 2.6 per- cent per year); by 2000, population in this case is 10 percent greater than in the low case. Population in the high case reflects the rapid economic growth assumed in this case. · Population grows by 97 percent; by 2000, population is 19 percent greater than in the moderate case~ The effects of the alternate government expenditure scenarios pro- vide as great a variance as the economic scenarios. By 2000, population in the moderate economic-moderate government scenario is 12 percent · greater than in the moderate economic-low government scenario. The moderate economic-high government scenario projects population in 2000 which is 8 percent greater than in the moderate government scenario. Population growth rates between 1980 and 2000 vary from an annual average of 2.0 percent per year in scenario ML to 2.6 percent per year in MM and 2.9 percent per year in scenario MH. C-41 J. ·~=-==Jo 1. 99~5 TABLE C.ll. POPULATION, 1980-2000 LE13. C!L 4 6 :-",' i-J. ~:; ... -~. .!-~·:;.~ov ~3t6 • ,., ·t --· -· -·· , . . "+~ ..• /~/ 5 () ::> y 9 .-:·i-~:-~ ~) 4 :.:.• .; •:_; Cf' \;) Hi:::;;;. C7H 55 0 + :5 ~}-4 64~:j ·i :.~~5 :i. (thousands of peoplk) ··----··---:] .. A2:1.. 737 ! - 4\::i J. v 34::s ~:5 :L i + c:.::)~5 HES.GH -t21. 73"71 !.=5 :i. ~:i ~ ~~ :·; '7 ~=j6 {r ·:· ::::6 ? ~".) ~~; E: -) ~~~ ~:.i ? 7 :~=; :J -:--4 -'1· ::) LE3 .• GH HES.GL 4;_:~1 . /3/·' ~:5 J. ::~ ·> ?' ::2 0 <S 6 0 .,. () _,·:;. 3 7~.3 -:L +!54~~ 421.737 ~l f~ ~i ~ -:~-6 3 51 s . -'t9 6 57~5 + 22 7 6.27 .1so HES. C1f-i 421. 73 i ----·-......,..------------ SCENARIO NAMES: LES.GL -Low Economic/Low Government LES.GM -Low Economic /Hodera te Government LES.GH -Low Economic/High Government MES .GL -Moderate Economic/Low Gov ernment MES.GM-Moderate Eco n omic/Moderate Government MES.GH -Moderate Economic/High. Government HES.GL -High Economic/Low Governmen t HES.GM -High Economic /Hod e rat e Government HES.GH -High Economi c /High Governme nt Note: Values in 1980 adjusted to be the same in all cases; adjusts for minor differences in exogenous series. C-42 In all scenarios, population grows at rates slightly lower than employment. This reflects, in part, the increased labor force participa- tion of both Alaska Natives and women and the changing age structure of the population. Total employment as a proportion of population is 49 percent in 1980. By 2000, this proportion is 56 percent in scenario HH, 54 percent in scenario MM, and 50 percent in scenario LL. The difference between scenarios results from the importance of migration in each scenario. Migration brings in fe>ver dependents per employee than in the existing population. Migration is more important as a source of population growth in the moderate and high scenarios. This is responsible for the greater increase in employment as a proportion of population in these scenarios. Table C.l2 shows the growth of households in each scenario. House- hold growth reflects two factors, the growth of the population and the changing structure of households reflected in an increased probability that certain sectors of the population will form households. All sce- narios follow the same pattern of increasing proportion of households in the population. The pattern of this change can be seen by examining the low economic-low government (LL), moderate economic-moderate government (MM), and high economic-high 'government (HH) scenarios. In scenario LL, the number of households reaches 210,790 by 2000; this is a 58 percent increase during the projection period. The number of households by 2000 is 24 percent greater in scenario MM than in LL; the number of house- holds has increased by 96 percent over the projection period in HM. In scenario HH, the number of households is 32 percent greater than in MM; C-43 ·' .... ,,..,r:·· .I. ~:.· () ,,; .-.... , '· .•·, .-:.: \,! '.) ~ . ..J ·::· .. ···.·'j.j TABLE C.l2. HOUSEHOLDS, 1980-2000 (thousands of households) /._Lb. Cl... SCENARIO NAJ.'1ES: . _:i. 3 3 :·.; .. ,~--:; 1 :1. ~:_:; ~-:) ~ 3 ~:; '? :L <~i/> t 9:1. ;3 1.~3(, v ~:)'7B 2l·J" /90 -v -.,. -! L :) . .:>. 043 j 164 ,, ~?9.:i :L87 .300 :l)'~?v:.347 I I''',-., _.,, ~ .. ·• J ..• J::. ;;) • l J 1'1 l .;~) 8 ~· :=5 ::j 4 1'7'~~ ~ ?:~:L ::::3-4 + 464 LES.GL -Lmv Economic/Lmv Government LES .GM -Low Economic/Moderate Government LES .GH -Lmv Economic/High Government MES.GL -Moderate Economic/Low Government MES.GM -Moderate Economic/Moderate Government MES.GH -Moderate Economic/High Government HES.GL High Economic/Low Government HES.GM -High Economic/Moderate Government HES.GH -High Economic/High Government L 1·-,... . ... , I -::. \:) '\ lJ , .. ol 0"1 .-, •'', ''• , .. , .l / )' • v ~ ........ : ~:2 <> 8 v E~ ::S 0 ::2 ~=5 ;,:.: ~ 1 0 () ,,, ... , .. , ... ,, J"l t::. ~:> .,. \.:t 1... ·f .t. ·-;.r .. ~. C)".:: .L ~.J / 4 '-·' , ....J :f.9 ~=j + 23'7 2 :; :~:; y :_:) ·4 /} 2/~} ·) ~:)07 .hEE;. GL '133.043 15<:;.~;,.~8 176. ?:1.2 20 3 . 8 9 3 2~):l ~ 897 HES . Gt'l :L33 . 0 43 :L75.0 79 .. , ' .. ., ·-'"1 0 .-_l\,'t i ,J t 26~2. 42 1 Note: Values in 1980 adjusted to be the same in all cases; adjusts for minor differences in exogenous series. .the number of households increased by 158 percent over the projection period. In all three scenarios, over 80 percent of the expansion of house- holds results from the increase in the population. In scenario LL, 82 percent of the household expansion results from population growth, 85 percent in scenario ~ill, and 84 percent in scenario HR. These dif- ferences reflect the different household age structures which result from rapid growth. The average number of people per household drops from 3.2 in 1980 to 2.7 in LL, 2 • .7 in MM, and 2.6 in HR. This approxi- mate 20 percent drop in the average people per household is consistent with the projected decline in the national level of number of persons per household (Bureau of the Census, 1979). REGIONAL PROJECTIONS Anchorage Region This section describes the projection of employment, population, and households for the Anchorage ·region. These projections are for the period 1980 to 2010; growth beyond 2000 is assumed to follow the state patterns for each of the major variables. The Anchorage region includes the Anchorage, Matanuska~Susitna, Kenai, and Seward Census Divisions. Three state scenarios were chosen for the regional economic and end-use projections. These scenarios are the high economic-moderate government, moderate economic-mod erate government, a nd low economic-mod erate govern- ment scenarios ; these scenarios were chosen since they reflect the most likely range of future growth. Table C.l3 shows the grmvth in Anchorage .. C-45 TABLE C.l3. ANCHORAGE ECONOMIC GROWTH, 1980-2000 Low Scenario 1 Moderate Scenario 2 High Scenario3 EmEloyment PoEulation Households 4 EmEloyment Poeulation Households 4 EmEloyment Poeulation Households 4 19805 102,529 219,303 68,224 102,529 219,303 68,224 102,529 219,303 68,224 1985 111,118 248,850 85,177 119,352 260,034 85,805 132,186 275,848 89,515 1990 116,939 265,539 94,528 128,267 282,766 97,827 148,498 314,247 108,048 1995 134,425 293,381 108,377 151,735 322,582 116,718 185,601 375,483 136,364 ('") 2000 157,268 329,865 127,099 173,021 361,239 137,172 211,011 427,146 163,560 I .p. 0\ 2005 ·165,290 346,691 133,582 191,029 398,837 151,449 248,203 502,433 192,388 2010 173,722 364,376 140,396 210,912 440,348 167,212 291,950 590,989 226,278 1 Growth beyond 2000 at 1 percent per year. 2 Growth beyond 2000 at 2 percent per year. 3 Growth beyond 2000 at 3.3 percent per year. 4Households exclude 3,212 on-base housing not included in energy projections. 5 1980 has been adjusted to be consistent ·among scenarios. The Anchorage region is of central importance to the Alaska economy. Because it contains Anchorage--the state's administrative, distribution, and finance center--much of the growth in the -state will be reflected in this region. In the past, many of the events which have influenced state growth have occurred in the region. Projected future growth will continue to follow these patterns; however, the projected future contains relatively more activity occurring out of this region than in the past. The low scenario reflects limited growth in the state and Anchorage region. Anchorage is assumed to grow at an annual average rate of 1.8 percent per year over the projection period (2.2 percent per year between 1980 and 2000). This is approximately the rate of growth in the state economy and reflects the fact that the growth of basic sector activity which is assumed promotes the existing distribution of activity. Population growth follows the pattern of employment. Population grows slightly less rapidly than employment; population gro,vs at an annual average rate of 1.7 percent per year between 1980 and 2010 (2.1 percent per year between 1980 and 2000). Finally, household growth is determined by the growth in population and the changing pattern of household com- position assumed at the state level. The number of households in the Anchorage region is projected to increase by 106 percent over the pro- jection period; as at the state level, over 83 percent of this growth results from population growth. The moderate scenario illustrates the effect of the increased basic sector activity outside of the Anchorage region; Anchorage growth, as C-47 measured by employment and population, is slightly slower than the state growth. Employment in this scenario grO\vS at an annual average rate of 2.43 percent (2.7 percent for the 1980-2000 period). Population grows at an annual average rate of 2.35 percent (2.5 percent for the 1980-2000 period). As at the state level, population grmvs less rapidly than employment as a result of increased labor force participation. The number of households in this scenario is 19 percent more than in the low ' scenario. Households increase by 145 percent between 1980 and 2010; 84 percent of this growth results from the increase in population. The growth of basic sector activity outside of Anchorage has a more profound effect on the growth of Anchorage relative to state growth. Employment in the Anchorage region grows at an annual average rate of 3.6 percent between 1980 and 2010 (3.7 percent between 1980 and 2000), which is .1 of a percent slower than state growth of 3.7 percent. Population in this scenario increases to 377,000 by 2010 and averages a 3.4 percent rate of growth over the projection period (3.4 percent between 1980 and 2000). As in the other scenarios, the change in the number of households is a result of changes in the population and in household size. Households increase by 221 percent in this scenario; 84 percent of this growth is a result of population growth. Fairbanks Region Table C.l4 presents the projections for the Fairbanks region for the low economic-moderate government, moderate economic-moderate govern- ment, and high economic-moderate government scenarios. The Fairbanks C-48 n I .s::- \0 TABLE C.14. FAIRBANKS ECONOMIC GROWTH, 1980-2000 . 1 Low Scenario Moderate Scenario 2 High Scenario 3 EmJ2lO:¥:ment PoEulation Households 4 Em2loyment PoEul a tion Households 4 EmJ2lO:¥:ment PoEulation 19805 29,641 59,268 17,114 29,641 59,268 17 '114 29,641 59,268 1985 36,508 70,276 21,152 38 '813 73,072 22,ll8 43,223 78,354 1990 37,270 74,187 23,530 40,485 78 '911 25,330 47,638 88,555 1995 41,729 81,966 27,433 46,840 89,398 30,414 57,492 104,871 2000 48,326 92,159 32,712 53,068 100,111 35,843 65,852 ll8,836 2005 50,791 96,861 34,381 58,591 110,531 39,574 77,459 139,782 2010 53,382 101,802 36,134 64,690 122,035 43,692 91,111 164,41 9 1 beyond 2000 1 Growth at percent per year. 2 beyond 2000 2 Grm..rth at percent per year. 3 Growth beyond 2000 at 3.3 percent per year. 4 Households exclude 3,062 on-base households not included in energy projections. Energy projections assume only 91 percent of households are served by electricity in 1980 (based on 1978 end-use inventory). This rate grows to 95 percent by 2010. 5 1980 has been adjusted to be consistent among scenarios. Hous eholds 4 17 '114 24,121 28 '711 36,287 43 '716 51,422 60,836 1 ' region contains the Fairbanks and Southeast Fairbanks Census Divisions. The projection period is between 1980 and 2010; employment, population, and households are assumed to grow at state rates after 2000. Fairbanks is a regional center for the Interior and Arctic regions of-Alaska. Its past growth has been connected with resource development in the region; most recently, Fairbanks has acted as a center for develop- ment of Prudhoe Bay and the trans-Alaska pipeline. Since it is a regional center, Fairbanks' future growth will be affected by growth of state government as well as resource development in the region. The growth in the Fairbanks region is only slightly faster than for the state; both major resource development and growth of state government affect the growth of the region. In the low scenario, employment grows at a rate of 1.9 percent per year between 1980 and 2010 (2.5 percent between 1980 and 2000). Population in this scenario reaches almost 102,000 by 2010; the growth is at an annual average rate of 1.8 percent per year between 1980 and 2010 (2.2 percent between 1980 and 2000). As at the state ievel, the increased labor force participation accounts for a slower rate of population growth. The number of households almost doubles, growing by 94 percent over the period. Eighty-eight percent of this growth results form population growth. In the moderate scenario, the Fairbanks region grows at approximately the same rate as the state; resource development is spread more evenly in this scenario, with fisheries and OCS development occurring out of the C-50 Fairbanks region. Employment grows at an average annual rate of 2.6 per- cent per year between 1980 and 2010 (3.0 percent between 1980 and 2000). Population in 2010 is 20 percent greater than in the low scenario; growth during the projection period is faster, averaging 2.4 percent per year (2.7 percent between 1980 and 2000). The number of households in the Fairbanks region increases by 132 percent in the moderate scenario; 88 percent of this change is a result of population growth. The high scenario has major developments--petrochemicals and agri- culture--occurring in the region. Because of this, growth (particularly in the 1980-2000 period) is faster than for the state. Employment in this scenario grows at an annual rate of 3.8 percent (4.1 percent for the 1980-2000 period). Population follows the typical pattern, growing slightly less rapidly than employment. The growth rate of population averages 3.5 percent per year between 1980 and 2010 (3.5 percent between 1980 and 2000). Households follow the same pattern; the number of house- holds more than doubles, with the majority of the growth resulting from population growth. Valdez Region The Valdez Region consists of the Valdez-Chitina-Whittier Census Division. This region has experienced major growth recently as a result of the construction of the trans-Alaska pipeline and tanker port in Valdez. Future growth of this region may result from expansion of industrial activity due to the location of the pipeline terminus. Table C.l5 shows the projected growth in Valdez. c-sl ("") I Vl N TABLE C.l5. VALDEZ ECONOMIC GROWTH, 1980-2000 Low Scenario 1 Moderate Scenario2 High Scenario 3 Em£loyment Po£ulation Households 4 Em£loyment Po£ulation Households 4 Em£loyment Po£ulation Households 19805 2,146 5,821 1,878 2,146 5,821 1,878 2,146 5,821 1,878 1985 2,967 6,739 2,255 3,782 8,063 2,698 7,464 9,660 3,182 1990 ·3,328 7,163 2,491 4,241 8,768 3,059 7,323 11,080 3,830 1995 3,532 7,914 2,853 4, 713 10,003 3,628 7,358 12,467 4,522 2000 4,033 8,898 3,354 5,237 11,201 4,197 7,717 13' 296 5,060 2005 4,239 9,352 3,525 5,782 12,367 4,634 9,077 15,640 5,952 2010 4,455 9,82~ 3,705 6,384 13,654 5,116 10,677 18,396 7,001 1 beyond 2000 1 Growth at percent per year. 2 beyond 2000 2 Growth at percent per year. 3 Growth beyond 2000 at 3.3 percent per year. 4 Energy projections assume only 71 percent of households are served by electricity in 1980 (based on 1978 end-use inventory). This rate is assumed to grow to 75 percent by 2010 . . 51980 has been adjusted to be consistent among scenarios. 6Because of the rapid growth assumed in the Valdez economy in this scenario (between 1980 and 1985, employment more than triples), we assume that not all of the new employees bring families but that they live in an enclave-type area and commute to a shift-work situation from other regions. We assume that in 1985, this amounts to close to 40 percent of total employment, but this drops to 20 percent by the end of the period. 4 Valdez is projected to grow rapidly in all scenarios. The rapid rate of growth results from the location of major projects in Valdez and the small population and employment base at the beginning of the period. In the lmv scenario, employment is projected to increase by 2,310 by 2010, at an annual rate of growth of 2.5 percent per year (3.2 percent per year between 1980 and 2000). Population increases at a slower rate of 1.7 percent per year (2.1 percent per year between 1980 and 2000). Households follow the pattern of population, increasing by 97 percent over the projection period. In the moderate scenario, the construction of a fuels refinery in Valdez results in a greater divergence from the state growth. Employ- ment increases at an average rate of 3.7 percent per year, tripling during the period. Population increases at a slower rate of 2.9 percent per year (2. 7 percent per year bet~veen 1980 and 2000) . Finally, house- holds follow the pattern of population and increase by 172 percent over the period. In the high scenario, a major petrochemical facility is developed in Valdez. This results in major growth in the region; employment almost triples between 1980 and 1985. It is assumed that, because of this major growth, not all employees bring families to Valdez but com- mute, on some shift basis, from other regions. We assume that 40 per- cent of the employees commute in 1985; this proportion is assumed to decrease to 20 percent by 2000 and remain at this level for the rest of the period. C-53 Employment in the high scenario increases by almost 300 percent between 1980 and 2010; this is an annual rate of 5.5 percent per year (6.6 percent per year between 1980 and 2000). Population, because of our assumption, increases much less rapidly, increasing at an annual average rate of 3.9 percent over the period (4.2 percent between 1980 and 2000). Households follow the pattern of population, increasing by 273 percent over the period; 85 percent of this is due to poulation. HOUSING STOCK PROJECTIONS The growth in population and households determines the growth of the housing stock in the three regions. Tables C.l6 through C.l8 illus- trate the projected growth in housing stock in each region. The growth in the housing stock parallels the growth in the number of households. Housing stock does not grow as rapidly as the number of households because each region begins the projection period with excess housing. In Anchorage and Fairbanks, minimal change in the housing distribu- tion is projected. In Anchorage, single-family units go from 52 percent to approximately 51 percent of the housing stock in all scenarios. In Fairbanks, the reduction in the proportion of single-family housing is somewhat greater, falling from 52 percent to 49 percent in each scenario. The other important distributional shift involves a shift in the type of multifamily housing from duplex to other multifamily units. c-54 Single Family Multifamily Mobile Home Duplex Total Single Family Multifamily ~tobile Home Duplex Total Single Family Multifamily Mobile Home Duplex Total TABLE C.16. ANCHORAGE HOUSI NG STOCK 1 1980-2010 Low Scenario 1980 1990 2000 37,422 50,130 65,506 19,061 25,409 36,430 9,239 11,725 16,032 5,871 6,226 8,958 71,593 93,490 126,927 Moderate Scenario 1980 1990 2000 37,422 53,309 71,837 19,061 27,530 40,772 9,239 12,561 17,890 5,871 6, 770 10,060 71,593 100 ,"170 140,559 High Scenario 1980 1990 2000 37,422 57,894 85,160 19,061 31,090 49,132 9,239 13' 982 21,331 5,871 7,101 11,996 71,593 . 110,667 167,619 2010 72,346 40,239 17,666 9,955 140,206 2010 87,555 49,689 21,760 12,337 171,341 2010 117,802 67,945 29,450 16,696 231,893 1Housing served, only off-base housing. The distribution in 2000 is assumed to remain constant after 2000. C-55 Single Family Multifamily Mobile Home Duplex Total Single Family Multifamily Mobile Home Duplex Total Single Family Multifamily Mobile Home Duplex Total TABLE C.l7. FAIRBANKS HOUSING STOCKl 1980-2010 Low Scenario 1980 1990 2000 9,009 11,462 15,446 4,792 6,550 10,005 2,252 2,964 4,178 1,272 1,278 1, 772 17,325 22,254 31,401 Moderate Scenario 1980 1990 2000 9,009 12,244 17,026 4,792 7,245 10,162 2,252 3,192 4,545 1,272 1,279 1,906 17,325 23,960 34,439 High Scenario 1980 1990 2000 9,009 13,529 20,436 4,792 8,519 13,586 2,252 3,617 5,543 1,272 1,505 2,410 17,325 27,170 41,975 2010 17,321 . 11,230 4,682 1,971 35,204 2010 21,048 13,592 5,624 2,343 42,607 2010 28,873 19,209 7,826 3,379 59,287 1Housing served, an increasing proportion of offbase households. The distribution in 2000 is assumed to remain constant after 2000. C-56 Single Family . Multifamily Mobile Home Duplex Total Single Family Multifamily Mobile Home Duplex Total Single Family Multifamily Mobile Home Duplex Total TABLE C.l8. VALDEZ HOUSING STOCK1 1980-2010 Low Scenario 1980 1990 2000 472 706 1,184 189 306 543 642 629 606 192 197 189 1,495 1,838 2,522 Moderate Scenario 1980 1990 2000 472 951 1,572 189 412 725 642 684 649 192 220 212 1,495 2,267 3,158 High Scenario 1980 1990 2000 472 1,231 1,902 189 533 892 642 792 763 192 259 250 1,495 2,815 3,807 2010 1,334 610 681 213 2,838 2010 1,950 902 808 263 3,923 2010 2,684 1,256 1,074 354 5,368 1Housing served, an increasing proportion of total households. The distribution of housing stock is assumed to change. in a straight- line manner from the 1978 distribution to that projected for 2000. Distribution in 2000 is assumed to remain constant beyond 2000. C-57 In Valdez, the change in housing stock distribution is somewhat more pronounced. It was assumed that housing preferences, which were projected to be much different than the beginning 1978 housing stock, would only slowly change the distribution as removal and growth of the population increased the demand for housing of different types. Because of this assumption, the proportion of housing which is single-family increases from 32 percent in 1980 to 47 percent in 2010. In all sce- narios, the proportion of housing stock which is mobile homes decreases; this reflects a stabilizing of the population over time. C-58 APPENDIX D. COMPONENTS OF THE END USE HODEL D.l. HOUSEHOLDS AND HOUSING STOCK The basic consuming unit for residential electricity consumption is the off-military-base household. Huch of the data available to analyze energy consumption, hmvever, is more closely associated with the number of housing units, which will generally be larger than the number of households for a number of reasons. Tables D.l and D.2 present information on housing units for the years 1960 and 1970 and indicate the different types of housing. Using the 1970 census definition as a guide, all housing units can be divided into year-round and seasonal units. The latter are not designed for year-round habitation. Of the year-round units, only a portion are occupied; and of those not occupied, only a portion are vacant in the sense of being for sale or rent. Second homes are not identified but are a component of both the seasonal category and the year-round category. For energy consumption purposes, there are three important housing stock measures: 1. Occupied housing units. Each household will occupy a housing unit, and this forms the basis for estimating the appliance electricity demand in the residential sector. t:l I N TABLE D.l. HOUSING UNIT ANALYSIS: 1960 Vacant Year-Round Year-Round Occupied Housing Units Population All Housing Housing Housing Not for Sale Per Occupied Census Division Units Units a Units or Rentb Housing Unit GREATER ANCHORAGE AREA Anchorage 23,972 23,564 21,853 788 3.4 Matanuska-Susitna 2,593 2,346 1,501 775 3.4 Kenai-Cook Inlet 2,504 2,339 1,686 551 3.4 Seward 1,494 1,294 966 146 3.0 GREATER FAIRBANKS AREA Fairbanks 12,598 11,928 11,056 550 3.3 GLENNALLEN-VALDEZ AREA Valdez-Chitina- Whittier 1,241 1,049 785 153 3.1 WESTERN REGION U.S. 3.2 aConstructed variable equal to occupied housing units plus vacant year-round housing units. bConstructed variable equal to year-round housing units minus occupied housing units minus vacant year-round housing units for sale or rent. This category thus includes 1) rented and sold awaiting occupancy, 2) held for occasional use, 3) held for other reasons, and 4) dilapidated. Notes: Second homes may be classified as either seasonal or year-round housing units. Southeast Fairbanks \<las not a separate census division in 1'960. SOURCES~ l960 Census of Rousing,.General Rousing Characteristics: Alaska. Tables 28, 29. Median Rooms 3.9 3.4 3.4 3.2 3.6 3.1 TABLE D.2. HOUSING UNIT ANALYSIS /: 1970 Vacant Year-Round Occupied Year-Round Occupied Housing Units Population Units All Housing Housing Housing Not for Sale Per Occupied Median Which Own Census Division Units Units Units or Rent Housing Unit Rooms Second Home GREATER ANCHORAGE AREA Anchorage 37,650 37,617 34,988 975 3.4 4.5 3,492 Matanuska-Susitna 4,214 3,355 1,826 1,073 3.4 4.0 177 Kenai-Cook Inlet 4,877 4,650 3,889 470 3.5 4.0 442 Seward 1,106 956 605 239 3.1 3.9 157 t:1 I GREATER FAIRBANKS AREA w Fairbanks/ S.E. Fairbanks 13,895 13,729 12,644 596 3.4 4.3 986 GLENNALLEN-VALDEZ AREA Valdez-Chitina- Whittier 1,447 1,405 947 411 3.2 2.9 257 WESTERN REGION U.S. 12,031,802 11,938,658 11,171,550 302,970 3.0 4.7 SOURCES: 1970 Census of Housing, Detailed Housing Characteristics: Alaska. Tables 60, 63, 66. 1970 Census of Housing, Detailed Housing Characteristics: United States Summary. Tables 1, 3. 2. Occupied plus vacant (but available for rent or purchase) housing units. Housing units which are occupied plus vacant units form the basis for space heating requirements because houses which are vacant, but available, must be heated in winter (although to a lower temperature) to prevent damage. 3. Second homes. Vacation homes and homes used seasonally will have different energy use characteristics than first homes. Before estimating the number of households and the housing stock for 1978, those individuals housed in group quarters must be identified and subtracted from total population since their consumption of elec- tricity is not reflected in the utility residential load . Table D.3 shows that in 1970 the population in group quarters was large in both Anchorage and Fairbanks. Although a large proportion of this population is military and the military population has declined since 1970, we assume the same number of individuals in group quarters in 1978 as in 1970. That is, the decline has affected military personnel not in group quarters. Table D.4 presents two estimates of railbelt households in 1978 and compares them with year-end electric utility residential customers. At least four factors contribute to the discrepancies between the household and utility customer figures: D-4 TABLE D.3. 1970 POPULATION LIVING IN GROUP QUARTERS GREATER ANCHORAGE AREA Anchorage Matanuska-Susitna Kenai-Cook Inlet Seward GREATER FAIRBAl~S AREA Fairbank s GLENNALLEN-VALDEZ AREA Valdez-Chitina- Whittier Population 124,542 6,552 14,250 2,021 50,262 3,116 Population in Population in Housing Units Group Quarters 118,809 5,733 6,208 344 13,719 531 1,870 151 42,682 7,580 3,023 93 Note: Group quarters are primarily institutions, boarding houses, military barracks, college dormitories, hospitals, religious centers, and ships. SOURCE: 1970 Census of Housing, Detailed Housing Characteristics: Alaska. Table 60. D-5 TABLE D.4. 1978 HOUSEHOLD ESTIMATES (thousand) PoEulation (1 2 000) Electric d Population Utility Ratio of a c Department b in Housing Population Residential Households of Department Units per Occupied 1978 and Rural to Commerce of Labor (1' 000) Housing Unit Households Customers Customers Greater Anchorage Area 215.7 226.3 209.0 219.6 61.6 64.7 77,000 .80 .84 Anchorage 179.0 185.5 173.3 179.8 3.4 51.0 52.9 57,916 .88 .91 Kenai-Cook Inlet 19.6 22.3 19.1 21.8 3.4 5.6 6.4 7,904 .71 .81 Matanuska-Susitna 14.2 15.4 13.9 15.1 3.4 4.1 4.4 10,152 .40 .43 Seward 2.9 3.1 2.7 2.9 3.0 .9 1.0 1,027 .90 1.00 Greater t:1 Fairbanks Area 59.4 60.8 51.8 53.2 3.3 15.7 16.1 17,524 .90 .92 I (j\ Fairbanks 54.1 55.5 Southeast Fairbanks 5.3 5.3 Glennallen-Valdez Area Valdez-Chitina- Whittier 5.9 5.0 5.8 4.9 3.1 1.9 1.6 1,539 1·.27 1.07 SOURCES: (a) State of Alaska Department of Commerce and Economic Development, Division of Economic Enterprise, "Numbers:· Basic Economic Statistics of Alaska Census Division," 1979. (b) State of Alaska, Department of Labor. (c) 1970 Census of Housing, Detailed Housing Characteristics: Alaska. (d) Federal Energy Regulatory Commission, Utility Power Systems Statements. 1. The household calculations and population estimates may be incorrect. 2. Some residential electricity hookups are for second and vacation homes. 3. Some residential electricity hookups are for units which are vacant since a minimal amount of electricity is necessary even in an unoccupied housing unit for such things as the heat distribution pump. 4. Some households may not have access to the electric utility. Table D.5 (compared to Table D.4) shows that in all areas of the railbelt the ratio of residential customers-to-households has apparently increased since 1970 unless the population and household size estimates for 1978 are very low. TABLE D.5. 1970 UTILITY HOOKUP RATES Greater Anchorage Area Greater Fairbanks Area Glennallen-Valdez A 1970 Occupied Housing Units 41,233 12,612 1,017 D-7 B 1970 Residential & Rural Utility Customers 41,151 10,756 561 B/A 1.00 .85 .55 Part of the differences among the census divisions in the ratios of households to utility customers arises from the fact that the utility· boundaries do not correspond to the census division boundaries used -in developing Table D.4. Specifically, Chugach Electric serves portions of Anchorage, Kenai -Cook Inlet, Seward, and Valdez-Chi tina-lfui ttier. Matanuska Electric serves a portion of Anchorage. A portion of Golden Valley Electric Association customers reside in the Yukon-Koyukuk Census Division. The only identifiable household concentrations not having access to utility service from the seven major railbelt utilities appear to be portions of Valdez-Chitina-lfuittier (Chistochina, Mentasta Lake, Tatitlek ), portions of Kenai-Cook Inlet (Tyonek, Seldovia), and portions of South- east Fairbanks (Tok). Vacancy rates for housing are not collected in a complete and accurate nianner. ·The rental housing vacancy rate for Anchorage in 197 8 was estimated at 14 percent.1 As of June 1979, the unsold inventory o f new houses was approximately 2 percent of the stock.2 It is not possib le from this information to develop an overall vacancy rate, but realtors generally agreed that the rate was higher in 1978 than in a normal market. A recent housing study by the Fairbanks North Star Borough estimated a vacancy rate of 14 percent overa11.3 Vacancy rates for other areas are unknown but assumed to be low·er than for Anchorage and Fairbanks.· D-8 The number of secon? homes within the railbelt served by electric u tilities is not known but is assumed to be concentrated in the Matanuska- Sus itna Census Division. In the 1970 census, 3,492 households in the . 4 Anc horage Census Division -indicated they owned second homes. Some are l o cated in the Matanuska-Susitna Census Division, and some of these would have appeared in the census as year-round housing units although there is no accurate information on actual numbers. We assume 2,000 second homes among the Greater Anchorage Area utility 5 c u stomers. Using this figure and the information from Table D.4 results in an estimate of the overall Greater Anchorage Area vacancy rate of be tween 14 and 18 percent, which would be about twice the normal rate. This may be somewhat high based upon the rate reported for Fairbanks. We have no vacancy information for Glennallen-Valdez. If we assume 25 percent of the population is not serviced by Copper Valley Electric Association, the vacancy rates derived using Table D.4 data range from zero to 20 percent, which probably brackets the actual value. The estimate of the number of households could be obtained either by estimating population and dividing by average household size or by using utility hookup numbers and adjusting for vacancies. Neither method is foolproof; but the latter involves only one estimate, while the former requires two. .We choose the latter and assume a 13 percent vacancy rate . The resulting household estimates are shown in Table D.6. D-9 TABLE D.6. HOUSING UNIT AND HOUSEHOLD ESTIMATES First Vacancy Rate Housing Units (percent) Households Greater Anchorage 75,000 13 65,250 Greater Fairpa?ks 17,524 13 15,245 Glennallen-Valdez 1,539 13 1,339 The housing stock may alternatively be calculated directly by a count or estimate, independent of the number of electric hookups. This method provides a check on the utility hookup method of housing stock estimation as well as providing information on the geographic distribu- tion of the stock (within the Greater Anchorage Area) and an estimate of the distribution of the stock by type. Table D.7 tabulates the housing stock analyses which have been done for the railbelt communities. From these analyses, it is relatively easy to construct an estimate of the 1978 housing stock for the Anchorage Census Division of 57,896, which is included in the table. Estimates for other Census Divisions must be developed more indirectly. Table D.8 shows the result of apply- ing the 1978-to-1970 population ratios to the 1970 year-round housing unit stock in each census division. These housing stock estimates can be adjusted to arrive at final estimates. D-10 TABLE D.7. HISTORICAL RAILBELT HOUSING STOCK DISTRIBUTION BY TYPE Single Hulti-Hobile Family Duplex Family Home GREATER _ANCHORAGE AREA Anchorage Census Division 1950: 3,325 964 1,128 202 1960 13,435 1,427 7,625 1,485 1970~ 15,572 3,813 13' 368 4,864 1978 (off base) 28,530 4,581 18,196 6,589 Anchorage Bowl 1975e (off base) 23,227 5,324 14,754 6,246 1975e (on base) 34 0 4,122 0 1975e (total) 23,261 5,324 18,876 6,246 1979f 26,300 24,203 6, 960 - Eagle River 1979f Gird~·.TOod 1978g Kenai-Cook Inlet Census Division 1960b 2,117 19 182 186 1970c 2,627 108 594 1,321 Seldovia 1970~ 102 29 22 1976] 153 20 41 Soldotna 1970~ 159 95 143 1976] 311 110 180 Homer 1970~ 310 39 18 1976] 251 31 134 Kenai 1970~ 574 371 231 1976] 684 350 274 D-11 Other Total 0 5,619 0 23 '972 0 37,617 57,896 0 49,551 0 4,156 0 53,707 0 57,463 0 3,524 198 0 2,504 0 4,650 0 153 15 229 0 397 0 601 0 367 16 432 0 1,176 0 1,308 Table D.7. (continued) Single Multi-Mobile Family Duplex Family Home Other Total Seward Census Division 1960b 1,307 86 77 24 o · 1,494 1970c 789 19 138 10 0 956 Seward 1976j 497 212 36 36 78 1 Matanuska-Susitna Census Division 1960b 2,336 20 149 88 0 2,59 3 1970~ 2,947 41 159 208 0 3,35 5 1978 310 7,616 GREATER FAIRBANKS AREA Fairbanks-Southeast Fairbanks Census Division 1950: (urban) 1,295 166 352 2 0 1,815 1960 6,527 671 4,547 853 0 12,598 1970c . 5,335 1,068 6,072 1,254 0 13,729 Fairbanks Census Division 1970c 4, 775 1,017 5,603 1,129 0 12 ,5 24 Southeast Fairbanks Census Division 1970c 560 51 469 125 0 1 ,205 Fairbanks North Star Borough (off base) 19701 9 ,884 1975m 11 ,324 1976n 15,200 1976° 1 6 ,89 4 1978° 8,787 1,232 5,616 2,306 0 1 7 ,94 1 1979° 17 '684 Fairbanks North Star Borough (off base) 1976p 49% 8% 25% 17 % 1% 1978q 52% 30% 12% 6% D-12 Table D.7. (continued) Single Multi-Mobile Family Duplex Family Home Other Total · GLENNALLEN-VA LDEZ AREA Valdez-Chitina-Whittier Census Division 1960b 803 31 392 15 0 1,241 1970c 881 34 278 212 0 1,405 Valdez 1970h 105 95 98 0 298 1978r 222 143 135 518 .12 1,030 1978s 314 171 521 16 1,022 Glennallen 1970 h 58 25 26 0 109 SOURCES: (a) U.S. Department of Commerce Census of Housing 1950: Alaska, General Characteristics, Table 14. These are all dwelling units. (b) U.S. Department of Commerce Census of Housing 1960: Alaska, Table 28. These are all housing units. (c) U.S. Department of Commerce Census of Housing 1970: Alaska, Table 62. These are all year-round housing units. (d) Estimated by author by nett i ng out 1978 housing units authorized for Anchorage Municipality from 1979 total and adding Eagle River and Girdwood (latter assumed all single-family units). (e) Anchorage Urban Observatory, University of Alaska, 1975 Housing Survey, Appendix I, p. 2. (f) Municipality of Anchorage, Planning Department. (g) Municipality of Anchorage, Planning Department. These are full-time residences only. Total residences weLe calculated at 729. (h) State of Alaska Department of Community and Regional Affairs, Division of Community Planning, Selected 1970 Census Data for Alaska Communities, 1974. (j) Kenai Peninsula Borough, Profile of 5 Kenai Peninsula Towns, 1977, Table 130. These are year-round dwelling units (vacant and occupied units designed for year-round living). This includes housing within the city limits of these tmms only and estimates 250 units outside Homer. D-13 Table D.7. (continued) (k) Matanuska-Susitna Borough Planning Department. (1) 1970 Census of Housing as reported in Fairbanks North Star Borough, FMATS Housing Study, draft, 1980. (m) E. Allen Robinson. "Situation Report: Fairbanks, Alaska," HUD Anchorage 1975, as reported in Fairbanks North Star Borough, F}~TS Housing Study, draft, 1980. (n) Jack Kruse, "Fairbanks Community Survey/' .Institute of Social and Economic Research, University of Alaska, Fairbanks, 1976, as reported in Fairbanks North Star Borough, ~TS Housing Study, draft, 1980. (o) William Rose, Fairbanks North Star Borough Planning Department, as , reported in Fairbanks North Star Borough, FMATS Housing Study, draft, 1980. (p) Jack Kruse, "Research Notes: Fairbanks Community Survey," Institute of Social and Economic Research, 1976, p. 2.1. (q) Jack Kruse, "Fairbanks Petrochemical Study," Institute of Social and Economic Res~arch, University of Alaska, Fair- banks, 1978, as reported in Fairbanks North Star Borough, FMATS Housing Study, draft, 1980. (r) Michael Baring-Gould et al. Valdez City Census, 1978. University of Alaska, Anchorage, Table 13. (s) Northrim Associates, Inc., CCC/HOK from the Environmental Impact Statement, Alaska Petrochemical Company Refining and Petrochemical Facility, Valdez, Appendix , Vol. II, p. 93 . This is total housing net of hotel-motel units and campers. D-14 TABLE D.8. 1978 YEAR-ROUND HOUSING STOCK ESTI}~TE BASED ON POPULATION RATIOS 1970 Year-Round 1978/1970a Ce nsus Division Housing Units Population Anchorage Kenai-Cook Inlet 4, 650 1.56 Matanuska-Susitna 3,555 2.35 Seward 956 1.53 Fairbanks/Southeast Fairbanks 13,729 1.25 Valdez-Chitina-Whittier 1,405 1.61 aBased on Alaska Department of Labor estimates for 1978. D-15 1978 Year-Round Housing Unit Estimate 7,277 8,356 1,463 17,207 2,268 In Seward, 40 percent of year-round housing units were unoccupied in 1970 and 35 percent in 1960.6 If we assume that 25 percent were unoccupi ed and did not require space heating in 1978, this results in an estimate of about 1,100 "first" year-round housing units (year-round housing units, as defined by the census, which are actually utilized on a year-round basis). Kenai-Cook Inlet had a 16 percent vacancy rate in 1970, dm.;rn from 33 percent in 1960. The growth in the number of units in the major com- munities of the census division bet~.;reen 1970 and 1978 was much slower 7 than would be indicated by the 1978 estimate of Table D.8. Thus, growth was more rapid outside these cities and may or may not have been accounted for by second homes. We assume 500 second homes and thus arrive at an estimate of the first home housing stock of 6,777. In the Matanuska-Susitna Borough, 44 percent of year-round housing units were unoccupied at the time of the 1970 census and 42 percent in 1960. If we assume that this rate has fallen since 1970 (Matanuska Electric Association had 678 seasonal rate customers in 1978, but this is not equivalent to second homes) to about 35 percent, it would be consistent with a 15 percent vacancy rate and 1,500 second homes. The Borough counted 7,616 dwelling units in 1978 but did not distinguish vacation homes. Netting out second homes would produce a first housing unit estimate of about 6,100 units. D-16 These census division estimates can be aggregated to arrive at an overall first h ousin g unit e stimate for the Greater Anchorage Area of 7 1,873, sho\vn in Table D.9. This is somewhat lower than the estimate d erived b y counting the number of electric utility accounts but is more reasonable as a basis for calculating ·electricity consumption on-an end _ u se basis. (For e x ample, there are some residences in Anchorage with t wo electric meters, each of which were counted as a customer during 1 978.)8 For the Greater Fairbanks area, the Fairbanks North Star Borough housing surveys closely correspond to the utility hook up data; hmvever, researchers admit that deficiencies exist in at least some of the surveys, \vhich could lead to an over count. 9 These house counts, however, would not include utility customer s located outside the Borough in the South- east Fairbanks a nd Yukon-Koyukuk Census Divisions. We assume that these effects, as well as the presence of some vacant, non-market housing in the North Star Borough and second homes outside the Borough, cancel one another out so that the utility hookup figure becomes our housing stock estimate for the Greater Fairbanks Area. In the Valdez-Chitina-Whittier Census Division, the vacancy rate was 33 percent in 1970, up from 25 percent in 1960. A large portion of this increase could be the decline in population of Whittier. Without additional information on the housing stock in the utility service area in the census division, we must use the electric utility residential hookup estimate of about 1,500. D-17 TABLE D.9. FIRST HOME HOUSI NG STOCK ESTIMATES FOR 1978 USED IN END USE CALCULATIONS Greater Anchorage Area Anchorage Kenai-Cook Inlet Matanuska-Susitna Seward Greater Fairbanks Area Fairbanks/Southeast Fairbanks Glennallen-Valdez Valdez-Chitina-Whittier 71,873 57,896 6, 777 6,100 1,100 17,500 1,500 The first housing unit housing stock is divided into four housing types which have very different space heating characteristics--single- family detached, duplex, multifamily, and mobile home. Information on the distribution of the housing stock by type comes primarily from the housing stock surveys shown in Table D.7. For Matanuska-Susitna, the planning department has estimated multi- family units and mobile homes. We allocate the remaining units between single-family and duplex units on the basis of the Anchorage proportions. In Kenai-Cook Inlet, the proportion of single-family units in the larger communities was representative of the census division as a whole. Thus, the 1976 proportion of 54 percent found in these larger communit i es D-18 is used for the 1978 estimate. For mobile homes, this was not the case as the proportion in the whole census division in 1970 was 28 percent, while it was only 20 percent in the larger cou@unities. In these larger communities, it grew to 24 percent by 1976, so we assume the same type of growth for the census division as a \vhole but that some of the rela- tive growth in the utilization of mobile homes is in areas inaccessible to the railbelt utilities. Thirty percent becomes our estimate. The most recent estimate of the distribution bet\veen duplex and multifamily units is the 1970 census. From the total data, a pattern toward single- family living is evident • so \ve assume that a majority of the growth since 1970 is in duplexes and that multifamily units are 600. For Seward, we assume the same distribution for the utility service area as indicated in the 1976 survey and that multifamily and duplex units are equal in number. For Fairbanks single-family units, we utilize information collected in the 1978 survey for the Borough and assume the same distributions for housing units outside the Borough. For trailers, \ve assume a downward trend in the percentage since 1976 and assume that the "other" category from the 1978 survey is not relevant for · our purposes. Thus, the 13 per- cent figure from the Borough count is taken. We further assume 21 percent of duplex and multifamily units are duplexes, which is an average of the various surveys. D-19 For Glennallen-Valdez, the data indicates a much higher proportion of mobile homes in Valdez than in Glennallen. lve use the 1978 Valdez City Census for Valdez and apply the 1970 Glennallen distribution to the remainder of the service area. The results of this analysis are shown in Table D.lO. D.2. RESIDENTIAL ELECTRIC SPACE HEATING Data on the proportion of housing units heating with electricity and average consumption levels for various housing types in different locations is fragmentary. None of, the electric utilities compile this information at present; and although some had special all-electric rates in the past, utility records of those customers have not been retained. The space heating distribution is currently relatively stable except in the outlying areas of the Greater Anchorage Area and in Fair- banks. In t4e fo~er, use of electricity fpr space heating. is growing relative to the primary alternative (fuel oil) because of the rising price of fuel oil and the relatively stable price of natural gas-genera electricity. In Fairbanks, there is a shift away from electric space heat toward fuel oil as the price of oil increases since incremental electricity isproduced by fuel oil. These shifts make it more difficult. to estimate the actual space heating mode split in these areas. Census data on fuels used for space heating presented in Table D.ll encompass the whole railbelt but are not current because .of the rapid D-20 ~ .. ·k· ... .t\1 ~ 'lri ~ ·l-·• i>·' .. .. ., • .j; :1!!~ . . o··· ¥r :w. 1\'l' !-,• ~ ~': t,._.,; G!!i :;, ~-- t;-· ,. 3···· . -./• t;.:, ~t· TABLE D.lO. 1978 FIRST HOME HOUSING STOCK DISTRIBUTION BY HOUSING TYPE Single Multi- Family Duplex Family GREATER ANCHORAGE AREA 37,357 . 5, 930 19,254 Anchorage 28,530 4,581 18,196 Kenai-Cook Inlet 3,660 484 600 Matanuska-Susitna 4,463 717 310 Seward 704 148 148 GREATER FAIRBANKS AREA Fairbanks/Southeast Fairbanks 9,100 1,285 4,840 GLENNALLEN-VALDEZ AREA Valdez-Chitina- Whittier 472 197 189 D-21 Mobile Home Total 9,332 71,873 6,589 57,896 7,033 6, 777 610 6,100 100 1,100 2,275 17,500 642 1,500 t:J I N N TABLE D.ll. PERCENT DISTRIBUTION OF SPACE HEATING FUELS IN THE RESIDENTIAL SECTOR Census Division Utility Gas Oil Electric Coal GREATER ANCHORAGE AREA Anchorage 1950 0 92 0 7 1960 0 82 0 15 1970 53 34 6 1 Matanuska-Susitna 1960 0 47 0 24 1970 0 62 1 14 Kenai-Cook Inlet 1960 0 69 0 8· 1970 31 52 4 3 Seward 1960 0 88 0 0 1970 0 92 4 0 SOURCES: 1950 Census of Housing, General Housing Characteristics: Alaska. 1960 Census of Housing, General Housing Characteristics : Alaska. 1970 Census of Housing, Detailed Housing Ch a racteristics: Alaska . Wood Propane 1 0 0 0 0 2 28 0 17 6 23 0 5 3 12 0 4 0 Table 17. Tables 29, 30. Table 63 . Other 0 3 4 1 0 0 2 0 0 Table D.ll. (continued) Census Division Utility Gas Oil Electric Coal Wood Propane Other GREATER FAIRBANKS AREA Fairbanks and Southeast Fairbanks 1950 0 30 0 54 16 0 0 1960 0 47 0 49 3 1 0 1970 3 61 7 20 2 1 6 GLENNALLEN-VALDEZ AREA t;:j I N Valdez-Chitina-Whittier (,.) 1960 0 57 0 0 17 0 26 1970 0 72 0 0 21 7 0 growth in the housing stock since 1970. Fuel oil is the predominant fuel except in those areas of Anchorage and the Kenai-Cook Inlet Cens~s Division where natural gas is now available. Coal was historically very important in Fairbanks, but it was surpassed by fuel oil in the 1960s. The census data does indicate a significant proportion of the occupied housing stock utilizing coal, wood, propane, and other fuels, even in 1970. Information on four electric utilities is available from a Federal Power Commission (now Federal Energy Regulatory Commission) report. This information is shown in Table D.l2. Utility personnel are unable to determine the source of this information but feel it is reasonable. This data shows that all electric customer growth in Anchorage in the early 1970s was more rapid than total customer growth. No such trend is apparent for Fairbanks. Additional published data on the residential space heating mode split in the railbelt is shown.in Table D.l3. This data tends to con- firm information gathered informally in conversations with utility and real estate personnel as well as analyses of utility monthly load curves. [Matanuska Electric Association (MEA) and Homer Electric Association (HEA) analyzed monthly.bills in an attempt to identify the number and average consumption levels for their electric space heating customers. The HEA analysis requires further work! The MEA analysis yielded an estimate of 2,685 space heating customers in 1978 and 18,172 kWh annual consumption for space heating per customer in 1979.] D-24 t::l I N Vt Customers Total All-Electric Non-All-Electric Chugach Electric 1963 1970 1971 1972 1973 1974 13,170 24., 682 25,761 28,687 29,077 31,779 120 1,280 1,475 1,756 2,010 2,605 Anchor a ge Municipal Light and Power 1963 1970 1971 1972 1973 1974 6,592 8,477 9,295 10,130 10,523 11,268 NA 381 700 700 928 1,123 Golden Valley Electric Association 1963 1970 1971 1972 1973 1974 NA 6,624 6,741 6,947 7,382 8,643 NA 802 850 850 1,448 900 Fairb a nks Municipal Utility System 1963 1970 1971 1972 1973 1974 4,120 4,532 4,443 4,540 4,443 NA 0 NA NA 19 19 NA 13,050 23,402 24,286 26,931 27,067 29,174 NA 8,096 8,595 9,430 9,595 10,145 NA 5,822 5,891 6,097 5,934 7,743 4,120 NA NA 4,521 4,424 NA Aver age Annual Consumption (kWh) Total All-Electric Non-All-Electric 6,137 8,057 9,194 9,386 9,887 9,621 4,681 6,431 6,782 7,080 7,855 7,982 NA 10,133 12,158 13,920 14,479 14,795 3,013 . 5,167 5,504 5,341 5,841 NA 37 .. 882 38,500 38,700 39,000 39,050 ·39,100 NA 18,045 18,127 18,127 17,985 17,355 NA 42,516 43,000 43,000 40,000 43,000 NA NA 46,316 46,316 NA 5,!345 6,392 7,402 7,455 7, 721 6,989 NA 5,884 5,858 6,260 6,875 6,944 NA 5,672 7,708 9,866 8,251 16,015 3,013 ·NA NA 5,169 5,667 NA He a ting Only For All-El e ctric Custome rs 29,600 29,100 29,600 29,750 29,900 NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA SOURCE: Compiled from Federal Power Commission, All Electric Homes, annu a l. TABLE D.13. RAILBELT RESIDENTIAL SPACE HEAT MODE SPLIT INFORMATION (percent) gas oil electric wood other-unknown Seward 1977 a 96 1 3 Kenai-Cook Inlet 1977 Seldovia a 88.2 3.9 3.9 4 Soldotna a City 70.4 25 2.8 .9 .9 Total 57.7 35.8 4.4 1.5 .7 Kenai 76.2 13.2 9.3 .7 .7 Total of Three Cities 66 25 7 1 1 Anchorage 1975 b Anchorage· Bowl house 68 17 11.5 .5 3 house new since 1970 77 .5. 6 16 0 0 trailer 48.5 36.5 7.5 1 6.5 trailer new since 1970 33.5 55.5 0 11 0 apartment 52 8.5 31 0 8.5 Fairbanks 1979c 0 70 11 12 7 SOURCES: (a) Kenai Peninsula Borough, Profile of 5 Kenai Peninsula Towns, 1977, Table 72; (b) Anchorage Urban Observatory, Anchorage Housing Survey, 1975, unpublished data. (c) Fairbanks North Star Borough, Community Information Center, Community Information Quarterly, Spring 1980, Vol. III., No. 1, p. 81. D-26 Additional information for drawing inferences about the number of electric space heating customers and their annual average consumption rates can be obtained from several sources. These.include the following: 1. natural gas consumption data, 2. "typical" home space· heating analysis, and 3. average annual and monthly residential consumption per customer information of electric utilities. In addition, information on housing unit sizes, in terms of average square feet, and the relative sizes and heating efficiencies of different- sized units can be helpful. Using data obtained from Alaska Gas and Service Company, an analysis of natural gas consumption for space heating was done. This analysis indicates that for. residential gas utility customers (assumed to be single-family, mobile home, and duplex units and all space heated with gas) average annual consumption for space heating (that portion of the load which varies over the course of the year with heating degree days) is about 84 percent of total consumption or 187 mcf/customer/year for the 1970s and 175 mcf/customer for 1978. Details of this analysis are shown in Table D.l4. The analysis further reveals no trend in consumption per customer in the 1970s, which might be the result of either larger homes or better insulation. Personnel at Alaska Gas and Service Company using national data estimate the space heat load at 75 percent of the total. Recog- nizing the imprecision of both the regression model and the national D-27 TABLE D.l4. DETERMINATION OF AVERAGE NATURAL GAS CONSUMPTION IN THE RESIDENTIAL SECTOR A. Alaska Gas and Service Company Residential Consumption Data Consumption Customers Average Consumption (mcf) (mcf) Eagle Soldotna/ River Kenai Anchorage Total 1970 2,615,042 12,097 216.2 1971 3,406,227 15,233 223.8 . 1972 3,817,869 16,f31 235.2 1973 4,162,662 17,983 231.5 1974 4,312,701 20,135 214.2 1975 5,402,111 22 '779 257 237 236 237.2 1976 5,765,871 25,659 225 225 225 224.7 1977 5,848,812 27,901 262 243 206 209.6 1978 6,367,015 30,629 199 196 209 207.9 1979 6,730,022 33,229 202.5 Average 1970,....1978 222.3 Normalized by Heating Degree Days 226 Heating Degree Days (annual) 10,137 11,879 12,016 11,.665 10,683 11,308 .10,361 9,394 9,131 - 10,730 10,911 SOURCE: Annual financial report to APUC and internal company records. D-28 > TABLE D.l4.· (continued) DETERMINATION OF AVERAGE NATURAL GAS CONSUMPTION IN THE RESIDENTIAL SECTOR Honthly Natural Gas Sales Data Sales/Customer (mcf) Heating Degree Days (Anchorage) Year Residential Commercial Industrial Current Lagged 1 Month 1974 26 100 NA 1,425 1,263 1975 37 140 872 1,643 1,425 1975 8 31 345 192 354 1975 7 29 360 252 192 1975 30 109 640 1,654 1,517 1976 36 136 889 1,485 1,654 1976 7 29 307 184 332 1976 7 25 310 262. 184 ,~. ' D'ec: 1976. 25 92 706 1,294 1,028 u~ti 1977 40 142 606 1,017 1,294 ~uly 1977 7. 26 306 75 208 ~~g 1977 6 26 337 144 75 .rlec 1977 32 117 836 1,659 1,486 'Jan 1978 31 118 813 1,349 1,659 quly 1978 9 31 290 186 308 '!\ug 1978 7 23 281 160 186 C(~. Dec 1978 25 90 708 1,344 1,153 Jan 1979 28 101 788 1,321 1,344 July 1979 7 24 285 NA 265 Aug 1979 6 19 179 NA SOURCE: Alaska Gas and Service Company records. D-29 TABLE D.l4. (continued) DETERMINATION OF AVERAGE NATURAL GAS CONSill1PTION IN THE RESIDENTIAL SECTOR C. Regression Results Equation: monthly consumption a + b * degree days a is interpreted as monthly, nonclimate related gas consumption in mcf or simply nonspace heat related consumption Dependent Variable a Value Standard Error of Equation Independent Variable monthly residential consumption monthly residential consumption monthly residential consumption monthly commercial consumption monthly commercial consumption monthly commercial consumption monthly industrial consumption monthly industrial consumption monthly industrial consumption 2.86 .93 2.87 .88 4.65 .86 10.81 .93 10.93 .89 17.70 .86 239.69 .91 242.94 .93 257.23 .92 D-30 3.60 4.49 5.06 12.58 16.18 18.15 74.09 63.36 68.49 heating degree days (HDD) lagged 2 month average HDD HDD HDD lagged 2 month average HDD HDD HDD lagged 2 month average HDD HDD data, we average the estimates to obtain 80 percent as the space heat load of residential natural gas sales. Table D.l5 shows the total number of residential natural gas space heat customers, including both gas utilities in the Greater Anchorage Area. Average annual consumption in the residential sector of the Kenai sy~tem was 90 percent of the Anchorage system in 1978. TABLE D.l5. 1978 RESIDENTIAL NATURAL GAS CUSTOMERS Anchorage Kenai Soldotna/North Kenai Eagle River Total 27,664 910 1,103 1,856 31,533 SOURCE: Alaska Public Utilities Commission records. The University of Alaska Fairbanks, Cooperative Extension Service, has developed a model which is capable of analyzing the fuel requirements necessary to heat a typical house with design specifications chosen and input by the model user. This model has calculated the annual fuel requirements shown on Table D.l6. D-31 ·;: TABLE D.l6. "TYPICAL" HOUSE SPACE HEATING FUEL REQUIREMENTS Single Family House 1. 2300 square feet (2 floors with daylight basement) natural gas (mcf) electricity (kWh) fuel oil (gallons) 2. 768 square feet (closed crawl space) electricity (kWh) fuel oil (gallons) 3. 768 square feet (heated crawl space) electricity (kWh) fuel oil (gallons) Mobile Home 1. 768 square feet (closed crawl space) electricity (kWh) fuel oil (gallons) 2. 768 square feet (heated crawl space) electricity (kWh) fuel oil (gallons) Anchorage 200 40,917 1,492 Fairbanks 52,392 1,910 29,042 1,059 26,620 970 34,873 1,272 32,761 1,194 SOURCE: Axel Carlson, Extension Engineer, Cooperative Extension Service, University of Alaska, Fairbanks, as reported in Fairbanks North Star Borough, Community Information Center, Special Report #2, 1978, and Special Report #4, 1976. D-32 Recent trends in average annual residential consumption per customer are depicted in Figure D.l. The rapid growth for Golden Valley Electric Association (GVEA), Matanuska Electric Association (MEA), and Homer Electric Association (HEA) is primarily due to space heating load. Table D.l7 shows electric utility monthly residential loads per customer in 1979 for.several utilities. The winter-to-summer month ratios provide some indication of the space heating load. Copper Valley Electric Association (CVEA), with no significant space heating load, has a winter-to-summer ratio of 1.48. The other utilities with space heating load have higher winter peaks. Unfortunately, this information is not precise enough to allow one to draw inferences about the amount of load devoted to space heat for the various utilities. An attempt was made to calculate the space heat load for the average space heat customer; but the results, shown in the final row, are implausible. Information on the average size of units in the housing stock is available for Fairbanks and is shown in Table D.l8 along with national averages. Anchorage retailers indicate that trailer dimensions have been increasing over time, although it is difficult to use sizes of trailers sold to estimate size of trailers in place. This is because people tend to add to trailers in place. Newer trailers are 924 square feet (14x66), while those sold in the early 1970s are 732 square feet (12x61). D-33 Figure D.l. Residential Electricity Consumption Per Customer II ~mh/ .·: year '1_ 1 _ -i r I ..... -,.- ,.-,..- ,.,. .. -,, -------:---.. ____ , L ___ , --------1····-----~-----t-----·------·----··-----.. , _______ -· - 6S t,(> 67 (,i ~1 70 7/ 7:J.. 73 7•{ YEAR D-34 \ :HEA ~~· /" GVEA """"-~ ~__.., CEA *.AP&T --I . --J---------· 71 7;?, TABLE D.l7. MONTHLY RESIDENTIAL ELECTRIC UTILlTY LOAD FOR 1979 CVEA CEA AMLP HEA MEA GVEA January 620 1,179 1,131 1,418 2,017 1,308 February 646 1,324 762 1,501 1,936 1,495 March 562 1,127 1,062 1,407 1,691 969 April 525 856 783 1,183 1,396 803 May 466 779 678 1,004 1,079 637 June 432 741 568 909 903 613 July 371 726 563 740 850 562 August 426 583 482 737 771 592 September 432 779 611 720 834 671 October 434 783 410 849 962 743 November 571 953 666 1,002 1,245 887 December 549 1,279 917 1,216 1,590 1,258 Monthly Average 491 871 716 1,054 1,270 877 Winter-Summer R . a at~o 1.48 1.84 1.74 1. 73 2.20 2.30 Total 5-,892 10,452 8,592 12,648 15,240 10,524 Nonspace Heat Load b 5,892 9,828 7, 726 11,429 12,090 8,464 Total Minus Nonspace Heat 0 624 866 1,219 3,150 2,060 Percent Space Heat CustomersC 0 .14 15 30 33 6 Space Heat Average 4,457 5,907 4,063 9,545 34,333 SOURCES: Utility Monthly Reports to Rural Electrification Association and internal utility records. (a) (December + January+ February)/(June + July + August) (b) Based upon the CVEA ratio of total annual sales to sales in the summer months of June, July, and August (4.79). (c) Author estimate. D-35 TABLE D.l8. AVERAGE SQUARE FEET FOR VARIOUS HOUSING TYPES FAIRBANKS AND NATIONAL DATA Fairbanks a ' Nationalb Single-Family 1,384 1,570 Duplex 796 1,370 Apartment 847 900 Mobile Home 919 720 Total 1,116 aFairbanks North Star Borough, Community Information Center, "1978 Fairbanks Energy Inventory," Special Report No. 4, July 1979, p. 40. b U.S. Department of Energy, Office of Buildings and Community Systems, Comprehensive Community_Energy Planning: A Workbook, Vol. 1. D-36 .New homes constructed in the Anchorage area tend to be in the range of 1,600 square feet on average according to real estate personnel. This includes a mixture of one-story ranch homes, split levels, and other types. The average house also seems.to be getting larger. This is consistent with the hypothesis that the Alaskan housing stock is being upgraded toward the national average. In 1977, the average size of new, single-family homes built in the United States wa~ 1,720 square 10 feet. Evidence that the average size of the Alaskan housing unit was smaller than the national average in 1970 can be inferred from earlier data from the 1970 census shown in Table D.2 on the median number of rooms and population per housing unit for Alaska. In all railbelt census divisions, median number of rooms was lower and population per occupied housing unit was higher than the average for the western region of the United States. • Real estate and electric utility personnel indicate that housing units are smaller in the outlying areas of the railbelt such as Seward, Homer, and Valdez. Finally, Table D.l9 shows national average estimates of average size and thermal requirements fo~ various types of structures. This table demonstrates the apparent variation in thermal requirements of different types of buildings. Also, the very high average electrical requirements calculated by the author for Anchorage, based upon the thermal requirements data, bring into question the use of nationally determined formulas and ratios for the Alaskan railbelt. D-37 TABLE D.l9. NATIONAL AVERAGES: RESIDENTIAL SPACE HEATING Anchorage a Thermal Electrical Average Size Requirement Type of Structure (square foot) (btu/sg.ft./HDD Single Family Detached 1~570 11.3 56~716 Single Family Attached 1,370 6.2 27,154 Multifamily High Rise 900 4.5 12~947 Multifamily Low Rise 900 5.0 14,386 Mobile Home 720 15.0 34,526 aCalculated on the assumption of 10,911 heating degree days. SOURCE: 11 Comprehensive Community Energy Planning: A Workbook," for the Office of Buildings and Co~~unity Systems, U.S. ment of Energy, Volume 1, pp. 4-7. D-38 To begin the actual determination of residential electric space heating, based upon all this fragmentary information, it is possible to first net out natural gas users in Anchorage and Kenai-Cook Inlet among single-family, duplex, and mobile home residents. We assume all resi- dential gas customers use gas for space heating. We can thus calculate the non-natural gas c~stomers according to Table D.20. In order to divide total gas customers among structural types, we note that trailers are somewhat less likely to be heated by gas according to the 1975 Anchorage survey and use this information for both census divisions. In 1970, fuels other than natural gas, electricity, and fuel oil accounted for 7 percent of space heating units in Anchorage and 13 per- cent in Kenai-Cook Inlet. If we assume no additions to the number of such units and no conversions, these percentages fall to about 3 and 8 percent, respectively, by 1978. It seems reasonable to assume that units burning these fuels would not be multifamily but might otherwise be randomly distributed among different types of structures. Netting these units out leaves only electric and fuel oil-heated units. Information on electricity and fuel oil for Anchorage consists of the census, the l975 Survey, and all-electric homes data; and these sources are not consistent. According to the census, 6 percent of residences were electrically space heated in 1970; while only 4 percent were all-electric homes in the Anchorage utility service areas. Growth in the proportions of all-electric homes was rapid in the early 1970s, based on new all-electric homes as a proportion of all new homes (about D-39 TABLE D.20. NON-NATURAL GAS CUSTOMER CALCULATION Natural Gas Non-Natural Gas Total Units Customers Customers Anchorage 39,700 29,520 10,180 Single Family 28,530 22,437 6,093 Duplex 4,581 3,603 978 Mobile Home 6,589 3,480 3,109 Kenai-Cook Inlet 6,177 2,013 4,164 Single Family 3,660 1,363 2,297 Duplex 484 180 304 Mobile Home 2,033 470 1,563 SOURCE: See text. D-40 17 percent). If that growth held through the mid-1970s, then using the all-electric homes information as a base, about 5,800 units would be all-electric presently, or about 10 percent of the total. Discussions with realtors indicate a significant proportion of multifamily units built in the middle 1970s were electrically heated. The number of building permits issued for multifamily units between 1974 11 and 1977 was 6,855 (including duplexes until 1977) or about 44 percent of total permits. If their electric space heat installation proportion was double that of the historical trend, then another approximately 1,000 multifamily units were built that were electric. Apart from these multifamily units (including condominiums), electric space heating is allocated based upon total units of each type (not heated by propane, wood, or coal) after an up•vard adjustment of the total by 3 percent to account for the discrepancy between the census and the Federal Power Commission's estimates of all-electric homes and the growth in housing in areas not served by gas. In Kenai-Cook Inlet, the census indicates 4 percent electric space heat in 1970; while the proportion from the 1977 Survey for the three cities of Kenai, Soldotna, and Seldovia. is 7 percent. Average residen- tial consumption data for Homer Electric Association (HEA) is indicative of a substantial growth in space heating load in the last few years. In 1977, a mail survey of their service area indicated a 33 percent electric space.heat proportion for 7,171 active accounts.12 Based upon consump- tion data, the 7 percent figure is obviously lm~, which is logical D-41 because Kenai and Soldotna have access to natural gas; while Homer itself does not. We scale down to 30 percent the REA figure on the assumption that it might be somewhat of an over-estimate for 1977 for first homes but that the electric heat load trend was upward throughout 1977 and 1978. The load is distributed among all struc.tures propor- tionately after netting out propane, wood, coal, and other fuels. For Matanuska-Susitna and Seward, natural gas is not an available option. Seward Electric Association estimated 2 percent all-electric 13 homes, and the census reported 4 percent. We compromise on 3 percent and allocate them all to single-family and mobile home units. For Matanuska-Susitna, the census reports 1 percent electrically heated homes, but much of the growth in the housing stock since then has been electrically heated units. If all housing added since 1970 was elec- trically space heated, the proportion could be over 50 percent even without· retrofitting electric systems. A portion of the Matanuska Electric Association (MEA) serVice area is in Eagle River where natural gas is available. Of total MEA residential customers in 1978, about 3,500 were located in the Anchorage Census Division and 1,856 of these were gas customers. Thus, when MEA calculates 2,685 to be the number of all-electric customers on their system (26 percent), this converts to about 40 percent in the Matanuska-Susitna Borough proper. On the basis of the criterion used by MEA to identify electric space heat customers (2,500 kWh or more on the December bill), they may have underestimated their electric space heating load. With this in mind, as "to7ell as a com- parison of the average residential bill with REA which is estimated at D ... 42 percent electric, we raise our estimate of electric space heating to percent. To distribute this among types of units, we assume that use coal, wood, and propane has fallen from 37 to 15 perc~nt of the total not utilized in multifamily units at all. Thus, the allocation of this amount. For Fairbanks, we have information from the census and from the · all-electric home data, both of which indicate about 7 percent electric mode split in 1970 and the latter which indicates that the proportion remains constant during the early 1970s. A recent mailback survey reported 11 percent late in 1979.14 The Fairbanks utilities say that growth was rapid ill. the mid-1970s but that now, because of the high price of peak electric power, they are discouraging the use of electricity for space heating. In April 1975, Golden Valley Electric Association (GVEA) put a prohibition on further electric space heat installations. In addition, they are assisting people to get off electric space heat and, as a consequence, the average residential consumption fell from a peak of 17,332 kWh in 1975 to 10,524 kWh in 1979. (Their average bill in 1970 was 10,785 lUfu.) On this basis, an electric space heat load of about 800 units (the 1970 number) would be reasonable for 1979 and a slightly higher number for 1978. GVEA estimates that they currently have about 750 electric space heat customers, which would be about 6 percent. ·we extrapolate back to 1978 and estimate 1,000 electric space heat units in that year whiah would be 8 percent of the total stock served by both utilities. D-43 To allocate the electric heat among units, we estimate 20 percent of the housing stock is now heated by coal, wood, propane, steam, and other but that those fuels are not used in multifamily units. The electric space heat is evenly distributed among all types of units net of these fuels. The resulting electric space heat mode splits are shown in Table D.21. Average consumption data is available for past years from the Federal Power Commission All-Electric Homes reports, from engineering analyses, and from inferences dra~vn from natural gas consumption. Since this last source contains the most recent information and is from a known source, it forms the basis for determining average consumption which will vary according to: 1. location (heating degree days), 2. type of structure, and 3. size of structure (square feet of floor space). The age of the structure and the habits of the occupants are impor- tant sources of variation which cannot be formally addressed at present. In addition, the heating load can vary considerably from year-to-year because of variation in weather conditions. Using the average annual heat load of 162 mcf calculated for Anchorage Natural Gas customers in 1978, we convert to k\~ of elec- tricity on the assumption of 65 percent efficiency for gas space heating and 95 percent for electricity resulting in an electric equivalent of approximately 32,400 k~. This is the average among three types of D-44 TABLE D.21. 1978 ELECTRICAL SPACE HEATING PERCENTAGES Single Multi-Mobile Family Duplex Family Home Total Greater Anchorage Area 18.8 18.4 19.9 18.4 19.0 Anchorage 12.9• 12.9 18.9 11.9 14.8 Kenai-Cook Inlet 30.0 30.0 32.7 29.4 30.0 Matanuska-Susitna ·49.5 49.5 58.7 49.5 50.0 Seward 4.0 0 0 5.0 3.0 Greater Fairbanks Area Fairbanks 7.3 7.2 10.0 7.3 8.0 ~ Glennallen-Valdez Area Valdez-Chitina-Whittier 2 0 0 0 0 D-45 structures of different characteristics and sizes. There are estimated to be 23,800 single-family units, 3,783 duplexes, and 3,950 mobile homes using natural gas, for a total of 31,533 units. Using this information as well as average structure size and space heating efficiency estimates, the average electric heating load for Anchorage by type of structure can be calculated. This calculation is shown in Table D.22. Floor space is the average of the values for Fairbanks and the national average. Heating efficiency factors for duplexes are based on the idea that the heating requirement is a function of wall and roof surface area, which increases less than proportionally as floor space increases. Specifically, two duplex units with 1,085 square feet each have a floor area of 2,170 which is 1.47 the area of the average single- family unit. The duplex wall and ceiling area, however, is about 1.38 times the single-family unit, indicating that heating requirements per square foot of floor space will be less. The ratio of outside wall-to- floor space for the duplex is about 1.85 and for the smaller single- family house, 2.05. On this basis, one can calculate that the duplex is about 10 percent more efficient to heat on a square foot basis. Studies in Fairbanks indicate that a mobile home requires 20 percent more energy to heat per square foot than a single-family unit of the same size. Using this assumption ~nd the fact that the average mobile home has a surface area-to-floor space ratio of about 2.35 (which would mean D-46 ,1.:, Type of Unit Single-Family Duplex Mobile Home t::l I .p. ....... Total TABLE D.22. CALCULATIO~ OF ANCHORAGE ELECTRIC SPACE HEAT LOAD BY TYPE OF UNIT Number of Units 23,800 3,783 3,950 31,533 Average Floor Space (square foot) 1,480 1,085 820 1,350 Total Floor Space 35,224,000 4,104,555 3,239,000 42,567,555 Average Heat Requirement Per Square Foot Re~ative to Total 1.0 .9 1.38 1.02 Average Space Heat Load/Square Foot (kWh) 23.53 21.18 32.47 24 Average Space Heat Load (kWh) 34,823 ·22,976 26,626 32,400 it consumes about 15 percent more energy per square foot of floor space f• than the average single-family home), the average heat ~equirement factor for mobile homes relative to single-family units thus becomes 1.38. The requirement for a multifamily unit is calculated similar~y to that of a duplex assuming an av·erage unit size of 900 square feet and that the average multifamily structure is 8 units. The ratio of surface area to floor space is calculated as 1.48. Compared to the single- family ratio of 2.05, the heating requirement on a square foot basis would only be 72 percent as large. Since the floor space of the multi- family unit is 61 percent that of the single-family unit, the overall energy requirement is calculated at 44 percent of the single-family unit. Adjustments for other parts of the railbelt are made on the basis of average size of units and average heating degree days. The former is directly available only for Fairbanks, but a_general idea of average size of unit is available from information on median rooms from the census. Outlying parts of the Greater Anchorage Area h~ve somewhat smaller units, and Glennallen-Valdez has considerably smaller units than Anchorage. We assume per-unit heating requirements outside Anchorage are 20 percent less for Kenai-Cook Inlet and Seward and 15 percent less for Matanuska-Susitna than in Anchorage on the basis of smaller average unit size. For Fairbanks, we assume the average size of units is 92 percent D-48 of Anchorage, based on actual survey data for Fairbanks on single-family, duplex, and mobile home units. No difference in heat requirements, based upon variation in heating degree days w~thin the Greater Anchorage Area, is assumed. The ratio of heating degree days in Fairbanks-to-Anchorage for the average of the 1977 and 1978 seasons was 1.43. This information, in addition to aver- age unit size i~formation, results in an estimate of Fairbanks unit requirements at 132 percent of those in Anchorage. For Glennallen-Valdez, we assume the average unit is 75 percent of the size of an Anchorage unit and average the heating degree days in the different parts of the census division to obtain a ratio to Anchorage of 122 percent. Combining these yields an estimate for Glennallen-Valdez which is about 91 percent that of Anchorage. The results of these cal- culations are presented in Table D.23. For projection purposes, it is necessary to adjust the figures on average annual electric space heating requirements to account for the fact that 1978 was a warmer-than-normal year. Adjustment factors, based on the ratio of normal to 1978 heating degree days, are used in the projections and presented in Table D.24. D-49 TABLE D.23. 1978 AVERAGE ANNUAL ELECTRIC SPACE HEATING REQUIREMENT GREATER ANCHORAGE AREA Anchorage Kenai-Cook Inlet Matanuska-Susitna Seward GREATER FAIRBANKS AREA Fairbanks/Southeast Fairbanks GLENNALLEN-VALDEZ AREA Valdez-Chitina- Whittier (kWh) Single Family Duplex Multifamily 34,800 23,000 15,300 27,800 18,400. 12,200 29,600 19,600 13,000 27,800 18,400 12,200 45,900 30,400 20,200 31,700 20,900 13,900 D-50 ,. ·. Mobile Home · 26,600 21,300 22,600 21,300 35,100 / TABLE D.24. HEATING DEGREE DAY COMPARISONS Ratio·of Average of 1977-78 Normal and 1978-79 Normal to.Recent Anchorage 9,548 10,911 1.14 Fairbanks 13,719 14,344 1.05 Glennallen-Valdez 11,621 12,241 1.05 D.3. MAJOR APPLIANCE SATURATION RATES Present appliance saturation rates (Table D.25) must be based upon past Alaskan trends in saturation rates, trends in other states, and national trends in the percentage of wired homes which have at least one of a particular appliance. This is because current data on saturation rates is not available. Because future saturation rate projections utilize the same methodology used to develop the present estimate~, they are also discussed in this section. D.3.A. Ranges The saturation rate is essentially 100 percent for ranges; it is not assumed to change in the future. , D. 3.B. .Refrigerators The penetration rate (number of households with at least one unit/ numbeh of households) for refrigerators is assumed to be approximately 100 percent. The saturation rate may exceed 100 percent, however, D-51 TABLE D~25. 1978 CALCULATED APPLIANCE SATURATION RATES Appliance Census Division Matanuska-Kenai-Glennallen- Anchorage Susitna Cook Inlet, Seward Fairbanks Valdez Hot Water 100 91 94 93 97 91 Clothes Dryers 71 . 65 76 73 66 48 Freezers 42 62 56 66 42 43 Clothes Washers 76 82 85 83 74 65 1:::1 I VI N Television Sets 156 108 104 147 149 80 Dishwashers 50. 35 31 45 36 11 Ranges 100 100 100 100 100 100 Refrigerators 100 100 100 100 100 100 Room Air Conditioners 0 0 0 3 1 0 because of multiple ownership. The California Energy Commission, for 15 example, estimates a saturation rate of between 113 and 116 percent. It can safely be assumed that the colder climate of Alaska reduces the incidence of second refrigerators below that of California, but a satura- tion rate in excess of 100 is still a possibility. Since no information is currently available on this possibility, however, it is assumed that the saturation rate remains at 100 percent. D.3.C. Air Conditioners The incidence of room air conditioners is rare in Alaska. The 1970 saturation rates are assumed to continue in all future years (Table D;26). D.3.D. Dishwashers The saturation of dishwashers should be rel&ted to the rate of growth of the housing stock since a large proportion of new housing is built with dishwashers. It should also be related to the incidence of households with two wage earners because it is a labor-saving device. On this basis, we assume that the growth in the dishwasher saturation rate since 1970 is slightly in excess of the national average. In 1970, the various Alaskan census divisions were close to or slightly lower than the national average and the Western Region U.S. saturation rates. Nationally, dishwasher saturation rates have grown at 6 percent annually since 1970, as calculated from data in Table D.27. For Alaska, .. we assume 7 percent for A~chorage, Fairbanks, and Seward and 8 percent D-53 TABLE D.26. Hot Water Range GREATER ANCHORAGE AREA Anchorage 1960 1970 Matanuska-Susitna 1960 1970 Kenai-Cook Inlet 1960 1970 Seward 1960 1970 GREATER FAIRBANKS AREA Fairbanks 1960 1970 GLENNALLEN-VALDEZ AREA Valdez-Chitina-ffi1ittier 95 99 64 82 53 88 79 87 88 94 1960 66 1970 57 WESTERN REGION U.S. 1960 1970 99 99 100 100 99 100 98 93 100 94 100 97 98 100 A~J>L.IANCE SATURATION __ RATESa _ Clothes Dryer 28 59 15 47 18 55 12 53 21 48 6 35 45 Clothes Dish Freezer Washer Washer 29 39 43 67 35 61 33 71 21 39 19 40 28 62 63 65 68 60 70 72 83 58 61 39 54 70 NA 29 NA 19 NA 17 NA 26 NA 21 NA 6 NA 27 . -----------·--------------------------------- Room Air Television Conditioning 99 121 61 84 43 81 6 114 88 116 26 40 122 1 0 0 0 0 0 0 3 1 1 0 0 15 aCalculated as the number of appliances divided by occupied housing units. A response of twq or more television sets or room air conditioners is counted as two sets. SOURCES: 1960 Census of Housing, General Housing Characteristics: Alaska. Tables 29, 30. 1970 Census of Housing, Detailed Housing Characteristics: Alaska. Tables 62, 63. 1970 Census of Rousing, Detailed Rousing Characteristics: United States Summary Tables 23, 24. TABLE D.27. PERCENTAGE OF ELECTRICALLY WIRED HOMES WITH SELECTED APPLIANCES 1960 1965 1970 1973 1974 1975 Rooni Conditioners 15.1 24.2 40.6 48.9 51.6 52.8 Dishwashers 7.1 13.5 26.5 34.3 36.6 38.3 Clothes Dryers (include gas) 19.6 26.4 44.6 53.9 56.5 57.7 23.4 27.2 31.2 37.9 41.7 43.5 Televisions (color) 9.5 42.5 67.1 71.5 74.4 Televisions (black & white) 89.4 97.1 98.7 99.9 99.9 99.9 Clothes \vashers 68.3 57.4 62.1 67.8 68.4 69.9 Refrigerators 98.2 99.5 99.8 99.9 99.9 99.9 1976 54.4 39.6 58.6 44.4 77.7 99.9 72.5 99.8 SOURCE: U.S. Department of Commerce, Statistical Abstract, annual. D-55 1977 55.3 40.9 59.3 44.8 81.3 99.9 73.3 99.9 for the other smaller census divisions with lower 1970 saturations. Subsequent to 1980, the saturation rate (St) is determined by the following equation: = S l + (1-S 1 ) * t-t-.02 The parameter. .02 is based upon a rough estimate of the long-run national trend. D.3.E. Clothes Washers With respect to clothes washers in 1970, most Alaskan railbelt census divisions had saturation rates relatively close to the national and Western Region U.S. rates. Thus, the same growth rate observed nationally between 1970 and 1977 is applied to Alaskan census divisions to generate 1978 saturation rates. This procedure is modified ,only in the case of Seward, where it would result in a saturation rate in excess of 100 percent. Th~ saturation rate for Seward is arbitrarily maintained at its 1970 level of 83 percent. Subsequent to 1978, the increase in saturation-rates is assumed to moderate as a practical upper limit on clothes washer. saturation would be less than 100 percent. This is based upon the assumption that most apartment units will have washer and dryer units for tenants but at less than a one-to-one ratio. Consequently, the growth of clothes washer satu.rations are based upon the following equation: D-56 D.3.F. Clothes Dryers The 1970 Alaskan saturation rates for clothes dryers are close to the national average except for Anchorage, where it is considerably higher. Thus, all census division saturation rates except Anchorage are assumed to grow at the national average rate between 1970 and 1978. Anchorage already had reached the· average nation~l saturation rate for 1978 of 59 percent in 1970. Also, Anchorage figures reflect a higher- than-average ratio between clothes dryers and clothes washers. We base the Anchorage clothes dryer estimate on the clothes washer estimate and on this 1970 ratio. Subsequent to 1978, it is assumed that the ratio between clothes dryers and washers will remain constant so that clothes dryer saturations grow at the same rate as clothes washer saturations. D.3.G. Freezers Alaskan freezer saturation rates were consistently and considerably above the national and Western Region rates for 1970. In addition, the rural census divisions had considerably higher saturations than the urban areas, suggesting that the high saturation might be partially attributable to the unavailability of adequate grocery facilities. It seems reasonable to assume that this will continue to be .an important factor in determining future freezer saturation rates and, in addition, that the strong hunting and fishing interests of the typical Alaskan will contribute to high freezer saturation rates. D-57 .. Nationally, the freezer saturation rate has grown about 5 percent annually since 1970, compared to 3 percent annually in the previous decade. These growth rates applied to 1970 Alaskan saturation rates would give unreasonably high values for 1978 and subsequent s<~turation levels. We assume that for the Anchorage, Fairbanks, and Valdez-Chitina- Whittier Census Divisions, the annual saturation rate growth rates are one percent. For the other census divisions with unusually high 1970 saturation rates, we assume that they decline by one percent annually to reflect growth in those areas close to adequate grocery facilities. Subsequent-to 1978, those census divisions which have increasing saturation rates continue to grow, based on the following equation: while those census divisions with declining saturation rates are assumed to maintain constant saturation rates in future years. D.3.H. Water Heaters Alaskan 1970 water heater saturation rates show surPrising varia- tion among census divisions. Here national data cannot serve as a guide because the national average is virtually 100 percent. This was the case in Alaska only for Anchorage. One possible explanation for the relatively low saturation rates might be the fact that a percentage of D-58 the population in 1970 lived outside utility service areas and also areas where petroleum products were readily available. This proportion population has been greatly reduced since 1970 as population grmv-th has centered in the urban areas. One pattern that does emerge from examinatio.n of the historical Alaskan saturation data is that between 1960 and 1970, the saturation rates grew rapidly toward 100 percent (except in the inexplicable case of Valdez-Chitina-Whittier). On this basis, it is reasonable to assume ~ a continuation of those growth trends in the 1970s. We assume that between 1970 and 1978 nonsaturation (1-S) is reduced by 50 percent in each instance, except for Valdez-Chitina-Whittier which is arbitrarily set equal to Matanuska-Susitna. Subsequent saturation rate growth results in 100 percent saturation by 1990. D.3.I. Television Sets The differences in television saturation rates between Alaska census divisions and the nation likely reflect the possibility of the household's receiving a television signal. Thus, Anchorage has a pat- tern close to the national growth, while the saturation of television sets in Seward grew dramatically between 1960 and 1970. Nationally, growth in saturation in the 1970s has been the result of increased ownership of color television sets, which nearly doubled between.l970 and 1977. We assume the same growth rates in the 1970s for Alaskan census divisions, except for Valdez-Chitina-Whittier which D-59 is assumed to follo\v the pattern of Kenai-Cook Inlet with an approximate ten-year lag. .. Future growth of television saturations will be modest if recent national trends can be used to gauge future trends. Alaskan growth will continue to lead national trends, but Alaskan saturations will not reach national levels. We assume a growth in saturation as follows: st = s 1 + (2-s . 1 ) "' . 02 t-t- D.4. MAJOR APPLIANCE FUEL TYPE MODE SPLIT Four. appliances--water heaters, ranges, clothes dryers, and refrigerators--may normally be designed to operate on fuels other than electricity. The appliance mode split, defined as the proportion of .. appliances using a particular energy fuel, is determined by current relative prices of fuels and appliances and consumer tastes, as well as the past values for these variables. That. is, the current mode split will partially be a reflection of past patterns of relative prices which are no longer relevant for the purchase of new appliances so that the mode split observed at any point in time may not be an equilibrium ·split. Because mode split is determined by these factors, national infor- mation is not particularly applicable to Alaskan conditions. In order D-60 to calculate mode split, we thus rely on Alaskan information. Because information on consumption of other fuels in the~e appliances is rele- . vant to the determination of what percentage of the appliance utilizes electricity, it is utilized wherever appropriate. (It is basically used to provide a check on the mode split by insuring that some alternative fuel has not been allocated an unrealistically large or small percentage of total appliances of a particular type.) Information to develop the mode split estimates sho~vn in Table D.28 comes primarily from the census of housing (Table D.29) and conversations with utility, horne construction, and real estate personnel. D.4.A. Within the Natural Gas Service Area The starting point of the analysis is to net out, where applicable, natural gas utility sales. No natural gas refrigerator sales have been identified, except for campers and recreational vehicles; so water heaters, ranges, and clothes dryers are the appliances which may be gas fired. From the 1970 census and more recent national data, one can calculate ratios between gas space heating customers and gas water heating and cooking customers. These ratios are shown in Table D.30. They indicate that the choice of gas water heat is closely asso- ciated with the gas space heat decision but that such may not be the case for cooking fuel. In particular, Anchorage is far ·below the national average for gas ranges among natural gas customers. D-61 TABLE D.28. 1978 ELECTRICAL APPLIANCE MODE SPLITS CENSUS DIVISION Greater Matanuska-Kenai-Anchorage Anchorage Susitna Cook Inlet Seward Area Fairbanks Glennallen Water Heater 33 43 41 35 34 43 40 Range 66 75 35 53 64 81 40 Clothes Dryer 91 96 78 70 90 98 75 t;:j Refrigerator 100 100 • 100 100 100 100 100 I 0'\ N PROPORTION OF APPLIANCES USING VARIOUS FUELSa Utility Gas Oil Electric Coal Wood Propane Other GREATER ANCHORAGE AREA Anchorage 1960 0 41 39 15 0 4 1 1970 45 13 35 1 0 4 2 Matanuska-Susitna I=' 1960 0 19 50 20 4 I 7 0 Cl' 1970 0 28 38 11 UJ 12 11 0 Kenai-Cook Inlet 1960 0 60 21 7 0 12 0 1970 34 17 37 2 0 9 1 Seward 1960 33 57 0 0 5 5 0 1970 0 60 35 0 0 5 0 a Proportions sum to 100. SOURCES: 1960 Census of Housing, General Housing Characteristics: Alaska. Tables 29' 30. 1970 Census of Housing, Detailed Housing Characteristics: Alaska. Table 63. TABLE D.29. (continued) ( ' WATER HEAT MODE SPLIT INFORMATION PROPORTION OF APPLIANCES USING VARIOUS FUELSa GREATER FAIRBANKS AREA Fairbanks and Southeast Fairbanks . '1960 1970 GLENNALLEN-VALDEZ AREA Valdez-Chitina- Whittier a 1960 1970 Proportions sum to 100. Utility Gas 0 3 0 • 0 Oil 23 33 56 41 Electric 19 37 4 36 Coal 48 15 0 0 SOURCES: 1960 Census of Housing, General ·Housing Characteristics: Alaska. 1970 Census of Housing, Detailed Housing Characteristics: Alaska. . ·. . ' .: . . . ', Wood 0· 0 0 0 Propane 9 8 0 23 Tables 29, 30. Table 63. 'ttrt·trttt&'i#!tt#ittttatiS·Jrw·tt¥Atrtw''t&Jfifi&rf#ttSifftr'~;.~tfir±t&ttre¥wrew&&t¥ri¥!t*¥#ijAW®Mm,'i%~*+w&ttwer·B¥k-·:'w,·t.~:,r:v?+mi4~~~~~~~~1i~~-ik_~~~~,,J:~~"'~~J.:~~,~-~~~~:-i~~;\1A~-,"~: .. ·( ... ~·.}.·. Other 1 4 40 0 PROPORTION OF APPLIANCES USING VARIOUS FUELS a Utility Gas Oil Electric Coal Wood Propane Other GREATER ANCHORAGE AREA Anchorage 1960 2 8 64 0 0 26 0 1970 20 1 65 0 0 14 0 Matanuska-Susitna t:;j I 1960 6 7 38 1 15 33 0 ~ \.J1 1970 0 1 44 0 16 38 1 Kenai-Cook Inlet 1960 21 17 23 8 0 31 0 1970 30 4 24 1 1 40 0 Seward 1960 0 29 35 0 4 32 0 1970 4 15 47 0 0 34 0 a Proportions sum to 100 •. SOURCES: 1960 Census of Housing, General Housing Characteristics: Alaska. Tables 29' 30 • 1970 Census of Housi\T' Detailed Housing Characteristics: Alaska. Table 63. TABLE D.29. (continued) COOKING APPLIANCE MODE SPLIT INFORMATION PROPORTION OF APPLIANCES USING VARIOUS. FUELSa Utility Gas Oil Electric Coal Wood "Propane Other GREATER FAIRBANKS AREA Fairbanks and Southeast Fairbanks 1960 3 2 62 2 2 29 0 1970 4 1 76 0 1 18 0 ~ I "' "' GLENNALLEN-VALDEZ AREA Valdez-Chitina- Whittier 1960 3 19 29 0 14 35 0 1970 6 6 18 0 5 65 0 aProportions sum to 100. SOURCES: 1960 Census of Housing, General Housing Characteristics: Alaska. Tables 29' 30. 1970 Census of Housing, Detailed Housing Characteristics: Alaska. Table 63. ·'· £." '~:; ._, A .... :. ;;. ;._ ~.. ', TABLE D.29. (continued) CLOTHES DRYING APPLIANCE MODE SPLIT INFORMATION PROPORTION ..OF APPLIANCES USING VARIOUS FUELSa Utility Gas Oil Electric GREATER ANCHORAGE AREA Anchorage 1960 1 0 99 1970 9 0 91 Matanuska-Susitna t;; I 0\ 1960 0 0 100 ....., 1970 7 0 93 Kenai 1960 15 0 85 1970 22 0 78 Seward 1960 18 0 82 1970 30 0 70 a Proportions sum to 100. SOURCES: 1960 Census of Housing, General Housing Characteristics: Alaska. Tables 29, 30. 1970 Census of Housing, Detailed Housing Characteristics: Alaska. Table 63. TABLE D.29. (continued) CLOTHES DRYING APPLIANCE MODE SPLIT INFORMATION PROPORTION OF APPLIANCES USING VARIOUS FUELSa Utility Gas Oil Electric GREATER FAIRBANKS AREA Fairbanks and Southeast Fairbanks 1960 1970 GLENNALLEN-VALDEZ AREA Valdez-Chitina- Whittier 1960 1970 a , Proportions sum to 100. 1 2 100 50 0 0 0 0 99 98 0 50 SOURCES: 1960 Census of Housing, General Housing Characteristics: Alaska. 1970 Census of Housing, Detailed Housing Characteristics: Alaska. Tables 29, 30. Table 63. TABLE D.30. PROPORTION OF GAS SPACE HEATING CUSTOMERS WITH OTHER GAS APPLIANCES Water Heater Range Anchorage 1970 .84 .38 Kenai-Cook Inlet 1970 .96 .94 United States 1970 1.00 .89 United States 1975 .79 SOURCES: 1970 Census of Housing, Detailed Housing Characteristics: Alaska. American Gas Association, Gas Facts, annual. We assume that since 1970 the ratios of gas water heating to gas space heating have approached 100 percent and the gas range ratio has increased in Anchorage and remained constant in Kenai-Cook Inlet. Growth in the housing stock in each census division has been about 50 percent since 1970 so that the differences between the national and local ratios could have declined by 30-to-50 percent. We assume 50 percent for water heaters to arrive at 92 percent for Anchorage and 98 percent for Kenai-Cook Inlet. Noting the declining ratio in gas ranges nation- ally, we estimate 50 percent for Anchorage and 90 percent for Kenai-Cook . Inlet. D-69 I This yields the mode spli~ estimates for gas water heaters and ranges for 1978, shown in Table D.31. TABLE D.31. 1978 GAS APPLIANCE MODE SPLIT ESTIMATES Gas Space•Heat Gas Hot Water Gas Range Anchorage .69 .63 .35 Kenai-Cook Inlet .33 .32 .30 Comparing these estimates to the 1970 census information on mode split indicates that the gas mode split has grown in Anchorage and has remained constant in Kenai-Cook Inlet. This is consistent with the observation that 'growth in Kenai-Cook Inlet has been balanced between areas accessible to gas and those not accessible to gas. To determine the electric mode split for these regions and appli- ances, use is made of an incremental mode split calculation using the 1960 and 1970 censuses. These incremental mode splits are calculated as the increments to the number of a particular appliance between 1960 and 1970 divided by the increment to total housing units over the same period (see Table D.32). Applying these incremental mode splits for electric water heating and cooking to the increase in housing units between 1970 and 1978 (about 50 percent) results in average mode splits which are reasonable when considered in relation to the natural gas mode splits, except for ranges in Kenai-Cook Inlet. These values are, therefore; D-70 TABLE D.32. INCREMENTAL MODE SPLIT CALCULATIONS BASED UPON HOUSING CENSUS DATA • Electric Electric Built-In Space Heating Water Heating Electric Electric· Electric Units Fuel Fuel Cooking Fuel Clothes Dryer GREATER ANCHORAGE AREA Anchorage .1403 '.1695 .2962 .6706 .9841 Matanuska-Susitna .0525 .0676 .2770 .7669 1.8919 Kenai-Cook Inlet .0815 .0638 .4989 .2378 .6410 Seward 0.0 -0.0486 .1991 .1644 .2431 t;;;l I "-J ...... GREATER FAIRBANKS AREA Fairbanks/Southeast Fairbanks .8320 .5553 1.6292 2.0463 2.3406 GLENNALLEN-VALDEZ AREA Valdez-Chitina- Whittier .0427 o.o .7974 -0.1897 . 7672 SOURCES: Census of Housing: 1970 Detailed Housing Characteristics Final Report HC(l)-B3, Alaska. U.S. Census of Housing: 1960 Vol. 1, States and Small Areas Final Report HC(l)-3, Alaska. used as the estimates, except that the Kenai-Cook.Inlet range estimate .. is increased based on Homer Electric Association survey data. For gas clothes dryers, the 1970 mode split is used for 1978. D.4.B. Outside the Natural Gas Service Area For Matanuska-Susitna, the information from. the census; is difficult to interpret because, bet>v-een 1960 and 1970, electric water heaters declined as a percentage, while those of wood and propane increased. Wood and propane also gained as cooking fuels, but electricity also increased at the expense of oil. One explanation of this phenomenon would be that the population increase was in relatively inaccessible areas and, thus, the same pattern of growth might not be expected to have occurred in the more recent decade. If we can assume that 90 percent of the appliances used by new residents since 1970 are oil or electric, then the proportions accounted for by coal, wood, propane, and other fall to about 18 percent for water heat and 30 percent for cookin& and 4 percent for clothes drying. This is based upon the observation that the number of residences has essent doubled since 1970. In addition, changeovers of existing appliances toward electricity and oil probably occurred, thus further decreasing the proportions of the other fuels. If this was about 25 percent, then the remaining'percentages would be 14 percent for water heat and 23 per- cent for cooking, with clothes drying assumed unchanged. D-72 In the absence of better information, we assume an equal mode split between oil and electricity for water heat and a continuation of the 1970 census ratio for cooking. In Seward, electric appliances made imp~essive gains over propane and oil in the 1960s as a proportion of the totals in spite of declines in the total housing stock. At the same time, oil maintained its share of the water heat market. In the absence of other information, we assume a continuation of the same mode splits for water heat and clothes drying. For cooking, we assume 75 percent of new households choose electricity, which brings the electric mode split up to 53 percent. Fairbanks electricity consumption for water heat increased in the 1970s, as did fuel oil consumption, both at the expense of coal which was rapidly phased out. We assume that no new housing since 1970 uses ·coal and that half the existing coal units have been converted. This leaves coal with about 4 percent mode split in 1978 since the housing units in Fairbanks have nearly doubled in the 1970s. We assume that the other fuels, except oil and electricity, maintain their 1960 mode splits and that oil and electricity maintain their 1970 ratio, which gives a 43 percent split for electricity. Electricity is preferred for cooking by the majority of Fairbanks households. If 85 percent of new households since 1970 chose it, then the electric cooking mode split would be about 81 percent. We assume the ·clothes drying mode split is unchanged from 1970. D-73 In the Glennallen-Valdez area, electric water heaters become much more pop~lar in the 1960s. On the other hand, the electric mode split for cooking fell. In the absence of other information, we assume the mode splits for the Glennallen-Valdez area move toward those of similar parts of the railbelt, as reflected in the 1970 census. D.5. ELECTRICAL APPLIANCE AVERAGE ANNUAL CONSUMPTION A reliable library of information on average electricity consump- tion by various appliances is just now beginning to emerge based upon manufacturer estimates, modeling efforts, and actual surveys which monitor consumer behavior. Some of the more widely circulated estimates appear as Table D.33. These estimates are all national averages. Examination of the various estimates reveals several important facts. First, only one--the Midwest Research Institute--is based upon actual metering of appliances; and it shows a considerable variance from other sources for the use of several appliances. Their results for ranges, dishwashers, and water heaters are below the others and for refrigerators, considerably above. This suggests actual use patterns may be considerably different from those assumed by manufacturers and other researchers. Second, the consumption of electricity by appliances depends upon the features of the appliance. In particular, larger refrigerators and D-74 ?:'A~L !;:__1;1_=-;i_~ AVERAGE ANNUAL ENERGY CONSUMPTION EST MATE FOR MAJOR APPLIANCES: VAR IOU S S OURCES (kWh /year ) 1976-1977 1968 1976 19 7 8 Midwest Edison Stanford 1970 1973 California California Research Electric ·Research Arthur D. Merchandising Energy Energy Institutea Instituteb InstituteC Littled Weeke Commissionf Commissiong Water Heater 4,046 4,811 4,490 5,626 4,515 3,025 4,876 Cooking 1,180 1,143 1,200 Range 782 700 2,071 778 Auto. Clothes Washer 88 103 98 88 90 70 76 including hot water 2,500 1,115 Clothes Dryer 1,032 993 990 937 993 950 1,212 Refrigerator (12 cu.ft.) 1,665 1,270 1,084 1,228 1,235 1,515 frostless (12 cu.ft.) 728 Refrig/Fr e ezer (12 cu.ft.) 1,217 t:l frostl e ss (12 cu.ft.) 1,500 I frostless (17 .5 cu.ft.) 2,250 ...... U1 Fr e ezer 1,342 1,348 1,480 1,294 regular (16 cu.ft.) 1,190 frostless (16.5 cu.ft.) 1,820 Di shwash e r 149 363 360 . 352 363 230 305 including hot water 2,100 925 Te levision -Black & White tube 350 360 solid state 120 general 362 140 129 Television -Color tube 660 490 solid state 440 g e ner a l 502 420 300 Television -General 439 Air Conditioner 978 860 1,389 SOURCES FOR TABLE D.33. (a) Midwest Research Institute, Patterns of Energy Use by Electrical Appliances, EPRI EA 682, January 1979, p. 5-3. (b) Edison Electric Institute. (c) Office of Science and Technology Executive Office of the President, Patterns of Energy Consumption in the United States. (d) Federal Energy Administration, Project Independence Blueprint Final Task Force Report, Residential and Commercial Energy Use Patterns 1970-1990. Under the Direction of Council on Environmental Quality, November 1974, Volume I, p. 98. (Presented in btu's and converted by author.) (e) Merchandizing Week as reported in Patterns of Energy Use by Electrical Appliances, EPRI EA 682, p. 5-23. (f) California Energy Commission, "Appendix A: Analysis of Residential Energy Uses," 300-76-034, 1976. (g) California Energy Commission, "California Energy Demand 1978-2000: A Preliminary Assessment," August 1979, Table D.5. These figures are for new appliances purchased in 1978.' f re ezers, as well as frost-free features, add considera bly to annual energy requirements. In contrast, solid state televisions use con- siderably less electricity than tube models. Third, the consumption of hot water, and thus of electricity by a wa ter heater, is related to the ownership and utilization of dishwashers and clothes •.;rashers. Because of these complications, it is not a straightfon.;rard matter t o apply national averages to Alaska. Furthermore, average annual consumption of electricity in various appliances will vary as a function of the following: 1. geographic location 2. size of household 3. income of household 4. age of appliance stock. Geographic location is obviously a factor iri room air conditioning consumption and appears also to be an important facet in refrigerator d f . 1 h h h . f . . 1. . 16 an reezer consumpt1on, a t oug t e 1n ormat1on lS pre 1m1nary. It is reasonable to assume that refrigerators need not work as hard in colder climates and will not be opened and closed as often. Water heating requirements may vary based upon the average annual inlet temperature of the cold water being heated . D-77 It is also reasonable to expect that use of electric ranges, clothes '"ashers and dryers, water heaters, and perhaps dishwashers would b; related to household size. There is some statistical information and h h . h h . 17 researc to support t ~s ypot es~s. The relationship appears most important for clothes washers where consumption might be estimated as a simple multiple of household size; while for the others, there is a significant "base load" independent of household size. Common sense and economic theory suggest that households with higher incomes will prefer appliances with features which may (frost- .free and large refrigerators) or may not (solid state television sets) be more energy using than the average. This idea is related to the age of the appliance stock which is a reflection of paftt purchase decisions based upon past income levels. As incomes grow over time, the existing stock of appliances will not accurately reflect present income levels of households but rate back present and past levels. There is no direct information available on Alaskan-specific elec- tricity consumption in various large appliances or how that consumption may differ from national norms. On the basis of climate, one might assume higher-than-average use for water heaters and ranges and lower- than-average use for air conditioners and refrigerators. On the basis of a slightly smaller average household size in Alaska, use of clothes washers and dryers, water heaters~ and dishwashers might be less than the average. On the basis of income and the appliance stock, Alaska could be less than or greater than the national average. Contributing D-78 a lower average for Alaska would be the fact ·that, historically, •Alaskan personal income per capita has been less than the national The higher-than-average per capita income in recent years in combination with rapid growth in the appliance stock could more than offset this, however. Since there is no clear direction to adjust the national data, the choice becomes essentially that of choosing among the various national series available. The choice makes significant difference only in the _cases of water heaters, ranges, refri-gerators, and dishwashers and clothes washers. We use the California Energy Commission's estimates of water heat consumption because they separate out consumption related to clothes and dish washing machines. The latter, we attribute directly to those appliances ba,.sed upon the probability of an mmer of an electric washer also owning an electric water heat·er. This could be an underestimate because of not counting hot water consumption for dish washing by those households without dishwashers. To adjust for this and the added heat- ing load caused by a colder-than-average water inlet temperature, we adjust the base figure upward by 15 percent·. We define the appliance entitled "range" broadly to encompass all heating for cooking purposes, including electric skillets, etc. This probably accounts for the discrepancies among the national estimates • ... D-79 Variation among estimates for refrigerator consumption is and is a function of the average size and type of refrigerator ass~med for the stock. The Midwest Research Institute estimate may be high because the sample chosen for metering was 97 percent single-family homes and duplexes.18 The other estimates fall within a much narrower band. Variation in freezer use is again a function of the stock although the various sources are in general agreement as to the average stock characteristics and electricity consumption. Anchorage merchants indi- cate most of their sales are in the manual defrost category, so we choose a lower-range estimate. Clothes dryer consumption may be nearer the high range of estimates in Alaska because of the cold climate. Air conditioner consumption will be considerably below the national average, and television consumption is a broad average of all estimates. The values used in the model are sum- marized in Table D.34. TABLE D.34. MODEL VALUES FOR MAJOR APPLIANCE AVERAGE CONSUMPTION FOR 1978. Water Heater Rc;tnge Clothes Dryer Clothes Washer \ Clothes Washer Hot Water Refrigerator Freezer Dishwasher Dishwasher Hot Water Television Air Conditioner D-80 Average kWh/Year 3,475 1,200 1,000 70 1,050 1,250 1,350 230 700 400 400 SMALL ELECTRIC APPLIANCE USE The average annual electricity requirements of several common, ·smaller electrical appliances are listed in Table D.35. Individually, the electricity used in each is not a large amount; but in the aggregate, it can be substantial for a household. Netting out the calculated quan- tities of electricity consumed in the residential sector for large appliances and space heating results in the amounts attributable to. small appliances in 1978 as shown in Table D.36. SOURCE: TABLE D.35. AVERAGE ANNuAL CONSUMPTION OF VARIOUS SMALL APPLIANCES (kWh) Trash Compactor 50 Iron 144 Electric Blanket 147 Humidifier 163 Hair Dryer 14 Clock 17 Sewing Machine 11 Vacuum Cleaner 46 Hi-fi (tube) 120 Hi-fi (solid state) 30 Radio (tube) 100 Radio (solid state) 10 Headbolt Heater 480 Garbage Disposal 36 Lighting 720 Edison Electric Institute, Canadian Wind Energy Program Papers by the National Research Council of-Canada, and Alaska Village Electrical Cooperative "Light Lines," May 1978. D-81 TABLE D.36. 1978 RESIDENTIAL ELECTRICITY CONSUMPTION ATTRIBUTABLE TO SMALL APPLIANCES· kWh/Customer Greater Anchorage Area ~ 2,010 Greater Fairbanks Area 2,466 Glennallen-Valdez Area 2,333 It is not possible at present to identify the specific small appli- ance mix in each region or to account for the interregional differences in small appliance consumption. We assume the Alaskan lighting load is considerably greater than the national average and arbitrarily set it at 1,000 kWh annually. The remainder is attributed to all other small appliances. D.7. COMMERCIAL-INDUSTRIAL REQUIREMENTS Commercial and industrial electricity requirements supplied by utilities are combined because of the small industrial base in the railbelt. Table D.37 presents the distribution of employment by category and indicates the relatively small size of manufacturing employment in the railbelt economy. Table D.38 further shows that the total number of manufacturing establishments is small (244), that they average 16 and that a large number are food and kindred products (at ·least 47) or D-82 t::l I 00 ... _______ ~ABLE D.37. 1978 WAGE AND SALARY EMPLOYMENT .. . . . . . .. ____ _ ·--·· ' --. . .... ··-.. -.. ······~·· .. , -·· . . .. ·····------"---·-·····--'----·-·---:·.,.. Government Civ. Mil.* Trade GREATER ANCHORAGE AREA Anchorage Kenai- Cook Inlet Matanuska- Susitna Seward 21,161 11,098 16,865 1,414 0 1,190 1,220 7 639 313 101 Finance- Services Ins.-R.E. 15,526 5,081 853 197 363 124 164 16 Trans.- Comm.- Utilities 7,924 574 307 ** Manuf. 1,683 989 Const. Mining 6,431 1,874 485 805 235 ** 12 ** Misc. 459 ** ** w GREATER FAIRBANKS AREA Fairbanks Southeast Fairbanks 7,218 570 GLENNALLEN-VALDEZ AREA Valdez-Chitina- Whittier 838 * Active duty 5,399 4,072 845 0 259 3,939 1,004 2,765 564 1,960 54 157 24 0 409 56 362 ** 89 ** )'cl'( Information withheld under regulations protecting confidentiality of data for individual firms. 86 ** SOURCE: State of Alaska Department of Commerce and Economic Development, Division of Economic Enterprise; "Numbers: Basic Economic Statistics of Alaska Census Divisions," 1979. Total 88,040 6,565 3,090 1,327 27,061 1,719 2,043 Food and Kindred Textile Mills Apparel and Other Textiles Lumber and Wood Furniture and Fixtures Paper Printing and Publishing Chemicals Petroleum and Coal Rubber and Misc. Plastics Leather and Products Stone, Clay, Glass Fabricated Metal Nonelectrical Machinery Electrical Equipment Transportation Equipment Instruments Miscellaneous Total Reporting Units Average Second Quarter 1979 Employment TABLE D.38. RAILBELT MANUFACTURING ACTIVITY Number of Reporting Units in 1979 Kenai-Matanuska-Southeast Valdez-Chitina- Anchorage Cook Inlet Susitna Seward Fairbanks Fairbanks Whittier 23 2 6 10 1 1 33 1 2 4 0 8 8 10 1 4 1 18 133 1,709 20 0 1 4 0 0 4 1 2 0 0 5 0 3 0 5 0 1 46 1,155 14 11 35 ,., 4 0 2 4 2 0 7 0 1 1 0 3 3 3 1 0 0 3 34 682 2 4 * 15 * Information withheld under regulation protecting confidentiality of data. SOURC~~ State of Alaska Department of Labor. Statistica1 Quarter1y, Second Quarter 1979. printing and publishing (at least 44). Petroleum processing and seafood processing are the most visible components of the manufacturing sector in terms of employees. Electricity and other energy end use in the commercial sector is most conveniently tabulated on the basis of the energy requirement per square foot of floor space. An accurate measure of this quantity--and its disaggregation into types of commercial floor space--is not available for any portion of the railbelt at the present time.19 The only fragment ~ of information available is an unpublished study conducted by the Muni- cipality of Anchorage Planning Department· which classified all nonresiden- tial floor space by use.20 An attempt was made, shown as Table D.39, to modify and update the survey results to make them compatible with the purposes of this study. The result_is an estimate of 37 million square feet of floor space in Anchorage in 1978. This figure is about 35 percent higher than would be calculated using Anchorage employment and national square feet per employee ratios. The difference may be attributable to a high vacancy rate in Anchorage office and retail space, a large pro- portion of newer construction, or the mix of employment within industries. This result is also considerably higher than a recent informal realtor 21 survey •. Estimates of floor space in other regions of the railbelt can be based either on national ratios of floor space per employee or on the 22 estimate developed for Anchorage~ The Anchorage ratio is preferred because it is based, however roughly, on Alaskan data. D-85 TABLE D.39. CALCULATION OF ANCHORAGE COMMERCIAL-INDUSTRIAL FLOOR SPACE AMATS Survey (Anchorage Bowl, 1975) Minus Non-energy Using (parking lots, cemetaries, etc.) Energy Using Floor Space 20 Percent Adjustment for Underreporting Sectors not Included in Survey: 1. Girdwood/Indiana b 2. Eagle River/Chugiak 3. Hotels/Motelsc 4. Assorted Cultural Buildingsd Retail Trade Warehousing Education Wholesale Trade Transport-Communication- Public Utilities Government Manufacturing Other 6,148 3,722 3,528 3,131 2,663 1,405 706 7",331 Growth Between 1975-1978e (approximately 25 percent) f 1978 Estimated Commercial-Industrial Floor Space General Education Warehousing Hotels Manufacturing See following page for table notes. D-86 25,120 5,000 4,520 1,500 860 42",067 18,918 23,149 4,630 27,779 53 300 1,000 500 29,632 7,400 37,000 TABLE D.39. (continued) (a) Twenty-five businesses in 1975 according to telephone book. Assume 2,500 square feet/business. (b) Based on the ratio of the housing stock in 1978 between Eagle River/Chugiak and Anchorage. (c) Assumes 2,000 rooms at 500 square feet/room. Based on Oak Ridge National Laboratory, "Commercial Energy Use: A Disaggregation by Fuel, Building Type, and End Use," Oak Ridge, Tennessee, p. 40. (d) Forty-six establishments identified in 1975 telephone book. Average size assumed to be 10,000 square feet. (e) This is based upon two indicators. The first is the growth in employment between 1974-75 and 1978. Civilian employment WqS as follows: 1974-58,700, 1975-69,650, and 1978-76,900. Employment growth was 31 percent in the period 1974 to 1978 and 10 percent in the period 1975 to i978. (State of Alaska, Department of Labor, Alaska Labor Force Estimates by Industry and Area, various issues.) ·The second is the growth in the appraised value of buildings over the period 1975 to 1978. After adjusting for inflation, the increase was 48 percent. Based on the assumption that the rapid employment increase in 1975 resulted in undersupply of floor space.in that year, we assume a 25 percent growth in floor space between the summer of 1975 and 1978. (f) Independent estimates of floor space in 1978 in the educational category and the hotel/motel category were available from the Anchorage School District and Anchorage Chamber of Commerce, respectively. The remaining growth was allocated proportionately among the other categories. D-87 On this basis, 1978 commercial-industrial floor space estimates for * the railbelt have been developed and are presented in Table D.40, using the Anchorage estimate of 480 square feet per employee. TABLE D.40. 1978 COMMERCIAL-INDUSTRIAL FLOOR SPACE ESTIMATES GREATER ANCHORAGE AREA Anchorage Kenai-Cook Inlet Matanuska~susitna Seward GREATER FAIRBANKS AREA Fairbanks Southeast Fairbanks GLENNALLEN-VALDEZ AREA Valdez-Chitina-Whittier Million Square Feet 42.3 37.0 3.2 1.5 .6 10.8 10.4 .4 1.0 Because different types of commercial establishments have different energy use characteristics, it is useful to categorize the commercial- industrial floor space by type of use. Unfortunately, it is not possible to do this accurately with the data that is presently available beyond the general categorization of the preceding Table D.39. Several studies at the national level and for other regions have broken the commercial sector into separated categories and have estimated the annual energy requirements for each on a square foot basis or a per- employee basis. Table D.41 reveals that no consistent pattern D-88 Offices Restaurants Retail Groceries Warehouse Elementary & Secondary Schools Health Hotel Misc. FEA (1974)b Offices Retail Schools Hospitals Other TABLE D.41. CATEGORIZATION OF THE COMMERCIAL SECTOR VARIOUS END-USE ANALYSES FEA (1977) c Retail -food -restaurants -office -other Wholesale -warehouse -non-ref. Finance-Insurance- Real Estate -office Offices Dispersed Retail CBD Retail Other State Bldgs. Health Care Education Local Gov't. Office PRNLf Retail-Hhlse. Educational Public Admin. Finance & Office Health Hotel Warehouse Religious Garage Other Forestry-Fishing- Construction Transportation & Public Utilities Wholesale Trade Retail Trade ·Finance-Insurance- Real Estate Services & Gov't. State Offices t::l -lodging -laundry I 00 \0 -recreation -automotive -outdoor activities SOURCE: (a) California Energy Commission, "California Energy Demand 1978-2000: A Preliminary Assessment," 1979. (b) Federal Energy Administration, "Residential and Commercial Energy Use Patterns, 1970-1990," 1974. (c) Federal Energy Administration, National, Energy Information Center, "~nergy Consumption in Commercial Industries by Census Division-1974," March 1977. (d) T. Owen Carroll et al. "The Planners Energy Workbook," Brookhaven National Laboratory and State University of New York Land Use and Energy Utilization Project. ? \0 0 TABLE D.41. (continued) (e) New York State Energy Office, "Development of a Commercial Sector Energy Use Model for New York State," 1979. (f) Jerry R. Jackson and WilliamS. Johnson, "Commercial Energy Use: A Disaggregation by Fuel, Building Type, and End Use," Oak Ridge National Labs., 1978. (g) NEPOL Load Forecasting Task Force and Battelle Columbus, "Report on Model for Long- Range Forecasting of Electric Energy and Demand to the New England Power Pool," 1977. in terms of the categorical breakdown utilized by different researchers either because of data restrictions, different objectives, or different perceptions and data availability concerning 0nd-use patterns in different sectors. Most researchers separate office, retail, warehousing, educational, and health-related buildings as having obviously different energy end-use characteristics. The most useful national source of information on energy consumption in the commercial sector for this study proved to be an analysis done by the Oak Ridge National Laboratory. Based upon their analysis of end use, Table D.42 was constructed, and it shows, for various components of the ·commercial sector, both proportion of energy consumed in various uses and the space heating load relative to office space. For Alaska, there is scattered information available on energy and electricity use on a square foot basis, essentially all of it from public buildings in Anchorage and Fairbanks. This information can be summarized as follows: 1. Based upon an inventory of 4.3 million square feet of space used by the Anchorage School District for the fiscal year 1978-1979, average electricity consumption (none for space heat) was 11.9 kWh/square foot/year. In addition, based upon 4 million square feet, the average natural gas consumption for space heat and related uses during the same period was .143 mcf/square foot/year. D-91 Commercial Sector Office Retail-Wholesale Garage & Service -Station Warehouse Education Hospital Religious Hotel Other TABLE D.42. ENERGY USE IN THE COMMERCIAL SECTOR NATIONAL PATTERN Percent of Fuel Used For Various Activitiesa Space Heat Water Heat Lighting Other 78 4 15 4 80 0 13 7 83 0 14 3 84 1 12 3 80 5 12 3 72 11 10 7 87 1 10 2 90 5 4 1 76 3 18 3 Space Heating Requirement per Square Foot Relati to Office 1 1.02 .45 -.45 .84 1.58 .67 1.58 .88 nationally weighted average 1.01 aDoes not include fuel used for air conditioning. SOURCE: Jerry R. Jackson a,nd William S. Johnson, "Commercial Energy Use: A Disaggregation by Fuel, Building Type, and End Use," Oak Ridge National Laboratory, 1978. D-92 2. Anchorage commercial real estate experts use 2-4¢/square foot/ a rule of thumb for electricity costs in the commercial sector. between·8 and 16 kWh/square foot/year if electricity 'averages 3¢ kWh. Gas consumption for space heating is also estimated to cost between .8 and 1.2¢/square foot/month which converts to between .062 and .093 mcf/square foot/year at a gas price of about $1.56 mcf. 3. Consumption of electricity on the Anchorage Community College campus in 1975 averaged 20.8 kWh/square foot/year; and for Kenai and .~· . v Matanuska-Susitna, it averaged 17.63 kWh/square foot/year. 4. Commercial consumption of electr~city in 1975 in the Fair- banks Area averaged 12.3 kWh/square foot/year for 552 thousand square feet of space surveyed. In addition, the 1.6 million square feet of space occupied by the University of Alaska averaged 17.4 kWh/square I 24 foot/year. Variation in use among consumers was dramatic, ranging from about 4 kWh to 50 kWh. 5. A 1978 inventory of Fairbanks schools.indicates·electricity consumption of 12.4 kWh/square foot/year based on 1.2 million square feet. In addition, space heating plus water heating requirements for the schools averaged 136 btu/square foot/year x 103 broken down by fuel type as follows: oil -103 thousand btu, coal -185 thousand btu, 25 steam -70 thousand btu. D-93 6. Another Fairbanks survey of public buildings done in 1978 -. indicates that the space heating requirement in two electrically heated buildings in Fairbanks is 302 thousand btu/square foot/year (88 kWh); and for four buildings heated by fuel oil, the average is 52 thousand.26 7. A sample of electricity and natural gas consumption of build- ings used by the Anchorage Borough indicates an average consumption of electricity of 21.8 kWh/square foot/year. The figure calculated for gas seemed unreasonably high and so is not reported. 8. The Anchorage International Airport reports electricity use at 51.8 kWh/square foot/year. No information is available on the incidence of the use of electrici as a fuel for space heating in the commercial-industrial sector, it is generally agreed that it is not significant. Gas supplies the majority of the load in the Anchorage area, supplemented by fuel oil; and fuel oil, coal, and steam supply the Fairbanks market. Fuel oil and propane serve Glennallen-Valdez. On the basis of this small amount of data, it is not possible to develop a plausible end use model of the commercial sector• It is, however, possible to substantiate the assertion that there is little electric space heating in the commercial sector in the Anchor-· age region and to develop a rough suggestion of electricity end use in I D-94 commercial sector. First, we calculate the use of natural gas in commercial-industrial category. The total of such consumption in 1978 was 6,917,786 mcf~ including apartments which are defined in this study to be in the residential sector. Netting out this component of 27 gas demand leaves 5,848,922 mcf. Using the school system consumption figures as a rule of thumb, this amount of consump- tion could provide space·heat to 40.9 million square feet of space, or 100 percent of the calculated space in the gas service area. Adjusting the·school district figure upward by the ratio of general · office-to-education building heat requirements taken from the Oak Ridge National Laboratory Report (1.5) yields 27.3 million square feet gas heated. (Natural gas use for process heat and air conditioning is very limited in the Anchorage market.) Thus, at most, somewhat between zero and one-third of commercial space heating requirements in the Anchorage area and within the gas utility service district could have electric space heat. Second, we confirm the predominance of gas fo~ space heating by allocating nonspace heat electricity consumption. Again using the Oak Ridge National Laboratory (ORNL) information on proportions of fuel us~d in various activities within a type of commercial building, we calculate the lighting and miscellaneous electricity requirement in various cate- gories of buildings relative to education. We then apply those ratios to the observed electricity consumption figures for electricity in Alaskan educational buildings. This results in Alaskan estimates of nonspace heating and cooling electricity consumption, which are sho'vn in Table D.43. D-95 TABLE D.43. ESTIMATES OF CO}illERCIAL ELECTRICITY CONSUMPTION Sector Education General Warehouse Hotel Manufacturingc Electricity Requirement Relative toa Education 1 1.5 .53 -~63 1.5 Calculated Electricity in kWh b Requiremerit/ft.2/year 13 19.5 7 8 19.5 aBased on Jerry Jackson and William Johnson, "Commercial Energy Use: A Disaggregation by Fuel, Building Type and End Use," Oak Ridge National Laboratory, 1978. bBased on 13 kWh/ft.2 /year assumed for the education sector (a rough-weighted average of the ayailable. information). c Assumed to have the same requirement as general commercial space. D-96 Using this data and the previously calculated information on square footage, it is possible to estimate total electricity use for nonspace in the commercial sector and compare it to the actual commercial sales~ This is done in Table D.44, which a.~so includes comparable cal- culations for other portions of the railbelt. By netting this use out of total sales, the amount remaining must be allocated to either space r Vheating, water heating, air conditioning, or process electricity. The 165,128 MWh calculated for Anchorage could, at 36 kWh annually per-square foot for space heat (based again on the Anchorage School· .District data converted from gas to electricity and adjusted to cover general commercial space using the ORNL data) heat about 4.6 million square feet of Greater Anchorage Area commercial space (about 11 percent). The comparable figures for Fairbanks and Glennallen would be about 12 and 23, respectively, if the electric heating requirement for Anchorage is converted on the basis of heating degree days in these other locations. The electric space heating load is lower than indicated by this procedure. There are some industrial and miscellaneous users of elec- tricity which consume much larger than average amounts of energy. The Anchorage International Airport, the petroleum-related facilities on the Kenai Peninsula, and the pipeline pump stations in Valdez are examples. D-97 TABLE D.44. CALCULATING THE POSSIBLE USE OF ELECTRICITY FOR SPACE HEATING IN THE COMMERCIAL- INDUSTRIAL-GOVERNMENT SECTOR Unit Electricity Total Actual 1978 Space Heat Floor Consumption Electricity Utility or Process Type of s3ace (non-space heat) Consumption Sales Electricity a Area Building (10 ft.2) kWh/ft.2 /year (mWh) (mWh) (mWh) GREATER ANCHORAGE AREA 721,620 886,748 165,128 Anchorage 618,270 67.7,021 58,751 education 5,000 13 65,000 general 25,120 19.5 489,840 hotel 1,500 7 10,500 warehouse 4,520 8 36,160 manufacturing 860 19.5 16,770 Kenai-Cook Inlet- ~ general 3,200 19.5 62,400 129,483 67,083 I 1.0 ,Matanuska-Susitna (.XI general 1,500 19.5 29,250 67,835 38,585 Seward general 600 19.5 11,700 12,409 709 GREATER FAIRBANKS AREA Fairbanks and Southeast Fairbanks . general 10,800 19.5 210,600 27i,726 61,126 GLENNALLEN-VALDEZ AREA Valdez-Chitina-Whittier general 1,000 19.5 19,500 28,604 9,104 ' a Difference between total utility sales and calculated non-space heat sales. ENDNOTES: APPENDIX D 1. Anchorage Real Estate Research Report Vol. III, Fall 1979, p. 44. Anchorage Real Estate Research Committee. 2. Ibid. , p. 55. 3. FMATS Study of Housing (draft), p. 16. 4. 1970 U.S. Census of Housing, Detailed Housing Characteristics: Alaska, Table 63. 5. In 1970, there were 2,757 year-round housing units unoccupied but not for sale or rent. Not all of these are vacation homes. In 1978, the Municipality estimated 729 residences in Girdwood~ of which 198 were full time. ~. Calculated from census data. 7. In 1970, Homer, Kenai, Seldovia, and Soldotna had 2,093 year-round housing units according to the census; and in 1978, 2,570 according to the Kenai Peninsula Borough, Profile of 5 Kenai Peninsula Towns. Applying this growth rate to the whole census division would yield only 5,710 un~ts in 1978. . . 8. Discussions with AMLP personnel. 9. Fairbanks North Star Borough, FMATS Housing Study (draf.t), 1980, p. 14. 10. U.S. Department of Commerce, Bureau of the Census, Privately Owned One-Family Homes Completed, 1977, Table 15. 11. Anchorage Real Estate Research Report, Ibid, p. 12. 12. Homer Electric Association, Power Requirements Study 1978, p. C.4. 13. CH 2M Hill, "Feasibility Assessment f9r Hydroelectric Development at Grant and Crescent Lakes," for Seward Electric Association 1979, . p. 10. 14. Fairbanks North Star Borough, "Community Information Quarterly," Spring 1980, p. 81. This survey may have substantial bias, but its direction is uncertain. 15. California Energy Commission, "Appendix A: Analysis of Residential Energy Uses," 1976, Table A.7.2. 16. Midwest Research Institute, "Patterns of Energy Use by Electrical Appliances," DPRI EA-682 project 576, January 1979, p. 5-3. D-99 ENDNOTES: APPENDIX D (continued) 17. Midwest Research Institute, Ibid., pp. 5-12; and California Energy Commission, Ibid. 18. Midwest Research Institute, Ibid., p. 3-2. 19. A partial tabulation is available for Fairbanks through the Municipality Planning Office in conjunction with the Fairbanks Metropolitan Area Transportation System (FMATS) Studies. 20. Anchorage Hetropolitan Area Transportation System (ANATS) Studies. 21. Alaska Center for Real Estate Education and Research and University of Alaska, Anchorage, Department of Real Estate •. 22. National estimates are available in Ide Associates Inc., "Estimating Land and Floor Area Implicit in Employment Projections," prepared for U.S. Department of Transportation, Bureau of Public Roads, 1970. 23. Based on studies conducted by Mark Fryer and Associates. 24. Ibid. 25. Ibid. 26. Fairbanks North Star Borough, Community Information Center, "Special Report 114," July 1979, p. 84. 27. Assume 90 percent of non-electrically heated units are gas (14,064). Converting the average heat load to gas yields about 76 mcf per unit or 1,068,864 mcf total apartment gas for space heating demand. D-100 APPENDIX E. ELECTRIC UTILITY SALES PROJECTION METHODOLOGY E.l. Residential Nonspace Heating Electricity Requirements The following appliances are identified in this model: 1. water heater 2. range (cooking) 3. clothes dryer 4. refrigerator 5. freezer 6. dishwasher 7. clothes washer 8. television 9. air conditioner 10. small appliances Time is measured in f1ve-year increments beginning in 1980, and there are three separate regions: Greater Anchorage, Greater Fairbanks, and Glennallen-Valdez. The electricity requirement for appliance type j at time t for region r (the r notation is dropped in all that follows) (REQ ) is j, t,r the product of five components. It begins with the number of households (HHt) multiplied.by the appliance saturation rate (sj,t). This yields the total n~~ber of appliances of type j (A. t). The portion of those J' appliances which use electricity is determined by the fuel mode split for electricity (FMS. t). The result is the number of electric appli-J,e, ances which is multiplied by average annuai consumption (KWH. ) to yield J, t . preliminary totalfconsumption. This is finally adjusted by an average household size adjustment (AHS. ) to yield total consumption. The ],t equa~ion is as follows: REQ. J. t HH * S * FMS . * KHH ;, AHS t j,t J,e,t j,t j,t (E.l) Total electricity requirement is the sum over all appliances j. L REQ. j -J, t (E. 2) The number of households (HH ) is determined by the demographic model. t The saturation rates (S. ) are exogenously entered. The average house-J,t hold size adjustment (AHS. t) applies to the following appliances only: J • clothes washer, water heater, clothes dryer. For these appliances, the adjustment is the ratio of average household size at t to average house- hold size in the base year, 1980. For other appliances, it is one. The fuel mode split (FMS. ) is the proportion of appliances of J,e,t type j which use electricity. FMS. J,e,t A. ],e,t A. t J. (E. 3) The number of appliances of type j in period t which use electricity is a function of the previous stock of such appliances (A. 1), the pro-J , e, t- portion of those appliances which are scrapped (d. 1 ), and the number J,e,t- of new appliances of type j purchased at time t which utilize electricity (NA. t). J,e, At this point, it is useful to distinguish different appliance "vintages," that is to identify the time period during which an appliance E-2 was manufactured (and sold). This is important because appliances of different vintages may have inherently different energy-use character- istics. This identification is accomplished by defining the appliance stock at any time t in terms of the initial stock and additions to the stock during each subsequent time period. Each vintage or appliances (m) may have a different scr~pping rate (d~ ) between time m and time t J,e,t because of different characteristics. The scrapping rate of the initial stock (d~980 ) will differ from that of subsequent vintages because the J,e,t initial stock is composed of appliances of different ages and, thus, vintages. For the existing appliance stock, the scrapping rate is also a function of the past grmv.th rate of the stock. Specifically, if gk is the historical growth rate of the stock and ex.k is the average lifetime, J then the scrapping rate in any year y can be approximated by: * (l+g.)y-1980 . J (E.4) This equation captures the phenomenon that the observed annual scrapping rate for a growing appliance stock will be less than (1/average lifetime) and that for a declining appliance stock the scrapping rate will be greater than .(1/average lifetime).a .,.aln practice, it was not possible to utilize this function. The scrapping rates for the existing appliance stock were based upon a cal- culation using the scrapping rate for new appliances extrapolated back- wards and weighted by the post-1970 growth rate of the stock. The stock of electrical appliances of type j at time t (A. ) is· J,e,t thus as follows: A. J,e,t A * (l-d:980) + NAJ.,e,l985 * (l-d:985) j,e,l980 J,e,t J,e,t + • • • + NA. 5 * (1-d :--5 ) + NA .. J,e,t-J,e,t J,e,t The number of additions to the appliance stock j at time t using electricity is the product of the total number of new appliances at time j (NA. ) and the incremental (or marginal) electrical mode split J,t (msi.' t) • This equation is simply, J,e, NA. = NA. t * msi. J,e,t J, J,e,t (E. The number of new appliances of type j at time t is the difference betlveen the stock demanded at time t, represented by the number of house- holds multiplied by the saturation rate, and the supply of those appli- ances which is the stock from the previous period net of scrapping. Thus, it is necessary to consider appliances using all types of fuels in calculating new appliance type j requirements at t. This demand can also be stated in terms of the initial appliance stock and all subsequent additions to the stock as follows: E-4 NA. =A. ],t ],t E A * (1-dl980 ) k j,k,l980 j,k,t E NA. k 1985 * (l-dl985 ) - k J, ' j, k, t -E NA. * (l-d:~5 ) k J,k,t-5 J,k,t (E. 7) The average annual electricity requirement in kilowatt hours of appliance type j at time t (KWH. t) is a function of the age distribution J' of the appliance stock and the electricity requirement for each vintage. Four factors are involved in the determination of the electricity re- ,quirement for each vintage. First, there is an average electricity requirement for the existing appliance stock (kwhj,1980). Second, there is an average electricity requirement for new additions to the stock in year t (kwh. t). Third, the average size of additions to the appliance J, stock j may change. This is accounted for by a growth rate on average appliance size subsequent to 1985 (l+kwhg.). Finally, mandatory improve- ] ments in appliance efficiency may reduce average electricity requirements for new vintages of appliances. These Federal conservation targets (cs. t) are implemented beginning in the interval 1980 to 1985. The J' average consumption of appliance j at time t can thus be expressed as the following: E-5 KWH. t J, kwhJ.,l9.80 * [A * (l-d:980 )]/A. + j,e,l980 J,e,t J,e,t 0 (l+kwhgj) * (l-csj,1985 ) * kwhj,1985 * [NAJ.,e,l985 * (l-d:985 )]/A. + .•• J,e,t J,e,t (l+kwhg.)t-1985 * (1-cs. t) *kwh. 1985 * ·J .J, J, [NA. ]/A. J,e,t J,e,t A major portion of electricity use in dish and clothes washer opera- tion is for water heating. We account for this by separately calculating· this component of electricity use in these appliances. The estimate of the use of electricity for water heating for these appliances is the product of the water heater saturation rate, the electric mode split, and the washer saturation rate. Small appliances are not differentiated. Thus, their electricity requirement is simply the product of the number of households, the ini consumption level (~~1980 ), and a growth factor [nkwh *. (t-1980)]. *KWH. t * [nkwh * (t-1980)] J, Electric light consumption is assumed to be a simple multiple of the current consumption level per household and the number of households. E-6 E.2. Residential Space Heating Electricity Requirements Total residential space heating electricity requirements in region r (not indicated in the algebraic notation for ease of exposition) at time t (SHREQ ) is composed of the requirements calculated separately e,t for four different types of structures. These are the following: 1. single family 2. duplex 3. multifamily 4. mobile home The electricity requirement for each type of structure j (SHREQ. ) is J,e,t based upon the required heat load measured in btu's 1TOTBTU. ), the J,e,t conversion efficiency of electric devices for producing space heat (EFF. t), and the conversion constant between btu and kHh. J,e, Algebraically, it is SHREQ. t J,e, TOTBTU. t/EFF. t/CONV J,e, J,e, e (E.lO) In the present use of the model, the conversion efficiency is assumed constant throughout (no utilization of heat pumps), so the energy require- ment can be calculated as kHh throughout. The space heating requirement for structu~e type j can be further defined as the product of the number of housing units of type j (HT.), J the proportion utilizing electric space heat (FMS. ), the average J,e,t level of consumption (KHH. t), and the utilization rate (UT. t) • . J, J,e, E-7 SHREQ. = HT. * FMS . * KWH. * UT . J,e,t J,t J,e,t J,t J,e,t The number of units of housing type j in year twill be a proportion. (HMSj,t) of the first home housing stock in year t (FHUt). The proportion of housing units of type j at year t using electric space heat will be simply FMS . = HT. t/HT. J,e,t J,e, J,t The number of housing units of type j using electric space heat at time t will be a function of the initial stock of such installations (HTj,e,1980) plus additions to the stock over time (NHTj,e,t) and net of scrapping from the stock over time (d~ . ) . The r.ate of scrapping J,e,t is vintage specific for the reasons mentioned in the appliance stock discussion. Thus, the electric space heat units for structure type j at time t can be written as follows: + . . . + NHT. 5 * (1-d~-5 ) + NHT. J,e,t-J,e,t J,e,t E-8 (E. New electric heating units for structures of type j at time t are determined by the electric incremental mode split (msi. t) applied to J 'e, all purchasers of space heating appliances for housing units of type j at timet (NHT. ), J,t NHT. = NHT. t * msi J,e,t J,. j,e,t New space heating requirements for structures of type j at· time t (E.l5) are determined by the difference between the total number of housing units type j required at time t (HT. ) and the existing stock w~th space ],t heating appliances. We assume the scrapping rate for the housing stock is zero but that there is a scrapping·rate for space heating appliances using different fuel types k. Thus, new space heating appliance demand in housing units type j at time t is as follows: NHT. = HT. ],t J,t r HT. k 19.80 * (l-d:980 ) -1: NHT. k 1985 * (l-d:985 ) k J, , J,k,t k J, , J,k,t t-5 1: NHT. k t-5 * (1-dJ.,k,t) k J, , Finally, the electricity requirement per unit (KWH. ) will be ],t the weighted average of .the per unit requirement of space heating appliances of all different vintages of the stock. The average per unit electricity requirement for any vintage will be the product of (E.l6) the base year requirement (kwhj,1985 ), the growth rate in basic unit consumption (b.rhg.), the mandated energy savings for the unit (1-cs. t), J J, E-9 and the energy savings as a result of system retrofitting (1-ret~ ) J ,e, t in the interval between m and t. As with appliances·, a difference between average existing and new unit energy consumption is recognized so that there is a separate and different unit consumption parameter for the existing stock (kwh. 1980). The average consumption calculation is . J , thus as follows: KWH = (1 tl980 ) * k h * [HT * j,t -re j,e,t w j,l980 j,e,l980 (l-d~980 )]/HT. + (l-ret~985 ) * J,e,t J,e,t J,e,t * kwhj,l985 * (l+kwhgj)o * [NHTj,e,l985 * (l-d~985 ) ] /HT. + . . . + (1-cs. ) * kwhJ. , 1985 * J,e,t J,e,t J,t (l+kwhg.)t-l985 * NHT. /HT. J J,e,t J,e,t (E .17) E-10 E.3. Commercial-Industrial Electricity Requirements Commercial-industrial-government electricity requirements are aggre- _gated into a single category because of data limitations. Total require- ments at time t in region r (dropped from equation for ease of exposition) for these combined sectors (CIREQt) is the product of nonagricultural wage and salary employment (EMt) and average consumption per employee (CIKWHt). (E.l8) Average consumption per employee varies as a function of time and the implementation of conservation standards. Specifically, new or incremental electricity users (EMt-EMt_5 ), who represent new commercial- industrial-government floor space, will have different electricity require- ments than existing customers. Thus, existing customers at the beginning of the projection will have an average consumption rate (kwhl980) different from incremental customers added in subsequent five-year intervals. A different consumption rate is assigned to incremental customers (kwh 1985 ) and this consumption rate grows over time linearly at a fixed amount (nkwh). Efficiency standards at timet (l-est) are effective on incre- ments to electricit~ requirements but not to the total. The general equation for the commercial-industrial-government load can then be written as follows: E-ll * (l-csl985) + (EM1990 -EM1985) * [kwhl985 + nkwh] (E .19) E.4 . Miscellaneous Electric Utility Sales Miscellaneous sales consist of two very small components of total sales: street lighting and recreational homes. Street lighting sales (SLREQt) is a fixed percentage (sl) ·of the total of residential and commercial-industrial-government sales. sl * [REQ + SHREQ + CIREQ ] t t t (E. 20) Similarly, recreational home consumption (RECREQ ) is calculated as t a fixed level of consumption (kwh) multiplied by a fixed proportion of households (rec). RECREQ t HHt * rec * kwh E-12 (E. 21) APPENDIX F. ELECTRIC UTILITY SALES PROJECTION MODEL PARAMETERS F.l. Projection Model Parameters for Base Case F.l.A. RESIDENTIAL NON-SPACE HEATING ELECTRICITY REQUIREMENTS Parameters used in this model are presented in Tables F.l and F.2. F.l.A.l. Conservation The Energy Po·licy and Conservation Act of 1975 and the National Energy Conservation Policy Act of 1978 direct the Federal Energy Admin- istration (Department of Energy) to promulgate energy efficiency targets for electrical appliances sold beginning in 1980. The targets are based upon an aggregate 20 percent increase in the efficiency of energy use for 13 appliances using 1972 as a base year. The targets for the various appliances differ because each is determined on the basis of economic and technical feasibility. For example, the efficiency improving changes in refrigerator design include improved compressor motor efficiency, improved insula- tion, improvement of door seals, provision of an on/off switch for anti- ' sweat heaters and elimination of the condenser fan motor. These changes were estimated to increase the retail price of the refrigerator by half the cost of electricity saved in the first year of operation.1 The efficiency targets are presented in Table F.3. F.l. MODEL PARAMETERS: RESIDENTIAL NON-SPACE HEAT APPLIANCES Parameter Region Greater Greater Glennallen- Anchorage Area Fairbanks Area Valdez Area Saturation Rates (S. ) J,t SWH 1980 .99 .97 .91 1985 ·1.00 .99 .94 1990+ 1.00 1.00 1.00 sc 1980+ 1.00 1.00 1.00 SCD 1980 .71 .67 .49 1985 .72 .69 .52 1990 .73 .71 .54 1995 .74 .72 .56 2000 .75 .73 .58 2005 • 76 .74 .60 2010 .77 .75 .62 SR 1980+ 1.00 1.00 1.00 SF 1980 .46 .45 .46 1985 .48 .48 .49 1990 .51 .51 .52 1995 .52 .53 .54 2000 .55 .55 .56 2005 .57 .57 .58 2010 .58 .59 .60 SDW 1980 .49 .38 .15 1985 .54 .44 .24 1990 .59 .50 .32 1995 .63 .55 .39 2000 .67 .60 .45 2005 .71 .64 .51 2010 .74 .68 .• 56 KEY: WH Water Heater DWW Dish\vasher \-later c Cooking cw Clothes Washer CD Clothes Dryer cww Clothes Washer Water R Refrigerator TV Television F Freezer AC Air Conditioner DW Dishwasher F-2 (Continued) ' Region Greater Greater Glennallen- Anchorage Area Fairbanks Area Valdez Area Saturation Rates (S. t) J' sew 1980 .77 .75 .66 1985 .78 .76 .68 1990 .79 .77 .70 1995 .80 .78 .72 2000 .81 .79 T' • J 2005 .82 .80 .74 2010 .83 .81 .75 STV 1980 1.50 1.51 .85 1985 1.55 1.56 1.00 1990 1.60 1.60 1.10 1995 1.64 1.64 1.19 2000 1.68 1.68 1.27 2005 1.71 1.71 1.34 2010 1.74 1.74 1.41 SAC 1980+ 0 .01 0 Incremental Electrical Appliance Hode Split (msi. ) J ,e, t ms~ 1980+ .35 .5 .4 msic 1980+ .66 .85 .4 msiCD 1980+ .90 .98 .75 msi (other) 1980+ 1.0 1.0 1.0 F-3 F.l. (Continued) Parameter Greater Anchorage Area Average Annual Appliance Consumption in 1980 (kwhj,1980) WH c CD R F DW DWW cw cww TV AC Average Annual New Appliance Consumption (kwhj,1985 ) WH c CD R F DW DWW cw cww TV AC F-4 Region Greater Fairbanks Area 3,475 1,.200 1,000 1,250 1,350 230 700 70 1,050 400 400 3,650 1,250 1,000 1,560 1,550 230 740 70 1,050 400 400 -(Continued) Greater · Anchorage Area Conservation Target for New Appliances (csj,1985) WH c CD R F DW cw TV AC Grmvth in Appliance Size (kwhg.) J WH c CD R F DW cw TV AC F-5 Region Greater Fairbanks Area .14 .03 .06 .29 .21 .18 .29 .32 .21 .005 0 0 .01 .01 .005 0 0 0 Glennallen- Valdez Area F.l. (Continued) Parameter Average Lifetime of Appliance (ex.) J WH c CD R F DW CT.V TV AC Historical Electric AEEliance Stock Growth Rates (gj) WH c CD R F DW cw TV AC Greater Anchorage Area .05 .05 .06 .05 .05 .09 .06 .07 0 F-6 Region Greater Fairbanks Area 10 10 15 15 20 10 10 10 10 10 .03 .03 .04 .03 .03 .04 .04 .04 .03 Glennallen- Valdez Area .15 .10 .14 .10 .11 .25 .12 .20 0 TABLE F.2. MODEL PARAMETERS: SMALL RESIDENTIAL APPLIANCES Parameter Region Greater Greater Glennallen- Anchorage Area Fairbanks Area Valdez Area Average Annual Con- sum:etion Level (KWh 1980) electric lights 1,000 1,000 1,000 assorted appliances 1,010 1,466 1,333 Annual Increment to Small A:epliance Con- sum:etion (nK\.fu) 50 70 70 F-7 TABLE F.3. FEDERALLY-MANDATED ELECTRICAL APPLIANCE EFFICIENCY Electrical Appliance 1. 2. 3. ·4. 5. 6. 7. 8. 9. 10. Refrigerator Freezer Dishwasher Clothes Dryer Water Heater Air Conditioner Television (black & white) Television (color) Range Clothes Washer Percent Reduction in Energy Consumption of New Appliance Using 1972 as a Base 32 23 20 7 15 23 65 35 3 32 SOURCE: Federal Energy Administration, Energy Conservation Program for Appliances. Federal Register, Vol. 42, No. 136. F-8 Independent studies of the potential for conservation of electricity in appliance design conclude that much greater savings in energy is pos- sible at modest cost. For example, a Danish study in 1979 concluded that a 64 percent savings in el~ctricity consumption could be-obtained from refrigerator design changes with a payback time of five years. The design included increased insulation thickness, elimination of automatic defroster, and increased condenser efficiency.2 More radical design changes could further reduce the electricity consumption of refrigerators and other appliances. This study assumes that the Federal guidelines will be implemented during the period 1981 to 1985.3 The target efficiencies are reduced by 10 percent to take account of the fact that the base consumption levels from which the targets are calculated are those of 1972; and in the interim betw·een 1972 and the present, some efficiency improvements have already appeared in new appliances. For example, the primary design improvement for television sets is a conversion to solid-state circuitry. This is already happening and, consequently, the applica- tion of the target value to the 1980 stock would overestimate the 1 . 4 actua energy sav1ngs. The rationale for assuming implementation of the Federal guidelines but no additional changes in appliance efficiency is based on the idea that these are improvements which will, in fact, occur, while further improvements will require addition-specific programs at the F-9 Federal or state level~ Should they occur, they can be subsequently introduced into the model. F.l.A.2. Appliance Lifetime Appliance lifetime estimates are available from the Home Appliance Manufacturers Association. Relevant appliance lifetimes are presented · in Table F.4. TABLE F.4. APPLIANCE LIFETIMES Appliance Lifetime in Years Freezer 20 Refrigerator 15 Electric Clothes Dryer 14 Electric Range 12 Television (color) 12 Dishwasher 11 Clothes Washer 11 Television (black & white) 11 Room Air Conditioner 10 Electric Water Heater 10 SOURCE: Richard B. Comerford, "PSE&G Method of Forecasting Kilowatthour Consumption," and George L. Fitzpatrick, "Peak Forecasting Methodology," in How Electric Utilities Forecast: ERPI Symposium Proceedings, March 1979. Other estimates of appliance lifetimes indicate some variation these figures, although it is not substantial.5 F-10 In this analysis, since appliances of different vintages have differ- ent energy-use characteristics,_it is important to identify not only the - average lifetime but also the probability distribution of lifetimes for specific appliances. One recent study which has investigated the dur- ability probability distribution for air conditioners concluded that it is a Weibull distribution which has a mean of approximately ten years with 50 percent of the population wearing out in the interval between . 6 4.5 years and 13.75 years. We assume the same probability distributions for the durability of other electrical appliances and use a simplified distribution to dis- tribute appliance· disposals around the mean lifetime. Since the model calculates appliance stocks on a five-year interval, appliance disposals are assumed to occur in the five-year interval preceeding and the ten- year interval succeeding the average lifetime which is adjusted to be a multiple of five years. The typical distribution is shown in Figure F.l. FIGURE F.l. PROBABILITY DISTRIBUTION FOR TYPICAL APPLIANCE DURABILITY % Scrapped 25% Average Lifetime -5 Years 40% Average Lifetime Years F-11 25% Average Lifetime +5 Years 10% Average Lifetime +10 Years F.l.A.3. Appliance Capacity Growth Rates The capacity of the existing appliance stock and consequently the average consumption level for those appliances is discussed in Appendix D, Section 5. The capacity of many major appliances has grown over the years. This contributes to higher electricity consumption. For example, it has been estimated that the capacity in cubic feet of the average refrig- erator sold in the United States has increased from 8.5 in 1950 to 12.0 in 1960 and to 14.0 in 1970.7 Analysis of past Alaskan electricity consumption patterns indicates that this phenomenon is occurring here and presumably will continue to occur in the future.8 This growth is a function of both increasing incomes and technological change. The latter factor is obvious in the example of refrigerators because in 1950 there were no refrigerators sold with a capacity of 15 cubic feet, while by 1970 the average was that size.9 The following appliances are assumed to maintain a constant capacity over time in terms of energy consumption: 1. Range 2. Clothes washer 3. Clothes dryer 4. Television sets 5. Room air conditioners F-12 There is no direct information available on changes in the average capacity of appliances in use in Alaska, and the information available nationally is not necessarily applicable to Alaska because of differences in existing stock as well as preferences. Some general assumptions may· be made, however, for those appliances which may be subject to capacity change. Dishwashers may include two features which affect electricity con- sumption for each load. These are the "pots and pans" feature which adds about 1 kWh per load and a "no-heat dry" feature which reduces consumption 10 by about .4 kWh per load. There is no current information on the utilization of these features, although the Department of Energy appli- ance efficiency guidelines include the "no-heat dry" feature as one of the components of their efficiency target. It is assumed here that the "pots and pans" feature becomes more commonplace over time on new dishwashers but is used relatively infrequently so that the average capacity growth rate is .5 percent annually for new dishwasher purchases. The average new dishwasher is assumed to use 1.05 times the electricity of the existing stock based roughly on this growth rate and our estimate of average appliance life. Refrigerator·volumes have been increasing over time as indicated above, and with increasing volumes has come increasing energy consumption. The inclusion of a freezer and a "no-frost" feature also add to average consumption rates. Elimination of these features is not suggested in the Department of Energy appliance efficiency standard targets. F-13 Nationally, the percentage of refrigerator sales by type for selected years is as follows: TABLE F.S. DISTRIBUTION OF DOMESTIC REFRIGERATOR SALES FOR SELECTED YEARsll (percentage) Single-Door Two-Door Freezer Manual Partial Non-Frost Side-By-Side 1950 100 0 0 0 1963 39 28 33 0 1966 23 20 so 7 1973 13 18 49 20 Using annual refrigerator sales volumes, the growth rate in the size of the average volume of refrigerators sold was 3.6 percent in the . 12 1950s, 2.3 percent in the 1960s, and 1 percent in the early 1970s. From this information, it is apparent that the trend nationally is for some moderation in the growth rate of refrigerator volumes with a simultaneous shift toward refrigerators with more energy-using features. Alaska is assumed to be experiencing the same trends with a resultant increase in the energy requirements of new refrigerators. The Californi~ Energy Commission's end-use model assumes that the average electricity consumption of new refrigerators is 1,609 k\Vh annually in 1974; while for the existing stock, it is 1,235 k\Vh annually, a ratio F-14 of 130 percent. It also presents information indicating that the growth rate in annual energy consumption of new refrigerators was 3.7 percent 13 between 1960 and 1974 and 2.2 percent between 1970· and 1974. On this basis and the assumption that the average age of the appliance stock is less in Alaska than nationally, it is assumed that new refrigerators operate at 1.25 times the consumption rate of the existing stock and that the growth rate in consumption of the new stock is 1 percent annually. Freezers vary in size, shape, and whether they have a "no-frost" feature. The "no-frost" feature adds about 50 percent to average 1 . . . 14 e ectr~c~ty consumpt~on. As with refrigerators, elimination of this optional feature is not a stated component of the appliance efficiency standards targets of the Department of Energy. There is no historical series nationally on the proportion of sales of freezers which are of the "no-frost" type and no data in Alaska on the characteristics of the existing stock. Because of this lack of data, the most reasonable assumption would be to assume the same growth characteristics for freezers as for refrig- erators; that is, the new freezers will be more likely to have a larger capacity and the 11 frost-free" feature than the existing stock. However, it is assumed that the standard deviation around the mean of the size distri- bution of existing freezers is smaller than for refrigerators so that the ratio of the average consumption of the new-to-existing stock will be smaller than in the case of refrigerators. It is assumed to be 1.15. F-15 Water heater size has increased over time, as indicated by manufac- turer shipments of domestic gas water heaters,15 which grew in capacity by about 1 percent annually during the mid-1970s. Since energy require- ments will not grow as fast as volume and because average per capita residential hot water consumption is relatively constant except for use in dishwashers and clothes dryers, continued growth in water heater energy use should be small. However, because the existing stock may .continue to be replaced by larger units, a small positive growth rate is assumed for energy use of additions to the stock. The average con- sumption of new units is .assumed to be 1.05 times the existing stock based upon this assumption and on estimate of the age of the existing stock. Small appliances. Electricity consumption from the use of small appliances will increase as additional appliances are purchased by the average household. Individually, such appliances do not constitute a large proportion of total demand, but the combined consumption of electricity through such appliances could continue to increase as new appliances, some not even now on the market, become available and are purchased. It is difficult to specify a growth rate for electricity consump- tion through these appliances because there is no information available on existing saturation rates and use patterns in Alaska and because of the following considerations: F-16 1. Use of electricity in some new appliances may substitute for electricity use in existing appliances. For example, increased use of microwave ovens may partially reduce electricity use for conventional ranges. 2. Use patterns for smaller appliances may ch~nge significantly in the future. For example, a more dispersed population would result in greater use of electricity in water pumps to bring well water to the surface. 3. It is not possible to anticipate all future uses of electricity in the home. Humidifiers, large-screen televisions, and trash compactors are examples of recent additions to small appliances in use in the residential sector. An annual increase of 5 percent of the 1978 consumption level for small appliances is assumed for future growth. The base figure used in this calculation varies between the regions because of different climate, preferences, and other unidentified factors. These differences are assumed to persist. The average household use of electricity for lighting is assumed to remain constant over time. F.l.A.4. Appliance Saturation Rates Deviation of appliance saturation rates is discussed in Appendix D, Section 3. F-17 F.l.A.5. Incremental Mode Split To calculate incremental mode splits for water heaters, ranges, and clothes dryers, we rely upon the same sources of information used in the development of the 1978 end use inventory. We begin by comparing the average mode split reported in the 1970 Census (Table D.29) to the incremental mode split calculated for the period 1960 to 1970 (Table D.32). When the average and incremental mode splits thus calculated are approximately equal, the market for the appliance is in equilibrium with respect to the fuel types used. Unfortunately, it was not gen- erally the case that such an equilibrium could be identified. Table F.6 shows the incremental mode splits used in the model. They remain constant throughout the projection period on the assumptions : '~ that the relative prices and availabilities of the various fuels will not change and that preferences for various fuels does. not change. The Anchorage splits have been relatively constant historically with only electric water heaters losing ground to gas. The census~ calculated incremental mode splits are utilized. The census-reported information for the Matanuska-Susitna Borough is not useful because of the rapid subsequent growth. there which has relied heavily on electricity. We calculate the water heater and range incremental mode splits on the assumption that the price advan- tage. enjoyed by electricity over fuel oil will persist and the majoritY of purchasers will choose electricity. Electric ranges will be slightlYf' F-18 TABLE F.6. INCRE~ffiNTAL ELECTRICAL APPLIANCE MODE SPLIT Water Heater Range Clothes Dryer Refrigerator Anchorage ~30a .67a .98a 1.00 Matanuska- .75d .sod .96e Susitna 1.00 Kenai- .soa .4ob .90c Cook Inlet 1.00 Seward .35e .75£ .70e 1.00 Fairbanks .sog .85g .98g 1.00 Glennallen-.40h .40h .soe 1.00 Valdez aCensus calculated incremental mode split·. bEA . H survey est~mate. cAssumes a shift away from gas toward the pattern observed for Anchorage. dBased on price advantage of electricity. e 1970 Census. £Assumes shift toward electricity. gBased on growth since 1970. hBased on shift toward electric range preference. F-19 more preferred than electric water heaters. The electric clothes dryer mode split is taken from the 1970 Census. In Kenai-Cook Inlet, we utilize the incremental mode split calcu- lated from the census for water heaters reflecting a shift in preference toward electricity. The electric range split is taken from an end use survey conducted by Homer Electric Association in 1977 because the census figures appeared low based upon the general pattern of growth since 1970. The clothes dryer mode split presumes a shift away from natural gas toward electricity in a reflection of Anchorage preferences. For Seward, the 1970 Census data was used to calculate water heater and clothes dryer mode splits while a shift toward a preference for electricity for cooking was assumed on the basis of cleanliness and con- venience. For Fairbanks, the 1970 water heater electric mode split was 37 percent and the end use inventory calculated a 43 percent split in 19~8. The incremental split over the interval would thus be about 50 percent. We assume this for future projections, although the recent electricity price increases might result in a shift in preference back to fuel oil in the future. Electricity is preferred for cooking in Fairbanks based upon the census-calculated incremental mode split which,shows a substan- tial electric range retrofit between 1960 and 1970. We assume a continua tion of the trend toward electric ranges with an 85 percent incremental mode split. For clothes dryers, we use the 1970 Census information. F-20 The Glennallen-Valdez census data is out-dated because of the rapid post-1970 growth, but subsequent information is not cl,lrrently available. We assume a shift toward electrical appliances occurs in reflection of trends observed elsewhere in the railbelt. The clothes drying mode split is based on the 1970 Census. F.l.A.6. Household Size Adjustment Factor Clothes washers, clothes dryers, and water heaters are used more intensively by larger households. A study conducted by the Midwest Research Institute calculated average annual electricity requirements for these appliances as a linear function of household size using d 1 . 16 metere app ~ances. These equations are converted to use in this model by: 1. annualizing them (they are based on daily consumption); 2. normalizing them by a 1980 average household size of three persons; and · 3. calculating a ratio by which to adjust calculated consump- tion to account for changing household size. The adjustment factor is a function of the ratio of average household size in year t to 1980 (AHHt) and is formed from the equations of Table F.7. F.l.B. RESIDENTIAL SPACE HEATING ELECTRICITY REQUIREMENTS Parameters used in the residential space heating model are pre- sented in Table F.8. F-21 TABLE F.7. EQUATIONS TO DETERMINE HOUSEHOLD SIZE ELECTRICITY CONSUMPTION ADJUSTMENTS Appliance Clothes washer Clothes washer water Clothes dryer Water heater Equation 1 * AHH .25 + .75 * AHH • 25 + . 7 5 * AHH .51 + .49 * AHH F-22 F.8. MODEL PARAMETERS: RESIDENTIAL SPACE HEATING Parameter Greater Anchorage Area .·· Average Annual Unit Consum:etion (Existing Units) (kwh. J' 1980) SF DP MF :MH Average Annual Unit Consumption . (New Units) (kwhj ,1985) SF DP MF :MH Growth in Unit Size (kwhg.) J SF DP MF. :MH Average Unit lifetime (ex.) J SF DP MF MH KEY: SF Single Family DP Duplex MF Multifamily :MH Mobile Home 36,500 24,200 . 17,100 27,300 40,100 26,600 18,800 30,000 F-23 Region Greater Fairbanks Area 48,200 31,900 21,200 36,900 53,000 35,100. 23,300 40,600 .01 .01 .01 .01 20 20 20 20 Glennallen- Valdez Area 33,300 21,900 14,600 25,400 36,600 24,100 16,100 "27,900 F.8. (Continued) Parameter Greater Anchorage Area Incremental Electrical Applinnce Mode Split (msi. ) -J,e,t msiSF,l980+ ms~P,l980+ ms~,l980+ ms\m,l980+ Conservation Target for New Appliances (csj,1985 ) SF DP MF MH Utilization Rates (UT. ) J ,e, t UTSF,l980+ UTDP,l980+ UTMF,l980+ UTMH,l980+ .19 .19 .19 .19 F-24 Region Greater Fairbanks Area .01 0 0 .01 .05 .05 .05 0 1 1 1 1 F.8. (Continued) Parameter Region Greater Greater Glennallen- Anchorage Area Fairbanks Area Valdez Area Retrofitting Coefficients m (ret. t) J ,e, . 1980 retSF,l985 .02 .04 .03 1980 retDP,l985 0 0 0 1980 retMF,1985 0 0 0 1980. retMH,l985 0 0 0 F-25 F.l.B.l. Conservation Conservation enters into the determination of the residential space heating load through assumptions about retrofitting of existing units with energy saving improvements and the application of efficiency standards to new housing units. Retrofitting existing structures to reduce the required heat load will be a function of the quality of the housing stock~ the expected length of housing unit ownership, the amount of information available to individuals interested in retrofitting, and the cost of fuel saved compared to the investment in supplies and labor required to do the retrofit. Several federal and some current state programs are designed to stimulate retrofitt~ng in the residential sector. Among the important federal programs are a tax credit of 15 percent of the first $2,000 expended for conservation measures, a home improvement loan program for energy conservation measures, and a weatherization grant program f 1 . f .1. 17 or ow ~ncome am~ ~es. The most important state program is a 10 percent residential fuel conservation tax credit for capital improvements to reduce the heat load of residential buildings. Several studies have attempted to assess the impact of retrofitting on energy requirements for space heating. In 1974 a study by Arthur D. Little estimated that nationally- applied retrofitting measures based upon current reasonable technology F-26 and cost could reduce the electric space heating load by 26 percent, 20 percent, and 17 percent for single family, multi-family, and mobile h . . 1 18 orne type un1ts, respect1ve y. A 1977 study estimated at 20 percent savings in energy consumption from retrofitting 20 million single family units· bet"t:veen 1977 and 1990.19 . Unfortunately, these estimates are not based upon actual,observed human behavior but, rather, are based upon simple engineering models. A study reported by the California Energy Commission indicates that the actual response to retrofit conservation programs and actual energy savings may be only about 50 percent of what engineering models would 20 suggest. The only information currently available concerning retrofitting of the housing stock is available from state tax returns for 1977 and 1978. The number of returns, percentage claiming credits, average credit claimed, and implied value of capital investment in retrofitting equipment are shown in Table F.9. Since this data is not regionally divided and specific fuel used is not specified, it is not possible to accurately estimate the impact of this program on electric space heat consumption. If we assume an even distribution statewide, an even distribution among all types of.fuels, and a 5-year payback for investments (with no discounting), then in 1978 about $975 thousand was spent for railbelt F-27 TABLE F.9. STATE OF ALASKA RESIDENTIAL FUEL CONSERVATION CREDITS 1977 tax returns 195,394 percent claiming credit 5.6% number claiming credit 10,942 average credit $74.10 ; implied capital investment (@ 15% credit) $741 implied total capital investment $8,108,070 1978 183,725. 8.1% 14,882 $57.61 $576 $8,571,873 Source: Alaska Department of Revenue, "Fuel, Conservation, and Industrial Credits Relative to the Individual Income Tax," 1980. F-28 electric space heat unit retrofits. This saved $195,000 in electric bills. If the average cost of electricty was 5¢, then about 3,900 ~Vh of' electricty were conserved by the retrofit program, or less than 1. percent of residential electric space heat requirements. This is ob- viously only· on order of magnitude estimate, but it suggests that the impact of the existing state retrofit program on aggregate consumption of electricity is probably modest. The impact could quite possibly be much smaller with a longer payback period or if a smaller percentage of credits were taken by electric space heaters than assumed. A further problem with using national estimates of the potential savings from retrofitting is that the thermal integrity of the typical Alaskan house may be much better than the national average. It is clearly a younger than average stock, so few homes would be without in- sulati9n as in the lower 48. The harsh winters would suggest more con- cern during construction for thermal integrity, but this may not have been the case in fact. On the basis of this spotty information, it is not·possible to assume a substantial impact on electric space heating of the existing federal and state retrofitting programs. Obviously, some of the impact has already occurred, and to project an estimate of the full impact of these programs into the future would involve some double counting of conservation. F-29 We assume that retrofitting is confined to single family residences and occurs on the existing housing stock during the period 1980 to 1985. It is twice as important in Fairbanks as in Anchorage because the higher price of electricity in Fairbanks creates an extra incentive there. The impact on Glennallen-Valdez falls midway between the 4 percent saving for Fairbanks and 2 percent saving for Anchorage. The application of mandatory construction or performance standards to new housing in order to improve their thermal integrity has been . under consideration for several years by the federal government. The sets of standards which may be implemented are either the American Society of Heating, Refrigerating, and Air Conditioning Engineers 90-75 standards or standards developed by the U.S. Department of Hous~ng and Urban Development (HUD). These standards now are supposed to become effective in 1980. National studies have estimated the impact of these mandatory stan- dards on energy consumption. The Arthur D. Little study estimated po- tential savings of 35 percent, 45 percent, and 40 percent in mobile homes single family units, and multi-family units, respectively. A 1977 study estimated savings of 11 percent for single family units and 46 ~ercent for multi-family units under the ASHRAE 90-75 standards and 20 percent and 51 percent savings under the HUD standards.22 Substantial savings are thus apparently possible, but there are no precise estimates of what the savings would be from standards. F-30 An attempt has been made to estimate the impact of the two afore- mentioned standards on Alaskan energy consumption, but the conclusions of the study were qualitative rather than quantitative and suggested 1 h b . 1 . ld b .bl 23 on y t at su stant~a sav~ngs wou e poss1 e. We assume-that a program of mandatory standards is implemented in 1981 which has the effect of reducing the h~at load in new construction (except for mobile homes) b~ 5 percent independent of other factors. This percentage takes into account the assumption that Alaskan housing is already more thermally efficient than the national avetage, the fact that actual savings observed will be less than savings in theory, and the idea that it will take some time to actually implement the program. No conservation is assumed for mobile home units. F.l.B.2. Heating Appliance Lifetimes For ease of calculation, the demolition rate for the existing housing stock is set at zero. This is not significantly different from actual ratios as indicated from building permit information. The assumption ~lso applies to the heat distribution system for the home. The heat generating system (boiler, furnace, etc.) is assumed to have an average lifetime of 20 years, independent of type or fuel used. This is based on Home Appliance Manufacturers Association data, and, as with other appliances, the actual time of scrapping of an appliance is determined by a probability_distribution centered at the average lifetime. F-31 F.l.B.3. Average Housing Unit Size and Consumption The average housing unit size was estimated in Appendix D, Section 2, and electricity consumption for space heating was assumed to be a function of the average unit square feet of floor space. (An adjustment factor was calculated to account for the fact that 1978 was a warmer than normal winter. See Table D.24.) We assume that the average electricity requirement for new units constructed after 1980 (independent of conservation) is 10 percent greater than existing units because new construction is assumed to con- sist of larger units on average than the existing stock of housing Two sources of information confirm the observation of an average size of the housing stock. In Table F.lO, the average tion of natural gas per heating degree day is presented The consumption growth in the 1970s of between 2 and 3 percent annually can be attributed to growth in the average size of stock (or to deterioration in the thermal integrity of the housing stock). Second, as noted in Appendix D, Section 2, the average size new single family units nationally is larger than the average for the existing stock by about 10 percent (1720/1570). We assume new housing of all types constructed after 1980 will be on average 10 percent larger than the existing stock based upon this national ratio. F-32 Year 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 TABLE F.lO. RESIDENTIAL CONSUMER NATURAL GAS CONSUMPTION FOR SPACE HEATING PER HEATING DEGREE DAY Average Annual Consumption/ Consumption for Heating Heating Degree Space Heating (mcf) Degree Days Day 172 10,137 .0170 180 11,879 .0152 191 12,016 • .0159 188 11,665 .0161 170 10,683 .0159 193 11,308 .0171 181 10,361 .0175 166 9,394 .0177 164 9,131 .0180 159 9,430 .0169 Source: Alaska Gas and Service Company annual financial reports and internal company records. F-33 In subsequent years, new additions to the housing stock are larger by 1 percent annually. This is a balance of several factors which can be identified but not quantified. Increasing disposable incomes will increase the demand for larger housing units, but the increasing cost of hou.sing will partially offset this. The role of the federal and state governments in stimulating home ownership through various programs could increase the size of new additions to the housing stock or reduce it, depending on the particulars of the program. The declining average household size in future years should reduce the demand for larger units somewhat. The Alaskan climate which requires that people spend a large proportion of their time indoors during the winter months sug- gests a strong preference for larger housing units. The Arthur D. Little study of 1975 assumed that these various factors would cancel one another out so that the size of new housing was projected to remain constant in future years and that only replace- ment of demolitions would increase the average size of the stock by 4 percent between 1975 and 1990. We assume the disposable the climate-related preference, and the presence of s_tate intervention into the housing market predominate and r.esult in an increasing size for increments to the housing stock. F.l.B.4. Incremental Mode Split* We assume that the Greater Anchorage Area space heat mode split is in equilibrium. Thus, the incremental mode split will be equal to *See also Section F.2 for a discussion of the space heating mode choice decision. F-34 the average mode split for electricity. The Greater Anchorage Area average of 19 percent electric residential space heat~ng is assumed for all housing types. This is a slight decline from the existing multi- family stock of 19.9 percent (see Table D.21) but a slight increase for the other type of units. This assumption presumes no shift in the· geographic distribtuion of new housing units either toward or away from areas where natural gas is available or the extension of natural gas service into new areas. Growth in the mid-1970s in these outlying areas has been relatively rapid, but it is not clear whether this is a tem- porary phenomenon or represents the emergence of a long-term trend. Growth has decelerated in the last year, but that could be a reflection of the gene~al softening of the Alaskan economy. In Fairbanks_, Golden Valley Electric Association (GVEA) has put a ban on new electric_space heat hookups. This is assumed to be permanent in the absence of new generating facilities powered by fuels other then fuel oil because of the high incremental cost of power from this source. Nevertheless, we assume that 1 percent of new and replacement single family and mobile home units are heated by electricity. This represents a gradual decline in the electric space heat load occurring over a period of about 20 years as existing units wear out and are replaced. GVEA in their own load growth estimates assumes that all of their resi- dential space heat customers will be shifted off of electrici~y by 1982. F-35 There is very little electricity used for space heating in the Glennallen-Valdez service area because of its relative price. ·We assume the same incremental mode split for electric space heating as the present average. F.l.B.5. Utilization Rates We assume utilization rates are unchanged from current levels. That is, people do not manually set back their thermostats at night, etc. F.l.C. CO~rnERCIAL-INDUSTRIAL-GOVERNMENT ELECTRICITY REQUIREMENTS Model parameters for the component are shown in Table F.ll. F.l.C.l. Conservation Conservation measures in the industrial sector consist of effi- ciency standards for new appliances, construction or performance stan- dards for new construction, and retrofitting of existing structures to increase the thermal efficiency and reduce the electricity load in the various building systems such as the heating, ventilating, and air con- ditioning systems (HVAC). Because detailed end use information is not currently available, it is not possible to identify in detail the impact on electricity consumption of specific conservation measures. Because federal conservation programs are and will be in effect, however, it is necessary to try to quantify their impact. The major conservation programs of the federal government specifi- cally targeted to the commercial, industrial, and government sectors F-36 TABLE F.ll. MODEL PARAMETERS: COMMERCIAL-INDUSTRIAL-GOVERNMENT SALES Parameter Region Greater Greater GlennalLen- Anchorage Area Fairbanks Area Valdez Average Consumption Rate for Existing Stock (KWhl980) 10.675 10.983 9.178 Average Consumption Rate for Increments to Stock (KI.fu 1985 ) 15.156 18.537 12.979. '!•. fi·. ·~-;.-;. Subseguent Increases to - -~~.:-_ Incremental ConsumEtion ~[;· Rate (nKvlh) · 3.020 3.707 2.596 :{. '>:; ·f ~(; Design and Performance Efficiency Targets (est) 1985 0 0 0 1990 .05. • 05 .• 05 1995+ .1 .1 .1 F-37 include grants to schools and hospitals for improving energy efficiency, a local public building energy audit program, conservation requirements for federal buildings, energy efficiency labeling of industrial equip-· ment, stimulation of cogeneration, business energy tax credits, and per.;.. formance standards for new commercial buildings similar to the residen 24 sector. National studies have attempted to measure the potential impact of these federal programs. A 1979 analysis of the National Energy Plan estimated the growth rate of energy use in the commercial sector could be reduced from 4 to 3.2 percent annually as a result of.implementation 25 of the plan. The Arthur D. Little study previously mentioned estimated potential energy conservation factors for several types of commercial buildings. These factors, shown in Table F.l2, are the proportion of energy savings possible using "practical methods and existing materials in addition to allowing for some technological improvements in selected HVAC and electrical components between now and 1990."26 These calculations are based upon a technical analysis of possi- bilities, but the study also includes a discussion of the institutional setting within which energy conservation in the commerciai sector would be addressed and provides some insight into the problems which imple- mentation of energy conservation would entail. Specifically, the relative complexity of the typical commercial structure makes it F-38 TABLE F.l2. ENERGY CONSERVATION FACTORS FOR COMMERCIAL BUILDINGS (1970 = 1.0) Existing Buildings Office Buildings Lighting .80 Auxilliary equipment .95 Space heating .78 Cooling .82 Hot water heating .95 Retail Establishments Lighting .70 Auxilliary equipment .95 Space heating .76 Cooling .77 Hot water heating .95 Schools, Educational Lighting .. .80 Auxilliary equipment .95 Space heating .79 Cooling .81 Hot water heating .95 New Construction .50 .90 .60 .53 .90 .50 .90 .50 .54 .90 .50 .90 .50 .59 .90 Source: Arthur D. Little, "Residential· and Commercial Energy Use Patterns: 1970 to 1990," for Federal Energy Administration, 197'•, p. 156. F-39 difficult to calculate actual energy consumption of the various systems in the building and to determine potential savings from design changes. (For example, the lighting system waste heat provides some space heat.) In addition, the design of a building normally involves the attempt to meet a large number of objectives, only-one of which would be energy efficiency. The implementation of this objective requires the close interaction of client, architect, and engineers designing the various building systems. It is-clear from the discussion in the Arthur D. Little report that energy conservation was not a major concern in design and maintenance in the early 1970s. This was reflected in the fact that architects consulted were sensitive to conservation issues but lacked "the detailed know·ledge to apply conservation measures with d f h . . . ,.27 any egree o sop ~st~cat~on. The heterogeneity of this component of electricity consumption is a further problem, making it difficult to analyze electricity use potential savings. Finally, consumption is dramatically effected by building use patterns. The same building, from a design standpoint, can have energy and electricity consumption differences of over 100 d d . h t f of the bu-lld-lng.28 perce~t, epen ~ng upon t e pat erns o . use ~ ~ We assume that the majority of electricity used in the commercial~ industrial-government sector is for lighting, in conjunction with space conditioning systems, and for auxilliary electrical equipment. The new construction conservation potential in these areas is significant, but we assume that the impact of currently-planned federal programs, F-40 including design or performace standards, will be more modest and will take considerable time in implementation due to institutional constraints to development of the standards and immediate implementation when they become available. We assume a 5 percent reduction in electricity re- quirements for new construction during the period 1985 to 1990 (indepen- dent of other growth factors), i~creased to 10 percent in the following decade. This suggests a higher conservation potential in this sector than the.residential space heat sector but a longer time for implementa- tion. At the same time, because of the absence, particularly in the Greater Anchorage Area, of a strong price incentive, retrofitting measures in the commercial sector are assumed to have no impact on. electricity consump- tion. (See next section concerning assumptions regarding growth of consumption by existing customers.) In other words, the _new construction standards program is the measure which results in conservation in this sector. F.l.C.2. Commercial-Industrial-Government Utility Sales Per Customer ·Historical annu~l utility sales per customer data for the major railbelt regions and the U.S. as a whole are compared in Table F.l3. The average Alaskan customer consumes about 70 percent more electricity in a year than those in the U.S. as a whole, and over the long run, the growth rate in average sales has been realtively equal for-Alaska and the U.S. In the period of the 1970s, the Alaskan growth rates have been more rapid, but this has been offset by apparently slower growth rates in the late 1960s. F-41 1950 1955 1960 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 TABLE F.l3 C0~1ERCIAL/INDUSTRIAL UTILITY SALES PER CUSTOMER (MWh) Greater Greater Anchorage Area Fairbanks Area· 47.2 41.2 52.0 40.5 55.2 57.7 53.0 61.7 57.8 62.7 62.8 69.1 69.0 73.7 67.9 80.6 73.4 79.8 72.1 85.1 90.0 87.6 86.6 85.8 84.8 86.5 85.0 aEdison Electric Institute, Statistical Year Book, annual. F-42 . a U.S .A. 9.3 12.7 17.0 27.4 40.0 41.9 44.5 47.6 46.6 49.0 50.7 52.9 It would be premature to identify the period of the mid-1970s as a per customer peak for commercial sales in Alaska, but there is a noticeable deceleration of growth rates in more recent years. This could partially be the result of more rapid than normal growth during the Alyeska oil pipeline boom years or a succession of abnormally warm winters (in Anchorage) in the late 1970s or both. These are temporary phenomena which should not form the basis for analysis of underlying trends. Examination of the year-to-year growth rates of commercial sales nationally reveals a ve~y rapid growth rate in consumption historically ·and also the possibility of a new long-term trend after the watershed years of 1973-1974 (the time of the great .recession and oil embargo). The average annual growth rate in the 1970s before the embargo was 6 percent, while afterwards it was 4.3 percent. Again, it is premature to emphatically call this a shift in long-term trend, but ft is sug- gestive of that. ) In order to facilitate analysis of various conservation programs and trends in new commercial structures, we calculate sales estimates for the existing stock of commercial-industrial-government buildings and for increments to the stock. For ease of calculation, we assume a zero rate of demolition of the stock. For the existing stock, the average consumption rates in 1978 are util~zed in the projections. This assumes, therefore, that all growth in consumption is the result of additional hookups at higher consumption rates and that commercial- F-43 industrial-government consumption is not sensitive to heating degree days. (1978 was a warmer than normal year.) The annual consumption per hookup for incremental customers is assumed to grow at a rate consistent with the period between 1973 and 1978. Specifically, the following equation was solved in Greater Anchorage and Greater Fairbanks to obtain the average sales per cus- tamer of customers added to the systems after 1973. average sales per customer for customers added after = 1973 total 1978 sales -total 1973 sales customers added after 1973 This value was then compared to the 1973 average to arrive at an esti- mate of the growth in average sales to incremental customers. These results are summarized in Table F.l4. Greater Greater TABLE F.l4. CALCULATIONS OF ELECTRICITY SALES TO INCREMENTAL COMMERCIAL-INDUSTRIAL-GOVERNMENT CUSTOMERS Average Sales Average Incremental 1973 Sales (1973-1978) Ratio (MWh) (MWh) Anchorage 80.557 97.903 1.22 Fairbanks 73.429 114.806 1.56 Thus, assuming that existing customer sales remained constant, new tamers between 1973 and 1978 on average purchased 22 percent more electricity inAnchorage and 56 percent more electricity in Fairbanks. F-44 Based upon the longer trend for the two areas from TAble F.l3 and the slower growth for Fairbanks in the early 1970s, we assume the same growth characteristics for the two regions and also for Glennallen- Valdez. Specifically, Table F.l5 shows the steps in calculating average and incremental rates of consumption. Consumption figures have been converted to a per employee basis (discussed in the next section). 1978 actual average consumption figues are converted to 1980 estimates on the basis of growth rates calculated from recent trends on average sales growth rates. This is multiplied by the ratio of incremental to average sales to obtain a base for calculating incremental consumption rates for the period 1981 to 1985. These rates are 25 percent above the 1980 base incremental rates. In subsequent five-year periods, the same amount is added to incremental sales per employee. F.l.C.3. Commercial-Industrial-Government Hookups The number of electric utility hookups is related to employment. Table F.l6 indicates the stability of that relationship historically except during the pipeline years of 1975 through 1977. Because of this stability between the level of employment and the number of hookups, it is possible to utilize sales per employee as the forecasting variable in this section of the model and link it directly to the employment forecast generated by the economic model. F-45 I I'Zj I ~ "' TABLE F.l5. COMMERCIAL-INDUSTRIAL-GOVERNMENT CONSUMPTION PER EMPLOYEE DATA Rate of 1978 1978-1980. 1980 19n-1978 average sales growth in average sales incremental to /emElo:lee c /emEloyee 1978 av. sales sales/emEl• (MWh) (%) (MWh) Greater Anchorage 10.071 6% 10.675 1.13 Greater Fairbanks 10.768 2% 10.983 1.35 Glennallen-Valdez 10.085a -9% 9.178 1.13b a 8,000 MWh of pipeline pump station sales netted out. b Assume the same relationship as Anchorage. cBased on recent trends in average sales growth. Av. sales/ Succeeding 5 employee for yr. increments 1981-1985 to sa.les to incremental incremental customers customers (MWh) (Mivh) 15.156 3.020 18.537 3.707 12.979 2.596 ' TABLE F .16.. RATIO OF NON-AGRICULTURAL HAGE AND EMPLOYMENT (NET OF MILITARY) TO COMMERCIAL-INDUSTRIAL-GOVERNMENT HOOKUPS Greater Anchorage Greater Fairbanks Glennallen- Area Area Valdez 1965 8.55 8.73 5.37 1966 8.63 8.02 1967 8.79 1968 8.34 8.43 1969 8.50 8.80 1970 7.79 7.61 3.58 1971 8.37 7.59 4.37 1972 8.46 7.78 1973 8.54 7.54 2.66 1974 9.12 8.81 4.32 1975 10.12 13.14 11.18 1976 9.45 12.15 17.69 1977 8.94 9.06 8.55 1978 8.59 7.90 4.12 Average 8. 72 8.89 6.87 Average net of 1975-1977 8.52 8.12 4.07 F-47 F.l.D. MISCELLANEOUS ELECTRICITY REQUIREMENTS This very small category consists of street light and second home sales. Street light sales are assumed to be 1 percent of the sum of all other components of sales. The electricity requirements for second homes is difficult to identify for several reasons. First, as discuss.::-d in Appendix D, Section 1, it is difficult to identify from the existing housing stock studies just what is a second home or a vacation home. Specifically, what the census defines as a year-round housing unit may actually be a second or vacation home. It is possible from the census to determine the number of households within a census division which own a second home, but not its location. Most utilities do not have separate rate schedules for second homes, and if they did, the utility definition would not necessarily be the appropriate one since it might cover seasonal units rather than units used year-round but infrequently. It is also difficult to estimate average electricity requirements for second homes because of this lack of data. We make the following very rough estimates to calculate second home electric utility sales: 1. 25 percent of households have second homes, based on census information; I 2. 50 percent of the second homes are located within the railbelt; F-48 3. 50 percent of the second homes in the railbelt are serviced by electric utilities. The average annual consumption per second home is 500 ~~, based upon conversations with utility personnel. F.2. Assumptions for the Price Induced Shift Toward Electricity Consumption Case F.2.A. FACTORS INVOLVED IN APPLIANCE CHOICE FOR SPACE HEATING The most important variable in a model of appliance choice for space heating or any other function is the system cost, including both the initial purchase price and lifetime fuel costs for system operation. Other characteristics are important and will affect the choice but will not be explicitly considered here. Some of these other considerations are as follows: Heating System 1. Heat distribution within the building 2. Amount of space occupied by the heating system 3. Multiple controls capability 4. Integration with other appliances (hot water, humidifier, air conditioner) 5. Comfort factor (annoyance of hot air, for example) 6. Reliability 7. Compatability with auxilliary heating systems Fuel 8. Fuel availability 9. Haintenance cost 10. Safety of fuel 11. Cleanliness 12. Convenience F-49 Because of these considerations, not all households will make the same appliance type and fuel choice even when faced with identical prices. In considering the least-cost space heating system, it is necessary to recognize that for the new residence there are at least two decisions after having decided in favor of a centralized heating system. The first is the type of heat distribution system and the second is the type of fuel to produce the heat. The choice of distribution system--hydronic (hot water or steam), hot air, electric resistance, etc.--affects not only the initial cost but also the cost of a subsequent retrofit to ,._n alternative system. Having once chosen a heating distribution system, the fuel to provide the heat is determined, basically on the basis of . 29 prJ.ce. - For a given set of desired heating characteristics, the distribution system and fuel type chosen will, in theory, be that which minimizes the present value of total future system costs. In practice, several factors tend to distort the decision in favor of the system with the least initial cost or the least cost over a shorter period than the system lifetime. 1. Homebuilders who build for others are concerned with minimizing construction costs and will opt for the system with the least initial cost if a related higher total operating cost is not reflected in a reduced market price for the house. 2. The same analysis applies for landlords to the extent that they are able to pass system operating costs on to tenants. F-50 3. For individuals who own their homes for only a short time before moving, the lifetime operating costs of the heating equipment in those homes is less important than initial system cost.30 4. Lack of information about the least-cost system may prevent people from switching to it. 5. Individuals may not act consistently with the actual opportunity cost of money. In other words, a system choice with a high initial cost may have a rate of return in terms of money saved (compared to the next best alternative which has a lower initial cost but a higher operating cost) which exceeds the return the purchaser could receive investing the same amount of money alternatively. Yet, for some reason, the con- sumer chooses the system with the lower initial cost but higher lifetime cost. In other words, the ob~ served discount rate used by the consumer exceeds his opportunity·cost.31 6. Promotional activities of utilities. In general, electric baseboard heating is the cheapest system to install followed by hydronic and then warm air systems. Both hydronic and warm air systems require a flue and a distribution system. The initial cost of oil relative to natural gas depends upon the cost of connecting the residence to the gas main compared to the cost of oil storage tanks and the somewhat higher cost of an oil burner. This may vary with location. Based on this discussion, it can be seen that the proportion of a particular type of heating system in place at any time may exceed what would appear to be economically justified based upon total system lifetime cost. F-51 Table F.l7 shows the prices of various fuels which equalize the operating cost of space heating. For example, if ele c tricity is 3~/kWh, then heating with fuel oil will be less expensive if it is available for under $.84/gallon. Table F.l8 shows the actual relative prices presently encountered in various parts of the railbelt. It is clear that natural gas is the cheapest fuel wherever it is available. Outside of the gas utility service area in the Anchorage region, the prices of fuel oil and electricity are comparable after adjusting for the higher conversion efficiency of electricity. This is because electricity is produced using cheap natural gas. In the Fairbanks area, fuel oil and electricity are comparable for one utility, while the price of electricity from the other exceeds that of fuel oil. This situation results from the fact that the lower cost electricity is produced by coal, while the higher cost electricity is · generated by a combination of fuel oil and coal. In Glennallen-Valdez, fuel oil is clearly cheaper than electricity. This pattern is confirmed by looking at historical data on relative fuel prices. In Anchorage, natural gas and electricity prices are largely determined by the long-term purchase contracts for Cook Inlet gas. The pricing clauses in these contracts have resulted in fairly stable prices during the 1970s. In contrast, fuel oil prices have risen with rising crude oil prices. F-52 $/106 btu 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 TABLE F.l7. EQUIVALENT DELIVERED PRICE FOR SPACE HEATING USING VARIOUS FUELS Electricity Fuel Oil Crude Oil ¢/kWh $/gallon $/barrel .5 .14 5.88 1 .28 11.77 1.5 .42 17.64 2 .56 23.54 2.5 .70 29.40 3 .84 35.28 3.5 .98 41.16 4 1.12 47.07 4~5 1.26 52.92 5 1.40 58.80 5.5 1.54 64.68 6 1.68 70.56 6.5 1.82 76.44 7 1.96 82.32 7.5 2.10 88.20 Notes: 1) Furnace conversion efficiencies: electric .95 gas and oil .65 2) BTU content of fuels: 1 kWh = 3,413 btu 1 gallon fuel oil = 138,000 btu 1 mcf gas = 1,005,000 btu 1 barrel crude oil = 5,800,000 btu 3) No r.efinery margin netted out of crude oil price F-53 Natural Gas $/me£ 1.53 3.06 4.59 6.12 7.65 9.18 10.71 12.24 13.77 15.30 16.83 18.36 19.89 21.42 22.95 TABLE F .18. PRICES OF ALTERNATIVE FUELS FOR RESIDENTIAL SPACE HEATING Natural Gas a Fuel Oilb Electricityc $/mcf $10 6btu $/gallon $10 6btu ¢/kWh $10 6btu GREATER ANCHORAGE AREA Anchorage 1980 .98 7.10 1979 2.98/d. 8.73/d 2.49/ 7.30/ 4.52 13.25 1978 1.89 1.88 Kenai-Cook Inlet 1980 .94 6.80 1979 3.52l 10.32/ 3.75/ 10.99/ 4.23 12.40 1978 2.01e 2.00 :Hatanuska-Susitna 1980 unavailable 1.07 7.75 1979 unavailable 4.52 13.25 GREATER FAIRBANKS AREA Fairbanks 1980 unavailable .84 6.09 1979 unavailable 3.50/g 10.26/ 7.97 23.37 GLENNALLEN-VALDEZ AREA 1980 unavailable .98 7.10 1979 unavailable 12.1/h 35.47/ 13.82 40.52 See following page for table notes and sources. F-54 Table F .18 (Continued) Table Notes: (a) 10 mcf monthly bill (b) 500 gallons delivered (c) 1,500 kWh monthly bill (d) CEA/AMLP/MEA (e) ANG (f) HEA/SES/HEA (Kenai) (g) FHUS/GVEA (h) CVEA (Valdez)/CVEA (Glennallen) SOURCES: Electricity: State of Alaska, Division of Energy and Power Development, "1980 Alaska Power Development Plan." Gas: Alaska Public Utilities Commission, Annual Report, 1978. Fuel Oil: Survey by authors and Fairbanks North Star Borough, Community Information Quarterly, Vol. III, No. 1, Spring 1980. F-55 In Fairbanks, fuel oil and electricity generated by fuel oil have ·. been most susceptible to pric.e increases in the 1970s. The price of electricity generated by coal has been less susceptible to increases in the 1970s. In Glennallen-Valdez, electricity is produced by fuel oil and, thus, the two prices move together. This review suggests that a substantial change in the existing fuel mode split for space heating would require a large change in the rela-· tive price of electricity. Specifically, for electricity to become the least expensive space heating fuel, the following price changes would be necessary; Anchorage -the relative price of natural gas would need to increase at least 3 times; Fairbanks -the relative price of fuel oil would need to increase at least two and one-half times; and Glennallen-Valdez -the relative price of fuel oil would need to increase at least three and one-half times. On the other hand, it is possible that a large increase in the price of all fuel will result in a shift away from the "conventional" fuels-oil, gas, and electricity--toward more conservation or auxilliary systems such as efficient fireplaces and wood stoves. This phenomenon may be beginning to occur ,.in the outlying parts of the Greater Anchorage area and in Fairbanks. F-56 This introduces the second group selecting a space heating fuel-- the retrofit market. This consists of households whose existing system has worn out as well as those whose systems are still functiorting but because of changed operating costs decide that a system replacement is cost effective. Fuel retrofits are relatively ·common when the switch is between oil and gas in hot air or hydronic systems. For example, a large portion of the Anchorage residential market has been retrofit from oil to gas. This required only switching the burner and connecting the unit to the gas main. Switches to an electric resistance furnace could similarly be relatively inexpensively accomplished. System retrofits in which one type of heating system is replaced with another are far less common, and the feasibility of such a switch will depend upon the construction of the building. For example, in a house built on a concrete slab, it would be virtually impossible to retrofit a hydronic or hot air system because there would be nowhere for the placement of th~ pipes or ductwork. In replacing an electric resistance system with a hydronic or hot air system, it is necessary both to locate a place for the furnace and to install a flue. Generally, the cheapest system retrofit is to electric resistance heating since the installation of the required wiring is less com- plicated than that of pipes or ductwork. F-57 Table F.l9 summarizes the conversion costs (prices as of the early 1970s, although the relative costs are unchanged) of v arious fuel and system retrofits. The table c.onfirms the discussion that retrofits to electric baseboard are relatively inexpensive, but retrofits from elec- tricity to oil or gas conventional systems are relatively expensive. Basically, it is easier to switch into the electric mode than out of it. Obviously, for the household contemplating a retrofit, the cost of the switch plus the present discounted cost of the fuel in the new system is the relevant variable against which the cost of using the current system must be compared. The higher the cost of retrofit, the higher the relative price of the existing fuel used for space heating can become and still be less costly than switching. Table F.20 shows what the actual railbelt heating s y stems have been historically. The pattern has been one of shifting from heaters to central space heating systems over time (which consume more energy). Hydronic systems were more common in Anchorage, Fairbanks, and Glennallen- Valdez in 1970, while hot air systems were more common elsewhere. This suggests that shifts to electric heat could involve either resistance systems or electric furnaces using the existing heat distribution systems (or possibly heat pumps). There is a historical example of a retrofitting phenomenon in the railbelt. In the early 1960s, natural gas from the Cook Inlet fields became available to the Kenai Peninsula and Anchorage. The natural gas F-58 'j1 V1 1.0 CONVERSION COSTS OF RESIDENTIAL HEATING SYSTEMS (dollars) ~ Gas Electricity Fuel System From Warm Air Hot Water Baseboard Warm Air Gas Warm Air X 3,300 1,500 1,500 Hot Water 2,600 X 1,500 3,100 Electricity Baseboard 2,600 3,300 X 2,700 Warm Air 1,000 3,300 1,100 X Warni Air 400 3,300 1,500 1,500 Oil Hot Water 2,600 400 1,500 3,100 SOURCE: Arthur D. Little, "Project Independence: Residential and Conunercial Energy Use Patterns 1970-1990," for Federal Energy Administration, 1974, p. 175. Oil Warm Air Hot Water 65o· 3,400 2,700 650 2,700 3,400 1,100 3,400 X 3,400 2,700 X TABLE F.20. HOME HEATING EQUIPMENT PROPORTION OF TOTAL Built-in Floor, Hall, Steam or Warm Air Electric or Pipeless Heaters, Hot Water Furnaces Units Furnaces FireElaces None GREATER ANCHORAGE AREA Anchorage 1960 46 20 0 8 26 0 1970 56 26 5 3 9 0 Matanuska-Susitna 1960 12 13 0 7 67 1 1970 16 30 1 1 53 0 Kenai-Cook Inlet 1960 9 14 0 2 72 4 1970 . 25 32 4 5 33 1 Seward >tj I 1960 18 12 0 14 56 1 ~ 0 1970 27 28 0 0 44 1 GREATER FAIRBANKS AREA Fairbanks 1960 50 25 0 3 22 1 1970 59 22 7 1 11 0 GLENNALLEN-VALDEZ AREA Va1dez-Chitina-Hhittier 1960 34 11 0 3 51 1 1970 28 20 1 1 50 1 WESTERN REGION U.S. 1970 5 40 9 22 22 2 Note: Based on all year-round housing units. SOURCES: 1960 Census of Housing, General Housing Characteristics: Alaska. Table 29. 1970 Census of Housing, Det~iled Housing Characteristics: Alaska. Table 62. 1970 Census of Housing, Detailed Housing Characteristics: United States Summarv. was cheaper than fuel oil; but in response to competition, the fuel oil distributors lowered their price. This, in combination with the fact that residents of short tenure would not recover their capital costs . of conversion from an oil to gas boiler in spite of the relatively low cost, kept the rate of conversions low. The limiting factor does not appear to have been the speed at which gas mains could be laid to the various neighborhoods.32 Over a period of 15-to-20 years, as indicated by Table F.21, there was a substantial shift toward natural gas space heating as a result of both new units choosing gas and retrofitting of gas burners where fuel oil had previously been used. The retrofitting of gas burners continues today but is asymptotically approaching zero. Unfortunately, because of a lack of intercensal housing stock data for the 1960s for Anchorage, it is neither possible to trace the exact time pattern of the retro- fitting of gas burners nor possible to correlat·e the rate of retrofitting with·the relative prices of fuels or other variables in an attempt to develop a model for predicting possible future responses to relative fuel price changes. F. 2. B. ASSUMPTIONS FOR A HIGH ELECTRIC SPACE HEAT SCENARIO Given the present structure of relative fuel prices in the rail- belt, the ·electric portion of the residential space heating load should remain relatively constant over time. This is predicated on the follow- ing assumptions: F-61 Year 1960 1965 1970 1975 1978 TABLE F.21. GROWTH OF USE OF NATURAL GAS FOR ANCHORAGE RESIDENTIAL SPACE HEATING Dwelling Units (not including Residential multifamily) Gas Customers 16,313 0 19,876 5,922 24,216 12,097 33,894 22' 779 39,702 30,629 Percentage of Residences U~ing Gas 0 30 50 67 77 SOURCES: 1960 Census of Housing, General Housing Characteristics: Alaska. 1970 Census of Housing, Detailed Housing Characteristics: Alaska. Anchorage Urban Observatory, University of Alaska. 1975 Housing Study. Alaska Gas and Service Company Annual Reports. F-62 1. Retrofitting of fuel oil for electric space heating in Fairbanks will continue, but more slowly than in the years since 1975 when a ban on new residential electric space heating was imposed by Golden Valley Electric Association (GVEA)~ This is because of the high cost of conversion from resistance electric heating to a fuel oil boiler. Some switching will be partial conversions not to an alternative central heating system but rather to room-by-room heating units. 2. Retrofitting of natural gas for electric space heating in Anchorage will continue but also at a slow pace because not only is the conversion cost high but also because a large portion of the electric space heating load is in multifamily units where conversion may be relatively more expensive and the incentive to the owner less if the tenant absorbs the cost of the fuel. The conversion of such units to condominiums might speed the conversion process but not guarantee it. 3.. The existing natural gas utility service areas do not expand significantly and thus capture a larger market share from the electric utilities. Such expansion might occur north into the Matanuska Valley, south into the Homer area, or into thinly populated portions of the Anchorage Borough such as the Hillside area and Girdwood. Such expansions would be a question of regu- latory policy and the economics of laying new gas mains. This relates to the fourth assumption. 4. The distribution of the population within the regions of the railbelt (particularly Anchorage) does not change significantly. If all future population increase asso- ciated with economic growth in the Municipality of Anchorage settled in the Matanuska Valley, then the use of electric space heating would expand relatively rapidly. However, this growth, particularly if it were accompanied by an increase in the density of settlement~ would be a stimulus to the extension of gas service into the Matanuska Valley, thus reducing the electric space heating proportion in favor of gas. 5. New natural gas utilities do not make natural gas an avail-· able alternative in either the Fairbanks or Glennallen- Valdez areas. 6. Presently available fuels, particularly natural gas, will continue to be available for residential space heating purposes. F-63 The amount of recoverable reserves of natural gas remaining in Cook Inlet is the subject of some controversy. This is understandable since exploration is still proceeding and new reserves may be discovered. The total is obviously finite, but the proportion of ~xisting reserves annually utilized for Anchorage space heating is relatively small. In 1979, for example, of 266 million mcf utilized, only 14 million was marketed to consumers for space heating and other uses.33 Thus, to the extent space heating is a priority use of natural gas, relatively modest reserves and reserve additions would be sufficient to satisfy even a rapidly growing market. To a certain extent, the space heating priority is automatically built into the gas distribution structure. Gas sales from the gas utility to the electric utilities are under interruptible contract· so that if supply constraints develop, the shortfall will occur in the availability of ga~ for electricity generation (at least in the· 34 short run). A reliable published estimate of currently proven recoverable of natural gas in Cook Inlet does not exist. Recent past estimates are as follows: 1. 7.044 trillion cubic feet as of January 1, 1977.35 2. 6.413 trillion cubic feet as of January 1, 1977.36 Estimates of undedicated reserves are as follows: 1. 3.919 trillion cubic feet as of January 1, 1977.37 F-64 2. 4.236 trillion cubic feet as of January 1, 1978, from the currently producing fields.38 3. 5.422 trillion cubic feet as of January 1, 1978.39 Estimates of potential additional resources are from 6 to 29 trillion b . f 40 cu ~c eet. On the basis of existing reserves, the supply of gas in Cook Inlet is sufficient to meet demand growth through 2000 if a large proposed LNG export .facility were not built. If it is built to preliminary de- sign capacity and consumes 3.2 trillion cubic feet of gas over a twenty- year lifetime, then the existi-ng supply will not be able to meet all 41 expected demands. If the availability of alternative fuels (including natural gas, fuel oil, but also wood) is reduced or if the relative prices of alter- native fuels rise and thus make electric space heating mo+e economically attractive, then the proportion of space heating needs met electrically might increase. It is clearly impo~sible to identify all of the conditions under ·which such a change might actually take place. Thus, it is also impossible to quantify the impacts on electricity use of all possible scenarios of changing relative electricity prices. The factors determining relative prices could be divided into four categories as follows: Alaskan market conditions (supply and demand), F-65 national and international market conditions, national energy regula- tions and policies, and Alaskan energy policies. Both Federal and state policies will undoubtedly alter the relative prices and availabilities of fuels which would result from normal market forces over the next thirty years. The state is in a central position in terms of being able to affect fuel prices and availabilities because of both its substantial surplus revenue position and its ownership of significant energy resources. The state government could, in the short run, easily subsidize the total energy requirements of the entire population out of excess revenues. It can provide fuel resources such as coal, oil, and natural gas to local markets at below market prices. Whether it will choose to do these things is a political question. Federal regulations and policies may act to make particular fuels more expensive or unavailable for specific uses. Environmental regula- tions on coal are one example of the former, while possible prohibition on the use of gas in the generation of electricity is another. Absent such government induced changes in energy fuel markets, the long-run trend in prices will likely be towards comparability and with world oil prices. In particular, it is plausible to assume that fuel oil prices and domestic oil prices will gradually approach the level of world oil prices as decontrol of prices is phased in. Prices of other fuels can thus be examined in relation to the fuel oil price. Examining region-by-region, the following price scenario .is reasonable: F-66 ~- Anchorage. Natural gas prices will rise relative to fuel oil as existing contracts expire and as pricing provisions of existing contracts are activated which require that all purchasers pay the same price as the highest priced purchaser from a field.42 The effect of these trends will be to raise the delivered price to the consumer of both natural gas and electricity since electricity is gas generated. Because of different contracts, the exact relative effects cannot be estimated accurately. The advent of the Pacific LNG facility has been estimated to have a larger price impact on electricity users rather than gas users, but 11 . . "1 . d 43 I overa pr~ce ~ncreases are not so eas~ y est~mate • n any event, as long as gas is used to generate electricity and current space heating practices are maintained, gas will be the less expensive space heating fuel. It is possible, but not likely, that gas prices will reach parity with fuel oil. Working against such parity is the cost of transportation of gas·to its alternative market on the West Coast. Working toward price parity is the fact that the market for gas on the West Coast may value the gas at its peak rather than baseload value. In this case, gas in Anchorage would lose its attractive price relative to fuel oil, but not electricity. In order for electricity to be the cheapest fuel of the three, it must be produced by a means other than natural gas or fuel oil such as coal or hydropower. Electricity thus generated has the potential for being least cost, although it is by no means assured. Fairbanks. As fuel oil prices rise, electricity prices will also increase because the present generating mode in Fairbanks utilizes fuel F-67 oil for incremental electricity generation. To the extent that electri- city consumption also grows, the cost of electricity will continue to exceed the fuel oil cost for space heating purposes. This link will broken, and the cost of electricity made independent of the fuel oil price if an alternative fuel such as coal or hydropower is used to generate electricity. In such a case, electricity may become a less costly fuel for space heating than fuel oil. Glennallen-Valdez. ·The cost of electricity is presently tied to that of fuel oil because fuel oil is used to generate electricity. the·future, this will no longer be the case since a hydroelectric facility is currently under construction in the area. In the short the integration of the electricity from hydropower is not anticipated reduce the price of electricity. If the price is stabilized at its current level, the price of fuel oil would have to increase by three times before electricity would be priced equivalent to fuel oil for space heating. This would occur in twenty years at a 6 percent rate fuel price increase, or ten years at a 13 percent rate of fuel price increase. Under these assumptions, a significant shif·t toward use of elec~ tricity could occur under the following conditions: Anchorage. Decreased availability and/or increased price of gas result in the addition of coal and/or hydroelectric baseload electric F-68 generating facilities by 1990. The electricity price does not fall relative to that of gas until 2000, however, because: 1. The initial cost of those capital-intensive facilities is large. 2. The main source of electricity will continue to be gas- fired turbines which, since the gas will have become expensive, will keep the average price of electricity high. Fairbanks. Increased demand.or very rapid increase in the price of fuel oil makes a coal plant attractive. If it is in place by 1990, it could immediately "back out" high priced oil, but the electricity price would not immediately fall relative to that of alternatives because: 1. The initial cost of the capital-intensive coal plant will be large. 2. The cost of the fuel oil-fired generation facilities will still be a part of the price of electricity. Glennallen-Valdez. By 1990, the fuel oil price may have increased sufficiently to make electricity from hydropower relatively attractive. For this to happen, any additional generating capacity requirements must also be met by low cost modes of generation. This study cannot predict these outcomes since they are obviously dependent upon not only the level of demand but also upon costs of supply, taking into account not only the cost of additions to systems but the impact on system cost of those additions. F-69 We can, however, analyze the case presented above where electricity becomes relatively inexpensive as a fuel. Based upon the foregoing, we assume a shift towards electricity for space heating beginning in the period 1995-2000 for Anchorage and Fairbanks and 1990-1995 for Valdez. The price advantage for electricity is assumed to be real and las but not of a substantial magnitude. Thus, the shift to electric space heating follows the pattern observed·in Fairbanks in the early 1970s, rather than the pattern in Anchorage in the 1960s during the shift to natural gas. That is, new installations are predominantly electric, existing units are not retrofitted to electric space heat. marily because of the cost of retrofitting to electric space pared to switching from an oil to a gas burner, for example) combined with the relatively short average tenure by an owner in a home. In addition, electric appliances became more attractive relative to natural gas and fuel oil. The electric incremental mode splits for water heaters, ranges, and clothes dryers increase at the same time that the shift to electric space heating occurs. The commercial- industrial-government sector projections are similar to those of the base case. The parameter changes for this case are summarized in Table F.22. F-70 TABLE F.22. PARAMETER VALUES: THE PRICE INDUCED SHIFT TOWAP~ ELECTRICITY CONSUMPTION IN THE RESIDENTIAL SECTOR CASE Parameter Region Greater Greater Glennallen- Anchorage Fairbanks Valdez SPACE HEAT incremental mode split (msi. ) J,t SF 1985 .19 .01 .02 1990 .19 .01 .02 1995 .19 .01 .9 2000+ .9 .9 .9 DP 1985 .19 0 0 1990 .19 0 0 1995 .19 0 .9 2000+ .9 .9 .9 MF 1985 .19 0 0 1990 .19 0 0 1995 .19 0 .9 2000+ .9 .9 .9 NH 1985 .19 .01 0 1990 .19 .01 0 1995 .19 .01 .9 2000+ .9 .9 .9 APPLIANCES incremental mode split (msi. t) J,e, WH 1985 .35 .5 .4 1990 .35 .5 .4 1995 .35 .5 .9 2000+ .9 .9 .9 c 1985 .66 .85 .4 1990 .66 .85 .4 1995 .66 .85 .9 2000+ .9 .85 .9 SF = single family MH = mobile home DP = duplex WH = water heating MF = multifamily C = cooking F-71 ENDNOTES: APPENDIX F 1. Eric Hirst and Janet Carney, "Residential Energy Conservation: Analysis of U.S. Federal Programs," Energy Policy, September 1977, pp. 211-222. 2. Jorgen S. Norgard, "Improved Efficiency in Domestic Electricity Use," Energy Policy, March 1979, pp. 43-56. 3. Battelle Columbus and Nepool Load Forecasting Task Force, "Report on Model for Long-Range Forecasting of Electric Energy and Demand to the New England Power Pool," June 30, 1977, p. 624. 4. California Energy Commission, "Appendix A: Analysis of Residential Energy Uses," Section A9. 5. Eric Hirst, William Lin, and Jane ·Cope; i'A Residential Energy-Use Model Sensitive to Demographic Economic and Technological Factors," Quarterly Review of Economics and Business, Vol. 17, No. ·2, p. 13. 6. Jerry A. Hausman, "Individual Discount Rates and the Purchase and Utilization of Energy Using Durables," Bell Journal of Economics,· Vol. 10, No. 1, p. 45. 7. Data from California Energy Commission, "Appendix A: Analysis of Residential Energy UseB," Appendix 7, Table A7.6. 8. Present average electricity consumption figures applied to 1960 saturation rate data results in a substantial overestimate of actual electricity consumption for that year. 9. California Energy Commission, Ibid. 10. California Energy Commission, Ibid., Section 12. 11. California Energy Commission, Ibid., Section 7, Table A7.5. 12. Compiled from California Energy Commission, Ibid., Section 7, Table A7.6. 13. California Energy Commission, Ibid., Section 7, Table A7.10. 14. California Energy Commission, Ibid., Section 10, Table Al0.2. 15. American Gas Association, Gas Facts, annual. 16. Midwest Research Institute, Patterns of Energy Use by Electrical Appliance, for Electric Power Research Institute, EPRIEA 682, 1979, p. 512. F-72 17. U.S. Department of Energy, Office of Public Affairs, "The Natural Energy Act," 1978. 18. Arthur D. Little, "Project Independence Residential and Commercial Energy Use Patterns 1970-1990," for Federal Energy Administration, 1974, p. 21-23. 19. Hirst and Carney, Ibid., p. 217. 20. California Energy Commission, Ibid., Section A3. 21. Arthur D. Little, Ibid. 22. Hirst and Carney, Ibid. 23. Larry Breeding. "Phase III Evaluation: ASHRAE 90-75 Energy Con- servation in New Building Design, HUD Minimum Property Standards," for State of Alaska, Energy Conservation Program, 1976. 24. U.S. Department of Energy, Ibid. 25. Eric Hirst and Bruce Hannon, "Effects of Energy Conservation in Residential and Commercial Buildings," Science, Vol. 205, August 17, 1979, p. 656-661. 26. Arthur D·. Little, Ibid., .p. 157. 27. Ibid., p. 137. 28. Ibid., p. 139 and conversations with local engineers. 29. Charles Rivers Associates, "Analysis of Household Appliance Choice," report for Electric Power Research Institute, 1979, Chapter 2 .• 30. In Alaska, this is definitely a factor. A 1976 survey in Fairbanks found the median length of residence in owner-occupied housing to be three years. John Kruse, Fairbanks Community Survey, Institute of Social and Economic Research, 1977, p. 5. A 1975 Survey of Anchorage found that 61 percent of households had lived at their present address less than three years. Diddy Hitchens et al, "Anchorage Municipal Housing Study," Anchorage Urban Observatory, 1976. 31. A recent study estimated the discount rate used by consumers in making appliance choice decisions to be Z5 percent, which is con- siderably above the opportunity cost of capital for many consumers. This suggests a bias in favor of systems with the least initial cost. Jerry Hausman, "Individual Discount Rates and the Purchase and Utili- zation of Energy Using Durables," The Bell Journal of Economics, Vol. 10, No. 1, p. 33. F-73 32. Based on conversations with Harold Schmidt of Anchorage Natural Gas Company. 33. Scott Goldsmith and Kristina O'Connor, "Alaska Historic and Projected Oil and Gas Consumption," prepared for the Alaska Royalty Oil and Gas Development Advisory Board and the Alaska State Legislature, 1980, p. 5. 34. Harold Schmidt, Ibid. 35. Stanford Research Institute, "Natural Gas Demand and Supply to the Year 2000 in the Cook Inlet Basin of Southcentral Alaska," for Pacific Alaska LNG Company, 1977, p. 37. 36. Alaska Department of Natural Resources estimate noted in Scott Goldsmith and Tom Lane, "Oil and Gas Consumption in Alaska 1976- 2000," for the Alaska Royalty Oil and Gas Development Advisory Board and the Alaska State Legislature, p. III.5. 37. W. H. Swift et al, "Alaskan North Slope Royalty Natural Gas: An Analysis of Needs and Opportunities for In-State Use," for State of Alaska, Division of Energy and Power Development, p. V.5. 38. Scott Goldsmith and Tom Lane, Ibid., p. III.7. 39. ·Ibid., p. IV.5. 40. Stanford Research Institute, Ibid., p. 36. 41. Scott Goldsmith and Tom Lane, Ibid., p. IV.5. 42. For a recent analysis of this, see Jack Kreinheder, State·of House Research Agency Report -Request No. 30, February 1980. 43. Ibid. F-74 APPENDIX G: PROJECTING MILITARY AND SELF-SUPPLIED INDUSTRIAL ELECTRICITY NET GENERATION G.l. Military Requirements Six major military installations, all of which produce and consume their own electricity, are located within the railbelt region of Alaska. The level of activity at these bases is a function of national defense policy and, as such, it is difficult to project. Historically , there has been considerable variation in the number of personnel stationed at these facilities with peaks occurring during 1\Torld \\Tar II and the Korean \\Tar. Table G.l sho\vs the variation in the level of military personnel for the whole state over the historical period. (Currently, about 80 percent of statewide military personnel are located in the railbelt.) The general trend since the late 1950s has clearly been for a gradual decline, but this may not be a reasonable guide to future personnel levels. We have not attempted to correlate military electricity consump- tion with the level of personnel or other factors for a number of reasons. A complete historical data series on military consumption would be difficult to obtain. A large and varying proportion of per- sonnel live off base, and, consequently, their residential consumption is satisfied by utility-supplied power. Finally, the difficulty in projecting military personnel nullifies the value of utilizing a functional relationship between personnel and electricity consumption. TABLE G.l. AVERAGE ANNUAL HILITARY PERSONNEL IN ALASKA (thousand) 1940 1 1960 33 1941 8 1961 33 1942 60 1962 33 1943 152 1963 34 1944 104 1964 35 1945 60 1965 33 1946 19 1966 33 1947 25 1967 34 1948 27 1968 33 1949 30 1969 32 1950 26 1970 31 1951 38 1971 30 1952 50 1972 31 1953 50 1973 27 1954 . 49 1974 26 1955 50 1975 25 1956 45 1976 24 1957 48 1977 25 1958 35 1978 23 1959 34 1979 23 Source: 1940-1959 -George Rogers, The Future of Alaska: The Consequences of Statehood, Resources for the Future, 1960-1965 -George Rogers, "Alaska Regional Population and Employment," ISER, 1967, p. 42. 1960-1969-MAP model.data. 1970-1979 -Alaska Department of Labor, "Alaska Population Overview, December 1979, p. 50. G-2 The Air Force conservation goal is to reduce their total energy requirements by 20 percent by 1985, according to the Alaskan Air Command. The Army may have a similar conservation goal. It is not clear what impact this policy will have on electricity consumption. Because of these difficulties which make detailed projections of military electricity requirements questionable, we assume the current level of net generation in all future years. Current requirements are shown in Table G.2~ G.2. Self-Supplied Industrial Requirements The largest industrial users of self-generated electricity in the railbelt are, with one exception, in the category of petroleum production, processing, and transportation. The University of Alaska, Fairbanks campus, is the only large public, non-utility generator of electricity. Table G.3 shows that most of the self-supplied electricity is centered in the Greater Anchorage Area. Offshore and onshore drilling and producing petroleum rigs contribute a major portion of the total load, along with the pipelines and other facilities for transporting and transshiping the petroleum. The major industrial facilities at North Kenai, consisting of two refineries, an LNG plant, and a fertilizer plant, complete the list of major consumers. In Valdez, the oil pipeline Pump Station 12 and the facilities at the pipeline terminal are the major consumers. G-3 TABLE G.2. RAILBELT MILITARY ELECTRICITY NET GENERATION (FY 1979) (103 MWh) Greater Anchorage Fort Richardson Army Base Elemdorf Air Force Base Total Greater Fairbanks Fort Greely Army Base Fort Wainwright Army Base Eilson Air Force Base Clear Air Force Base Total Total Railbelt 56.7 9?.8 155.5 14.4 36.8 47.0 80.0 178.2 333.7 Source: Military and Alaska Power Authority records. G-4 . ~. ' TABLE G.3. RAILBELT SELF-SUPPLIED INDUSTRIAL NET GENERATION (103 MWh) Area 1977 1978 1979 North Kenai 69.5 94.6 94.6 Valdez 39.4 54.8 54.8 Cook Inlet 208.9 _226.7 226.7 Interior Alaska 25.7 37.9 37.9 Total 343.5 414.0 414.0 Source: Alaska Power Authority worksheets • G-5 In Fairbanks, the Univer$ity and· pipeline Pump Stations 8 and 9 are the major consumers. In some cases, an industrial facility will both generate its own electricity and purchase power from the local utility. For example, Alyeska Pump Station 12 uses electricity provided by Copper Valley Electric Association for all its needs except the pumps themselves, which are powered by self-supplied generation. In other instances, the facility may purchase power but maintain its own backup generation capability. The difficult-ies in attempting to project self-supplied industrial electricity are that additions over time have been "lumpy" (large but infrequent) and that there is not always a clear criterion to deter- mine whether a particular consumer will choose to provide his electricity from self generation rather than from utility purchased power. In some instances, such as the case of offshore drilling and production plat- forms,_ self-supplied electricity is the only practical method of ob- taining power. In other cases, however, the industrial facility faces a choice, and the decision will depend upon the cost of self-generated electricity vs. the price of purchased electricity. Each instance will be different depending upon, among other things, the type of load, the capacity and load characteristics of the utility, and the resources available to the facility for generating electricity compared to those which the utility has available. G-6 The utilities are presently supplying a portion of the large indus- trial consumer load, even though in total the load is relatively small. This is the case in Fairbanks for the refinery and a portion of oil pipeline requirements, in.Valdez for a portion of oil pipeline require- ments, and in the Greater Anchorage Area for the refineries and the LNG plant as well as some petroleum production and transportation facilities. Thus, a portion of increments to industrial requirements is already included in the utility load projections. Self-supplied industrial _requirements should be limited to new industrial uses that would not normally be picked up by the utilities given the same general market conditions in the future as in the past. Having thus narrowed the definition of self-supplied industrial requirements, it is still possible to identify two types of industrial facilities. The first is any facility ~.;rhich chooses to locate in the railbelt independent of the price of electricity, while the second is any facility which chooses to locate in the railbelt because of price of electricity or the availability of electricity. We address ourselves to and consider only the first type of facility because the determina- tion of whether such "electricity intensive" industries will locate in Alaska in the future is a function of, among other things, future electricity price, which is beyond the scope of the present study. Table G.4 presents the self-supplied electricity requirements for . those facilities identified in the economic scenarios. Three have relatively small requirements which we assume to be incorporated in G-7 TABLE G.4. PROJECTED ADDITIONS TO RAILBELT SELF-SUPPLIED INDUSTRIAL ELECTRICITY REQUIREMENTS (103 MWh annual) Facility Economic Scenario Minimum Likely Maximum ,Consumption Start . (lo3 Ml;.fu) Date Consumption Start (103 Ml;.fu) Date Consumption Start (103 Ml;.fu) ~ Pacific Alaska LNG a 0 127 1985 127 Alpetco Refinery b 0 30 1985 306 Cook Inlet Oil 0 0 46 c Development Fairbanks aetro- chemicals 0 0 88 Northwest Alaska ( Gas Pipeline Incorporated in utility sales forecast. State Capital Move Incorporated in utility sales forecast. Beluga Coal Development Incorporated in utility sales forecast. ~omer Electric Association, Power Requirements Study, 1977. bCalculated from Alpetco Refin~ry Environmental Impact Statement, Volume II, p. 343 and, 413. Assuming 51.5 MW operating at. 80 percent capacity, 85 percent of the year for high case. Authors' estimate of basic refinery requirement for medium case. 1985 cAssumes a 20 percent increase from present requirements. dWith a capacity for processing 200 million cubic feet/day of royalty gas and if 8 percent of input is used for processin~ and if 25 percent of processing is steam and electricity, 4.0 x 10 btu daily would be used. (Based upon same proportions as the Alpetco refinery.) This can be converted to electricity by assuming a 25 percent conversion efficiency and 80 percent load factor. G-8 the utility projections. Of the other facilities, they all could have their electricity requirements met by utilities if it were available, except perhaps in the case of additional oil development in the Cook Inlet. Their requirements are large enough, however, that they should be treated, for projection purpose, as additions to, rather than com- ponents of, the utility electricity requirement. G-9 APPENDIX H HISTORICAL ELECTRICITY SALES AND ECONOMIC DATA ::X:: I N RAILBELT TOTAL ELECTRICITY SALES (GWh) GREATER ANCHORAGE AREA 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979p Chugach Anchorage Matanuska Electric Municipal Electric. Association Light & Power Association 165 133 34 190 143 36 207 159 39 235 170 39 262 190 44 309 222 50 368 254 61 435 288 72 485 326 81 517 350 92 596 405 118 668 468 147 727 492 172 781 498 223 803 523 232 PPre1iminary, based on data from various sourcesf * Approximate. TO FINAL CONSUMERS Homer Seward Electric Electric Association System Total 31 6 369 39 7 415 49 7 461 67 8 519 82 9 587 93 10 684 103 11 797 99 13 906 105 14 1,010 112 14 1,086 133 17 1,270 161 18 1,463 194 17 1,603 224,.~ 20 1,747 247 TABLE H.l. (continued) RAILBELT TOTAL ELECTRICITY SALES TO FINAL CONSUMERS (Continued) (GWh) GLENNALLEN- GREATER FAIRBANKS AREA VALDEZ AREA RAILBELT TOTAL Fairbanks Alaska Power Golden Valley Municipal and Copper Valley Electric Utilities Telephone Electric Association System Tok Total Association 1965 50 47 1 98 6 473 1966 59 49 ** 108 "'* 523 1967 66 *~" ** 66 'lc'l; 527 ::d 1968 84 58 *-1• 141 ~""~• 661 I 1969 104 66 w ~"* 170 '"* 758 1970 136 75 2 213 9 907 1971 175 . 76 ** 251 10 1,059 .1972 190 70 2 262 6**," 1,174 1973 206 81 3' 290 ll 1,311 1974 231 88 3 322 14 1,422 1975 300 110 3 413 24 1,707 1976 306 114 3 423 33 . 1,920 1977. 324 118 5 447 42 2,092 1978 310 116 5 432 38 2,217 1979p 302 37 PPreliminary, based on data from various sources. TABLE H.2. HISTORICAL RESIDENTIAL UTILITY SALES H-4 ~- .. GREATER ANCHORAGE AREA RESIDENTIAL ELECTRICITY CONSUMPTION Chugach Electric Association (CEA) Energy Delivered Year-End To Final Customers Customers Consumption/Customer (MWh) (kWh) 1965 111,587 15,446 7,224 1966 128,484 16,487 7,793 1967 134,985 17,037 7,923 1968 148,591 19,893 7,470 1969 . 166~146 22,036 7,540 1970 198,856 24,682 8,057 1971 236,857 25,761 9,194 1972 269,252 28,687 9,386 1973 287,484 29,077 9,887 1974 305,739 31,779 9,.621 19.75 359,922 34,031 10,576 1976 397,846 35,960 ll,064 1977 432,070 41,025 10,532 1978 472,040 • 43,542 10,841 1979p 477,189 42,761 l1,161 PPreliminary, based on data from various sources. SOURCE: Federal Energy Regulatory Commission, Power System Statement. H-5 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979p GREATER ANCHORAGE AREA RESIDENTIAL ELECTRICITY CONSUMPTION Anchorage Municipal Light and Power (AMLP) Energy Delivered Year-End To Final Customers Customers Consumption/Customer (H\Vh) (kWh) 34,656 6,664 5,201 35,o56 6-,516 5,380 38,426 6,894 5,574 42,825 7,544 5,677 47,781 8,043 5,941 54,518 8,477 6,431 63,038 9,295 6,782 72,993 10,130 -7,206 82,663 10,838 7,627 89,946 11,674 7,705 105,214 11,803 8,914 119,475 12,353 9,672 117,986 13,605 8, 672 115,638 14,374 8,045 116,211 13,517 8,597 PFreliminary, based on data from various sources. NOTE: Year-end customer data overstated in 1977 and 1978 due to a computer error within Municipality of Anchorage. SOURCE: Federal Energy Regulatory Commission, Power System Statement. H-6 -·:·, GREATER ANCHORAGE AREA RESIDENTIAL ELECTRICITY CONSUMPTION Matanuska Electric Associati~n (MEA)a Year-End Energy Delivered b To Final Customers Customers Consumption/Customer (MWh) (kWh) 1965 16,628 2,688 6,186 1966 18,012 2,707 6 ,'554 1967 19,623 3,022 6,493 1968 20,760 3,174 6,541 1969 24,861 3,611 6,885 1970 29,416 3,975 7;400 1971 37,791 4,281 8,828 1972 44,147 4,669 9,455 1973 51,026 5,045 10,114 1974 59,764 6,153 9, 713 1975 77,592 6,834 11,354 1976 96,280 7,681 12,535 1977 112,662 8,321 13,539 1978 152,133 10,152 • 14,986 1979p 157,889 10,362 15,237 aPalmer and Talkeetna stations. bFarm (including irrigation) and nonfarm residential. PPreliminary, based on data from various sources. SOURCE: Federal Energy Regulatory Commission, Power System Statement. H-7 GREATER ANCHORAGE AREA RESIDENTIAL ELECTRICITY CONSUMPTION Homer Electric Association (HEA)a Year-End Energy Delivered b To Final Customers Customers Consumption/Customer (MWh) (kWh) 1965 7,526 1,569 4,797 1966 9,809 1,753 5,596 1967 12,402 2,441 5,081 1968 17,673 3,182 5,554 1969 20,200 3,296 6,129 1970 22,768 3,312 6,874 . 1971 27,267 3,431 7,947 1972 28,299 3,491 8,106 1973 30,849 3,708 8,320 1974 33,752 4,215 8,008 1975 44,008 4, 773 9,220 1976 55,859 5,508 10,141 1977 70,742 7,346 9,630 1978 94,846 7,904 12,000 1979p 108,973 8,764 12,434 aHomer and Kenai stations. bFarm and nonfarm residential customers. PPreliminary, based on data from various sources. SOURCE: Federal Energy Regulatory Commission, Power System Statement. H-8 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 . 1976 1977 1978 1979 * GREATER ANCHORAGE AREA RESIDENTIAL ELECTRICITY CONSUMPTION Seward Electric System (SES) Energy Delivered Year-End To Final Customers Customers Consumption/Customer (Mllli) (kWh) 3,169 649 4,883 3,073 656 5,439 2,987 634 4,711 3,179 650 4,891 3,481 667 5,219 3,771 705 5,349 4,101 718 5,712 4,535 730 6,212 4,711 765 6,158 4,664 785 5,941 5,120 885 5,785 5,632 911 6,182 6,020 978 6,155 6,807 1,027 6,628 * *" * Not available SOURCE: Federal Energy Regulatory Commission, Power System Statement. H-9 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 ·1978 GREATER ANCHORAGE AREA RESIDENTIAL ELECTRICITY CONSUMPTION Total Anchorage Area System Energy Delivered Year-End To Final Customers Customers Consumption/Customer (MWh) (kWh) 173,566 27,016 6,425 194,434 28,028 6,937 208,423 30,028 6,941 233,028 34,443 6,766 262,469 37,653 6,971 309,329 41,151 7,517 369,054 43,486 8,487 419,226 47,707 8,788 456,733 49,433 9,239 493,865 54 ·606 9,044 591,856 58,326 10,147 675,092 62,413 10,817 739,480 71,275 10,375 841,464 76,999 10,928 H-10 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979p GREATER FAIRBANKS AREA RESIDENTIAL ELECTRICITY CONSUMPTION Golden Valley Electric Association (GVEA) Edergy Delivered Year-End To .Final· Customers Customers Consumption/Customer (MWh) (kWh) 23,142 4,036 5,734 29,184 4,213 6,927 33,444 4,402 7,597 41,917 . 4, 95'7 8,456 54,569 5,459 9,996 67,123 6,224 10,785 81,960 6,741 12,158 96,702 6,947 13,920 . 106,882 7,382 14,479 127,873 8,643 14,795 160,199 9,243 17,332 162,369 10,680 15,203 168,275 12,443 p,524 150,804 13,030 11,574 142,960 13,711 10,427 PPreliminary, based on data from various sources. SOURCE: Federal Energy Regulatory Commission, Power System Statement. H-11 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 * GREATER FAIRBANKS AREA RESIDENTIAL ELECTRICITY CONSUMPTION Fairbanks Municipal Utility System (FMUS) Energy Delivered Year-End To Final Customers Customers Consumption/Customer (MWh) (kWh) 16,172 4,147 3,900 17,485 3,957 4,419 * * * 19,461 4,387 4,436 22,327 4,564 4,892 23,419 4,532 5,167 24,456 4,443 5,504 24,248 4,540 5,341 25,952 4,443 5,841 25,909 4,618 5,610 30,181 4,634 6,513 31,302 4,739 6,605 29,497 4,754 6,205 27 ,109· 4,494 6,032 * * * Not available SOURCE: Federal Energy Regulatory Commission, Power System Statement. H-12 GREA~ER FAIRBANKS AREA RESIDENTIAL ELECTRICITY CONSUtWTION Alaska Power and Telephone, Tok Energy Delivered • y;ear-End To Final Customers Customers Consumption/Customer (MHh) (kWh) 1965 142 * * 1966 * * * 1967 * * * 1968 * * * 1969 * * * 1970 279 * * 1971 * * * 1972 396 * * 1973 411 * * 1974 470 * * 1975 603 * * 1976 730 * * 1977 795 155 5,129 1978 870 * * 1979 * * * * Not available SOURCE: Federal Energy Regulatory Commission, Power System Statement. H-13 1965 1966 1967 1968 1969 .1970 1971 1972 1973 1974 1975 1976 1977 1978 a * GREATER FAIRBANKS AREA RESIDENTIAL ELECTRICITY CONSUMPTION a Total Fairbanks Area System Energy Delivered ·Year-End To Final Customers Customers Co~sumption/Customer (HWh) (k\fu) 39,314 8,183 4,804 46 '669' 8,170 5, 712 * * * 61,378 9,344 6,569 76,896 10,023 7,672 90,542 10,756 8,418 106,4.16 11,184 9,515 120,950 11,487 10,529 132,834 11,825 11,233 153,782 . 13,261 11,600 190,380 13,877 13,719 193,671 15,419 12,561 197 '772 17,197 11,500 177,913 17,524 10,153 Net of Tok Not available H-14 GLENNALLEN-VALDEZ AREA RESIDENTIAL ELECTRICITY CONSUMPTION Copper Valley Electric Association (CVEA)a Energy Delivered Year-End To Final Customers Customers Consumption/Customer (~fu) (kWh) 1965 1,445 432 3,345 1966 * * * 1967 * * * 1968 * * * 1969 * * * 1970 2,133 561 3,802 1971 2,611** 676** 3,862 1972 1,528 324 4, 716 1973 2,887 680 4,246 1974 3,751 935 4,012 1975 7,656 1,48.7 5,149 1976 10,234 1,758 5,821 1977 10,895 1,601 6,805 1978 9,545 1,539 6,202 1979p 9,354 1,588. 5,890 a Glennallen and Valdez stations. PPreliminary, based on data from various sources. * ** Not available Valdez only SOURCE: Federal Energy Regulatory Commission, Power System Statement. H-15 TABLE H.3. HISTORICAL CONMERCIAL-INDUSTRIAL- GOVERNMENT CONSUMPTION DATA H-16 GREATER ANCHORAGE AREA COMMERCIAL-INDUSTRIAL-GOVERNMENT ELECTRICITY CONSUMPTION Chugach Electric Association (CEA) Energy Delivered To Final Customersa Year-End b Customers Consumption/Customer (m.Jh) (kWh) 1965 52,920 964 51,605 1966 60,601 1,047 53,626 1967 71,561 1,135 57,532 1968 84,513 1,381 55,652 1969 94,565 1,678 51,544 1970 108,298 2,040 53,087 1971 128,675 2,126 60,525 1972 163,566 2,449 66,789 1973 194,973 2,579 75,600 1974 208,855 2,835 73,670 1975 231,377 3,036 76,211 1976 264,731 3,494 75,767 1977 289,394 4,208 68' 772 1978 303,263 4,331 70,021 1979p 320,365 4,414 72,579 aCommercial and other (public authorities). b Other only since 1970. cOther only since 1970. PPreliminary, based on data from various sources. c SOURCE: Federal Energy Regulatory Commission, Power System Statement. H-17 _GREATER ANCHORAGE AREA COMMERCIAL-INDUSTRIAL-GOVERNMENT ELECTRICITY CONSUMPTION Anchorage Municipal Light and Power (AMLP) · Energy Delivered Year,-End To Final Customersa Customers Consumption/Customer (MWh) (k'tfu) 1965 92,889 2,071 44,852 1966 104,663 2,058 50,857 1967 116,157 2,060 56,387 1968 121,490 2,107 57,660 1969 135,306 2,115 63,974 1970 159,538 2,159 73,894 1971 181,374 2,226 81,480 1972 205,288 2,315 88,677 1973 233,312 2,350 99,282 1974 250,409 2,417 103,603 1975 289,296 2,464 117,409 1976 339,550 2,675 126,935 1977 365,510 2,800 130,539 1978 372,511 2,885 129,120 1979p 396,811 2,933 135,292 aCommercial and industrial. PPreliminary, based on data from various sources. SOURCE: Federal Energy Regulatory Commission, Power System Statement. H-18 GREATER ANCHORAGE AREA CO}lliERCIAL-INDUSTRIAL-GOVERNMENT ELECTRICITY CONSUMPTION Matanuska Electric Association (MEA)a Energy Delivered Year-End To Final Customers Customers Consumption/Customer (MHh) (kWh) 1965 16,441 412 39,905 1966 17,187 425 40,440 1967 18,172 490 37,086 1968 17,471 500 34,942 1969 18,148 511 35,515 1970 19,311 594 32,510 1971 22,239 599 37,127 1972 26,264 675 . 38,910 1973 28,252 739 38,230 1974 30,630 800 38,288 1975 38,756 980 39,547 1976 48,296 1,128 42,816 1977 57,263 1,265 45,267 1978 66,699 1,307 51,032 1979p 71,255 1,315 54,186 aPalmer and Talkeetna stations. PPre1iminary,.based on data from various sources. SOURCE: Federal Energy Regulatory Commission, Power System Statement. H-19 GREATER ANCHORAGE AREA COMMERCIAL-INDUSTRIAL-GOVERNMENT .ELECTRICITY CONSUMPTION Homer Electric Association (HEA)a Year-End Energy Delivered b Tc Final Customers Customers c Consumption/Customer (~.fu) 1965 23,419 416 1966 28,707 493 1967 30,705 561 1968 49,421 687 1969 63,155 705 1970 69,887 813 1971 75,955 861 1972 70,382 797 1973 74,194 831 1974 78,517 981 1975 88,714 1,066 1976 105,239 1,232 1977 122,512 1,355 1978 129,493 1,422.* 1979p 137,727 ** a d K · · Homer an ena~ stat~ons. bCommercial, industrial, and public buildings. cPublic buildings only since 1970. dPublic buildings only since 1970. PPreliminary, based on data from various sources. * ** Yearly average Not available (kWh) 44,647 48,032 .52' 989 63,086 72,852 85,962 88,217 88,308 89,283. 80,038 83,221 85,421 90,415 91,064 ** d SOURCE: Federal Energy Regulatory Commission, Power System Statement. H-20 -.·"·{ GREATER ANCHORAGE AREA COMMERCIAL-INDUSTRIAL-GOVERNMENT ELECTRICITY CONSU}~TION Seward Electric System (SES) Energy Delivered To Final Customersa Year-End b Customers Consumption/Customer (MHh) 1965 2,989 131 1966 4,109 124 1967 4,267 117 1968 4,367 129 1969 4,814 116 1970 5,695 178 1971 6,513 194 1972 . 7,680 184 1973 8,436 194 1974 8,640 199 1975 11,174 204 1976 12,080 260 1977 10,842 232 1978 12,409 274 1979 * * aCommercial, industrial, and public authorities. bPublic authorities only since 1970. cPublic authorities only since 1970. * Not available (k\Vh) 15,349 17,702 14,197 14,969 18,552 31,994 33,572 41.739 43,485 43,417 54,775 46,462 46,733 45,288 * c SOURCE: Federal Energy Regulatory Commission, Power System Statement. H-21 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 GREATER ANCHORAGE AREA COMMERCIAL-INDUSTRIAL-GOVERNMENT ELECTRICITY CONSUMPTION Total Anchorage Area System Year-End Energy Delivered To Final Customers Customers a Consumption/Customera (Mt.fu) (kWh) 188,658 3,994 47,235 215,267 4,147 51,909 240,862 4,363 55,206 277,262 4,804 57' 715 315,988 5,125 61,656 362,729 5,784 62,.713 414;756 6,006 69,057 473,180 6,420 73,704 539,167 6,693 80"557 577,051 7,232 79~791 659,317 7,750 85,073 769,896 8,789 87,598 845,521 9,860 85,753 884,375 10,219 86,542 aNumber of customers are slightly underreported and average consumption calculations are slightly overestimated prior to 1970 due to incomplete information prior to 1970 on the following customer categories: Chugach "other," Homer public buildings, and Seward public authorities. H-22 .';~~ J 'l . ' GREATER FAIRBANKS AREA COMMERCIAL-INDUSTRIAL-GOVERN}ffiNT ELECTRICITY CONSUMPTION · Golden Valley Electric Association (GVEA) Energy Delivered Year-End To Final Customers Customers Consumption/Customer (MWh) (kWh) 1965 25,850 523 49,426 1966 28,982 591 49,039 1967 30,830 576 53,524 1968 41,585 634 65,915 1969 49,284 703 70,105 1970 69,064 844 81,829 1971 84,295 914 92,226 1972 92,758 916 101,264 1973 98,744 973 101,484 1974 102,342 1,132 90,408 1975 133,972 1,209 110,812 1976 138,735 1,365 101,637 1977 155,426 1,649 94,255 1978 157,202 1,675 93,852 1979p 155,436 * * PPreliminary, based on data from various sources. * Not available SOURCE: Federal Energy Regulatory Commission, Power System Statement. H-23 GREATER FAIRBANKS AREA C0~1MERCIAL-INDUSTRIAL-GOVERNMENT ELECTRICITY CONSUMPTION Fairbanks Municipal Utility System (FMUS) Energy Delivered Year-End b To Final Customersa Customers (MWh) 1965 29,348 795 1966 30,.394 876 1967 * * 1968 36,321 835 1969 41,928 876 1970 49,496 1,044 ** ** 1971 48,761** 1,015** 1972 43,115 1,086 1973 52,079 1,081 1974 59,273 1,110 1975 76,787 1,133 1976 80,440 1,165 1977 85,037 1,185 1978 85,466 1,179 1979 * * aCommercial and other (public) categories. b Other only since 1970. c Other only since 1970. Consumption/Customer (kWh) 27,810 25,862 * 33,101 36,397 47,410 48,040 39,701 48,177 53,399 67 '773 .69,047 71,761 72,490 * * ** Not available. Based on author's estimate of street lighting requirements due to aberration in reporting method for these years. c SOURCE: Federal Energy Regulatory Commission, Power System Statement. H-24 GREATER FAIRBANKS AREA CO}lliERCIAL-INDUSTRIAL-GOVERNMENT ELECTRICITY CONSUMPTION Alaska Power and Telephone, Tok Energy Delivered Year-End To Final Customersa Customers Consumption/Customer (MWh) 1965 1,343 * 1966 * * 1967 * * 1968 * * 1969 * * 1970 1,537 * 1971 * * 1972 1,949 * 1973 2,663 * 1974 2,526 * 1975 2,652 * 1976 2,730 * 1977 4,089 75 1978 4,514 * 1979 * * aCommercial, industrial, and other categories. * Not available (kWh) * * * * * * * * * * * * 54,520 * * SOURCE: Federal Energy Regulatory Commission, Power System Statement. H-25 GREATER FAIRBANKS AREA COMMERCIAL-INDUSTRIAL-GOVERID-lENT ELECTRICITY CONSUMPTION Total Fairbanks Area Systema Energy Delivered Year-End b Consumption/Customerb To Final Customers Customers (MWh) (kWh) 1965· 55~198 1,318 41,880 1966 59,376 1,467 40,474 1967 * * * 1968 77 ~906 1,469 53,033 1969 91,212 1~579 57,766 . 1970 118,560 1,888 62~797 1971 133,056 1,929 68,977 1972 135~873 2,002 67,869 1973 150~823 2,054 73,429 1974 161,615 2,242 72,085 1975 210,759 2,342 89~991 1976 219~175 2,530 86,630 1977 240,463 2,834 84,849 1978 242,668 2~854 85,027 a Net of Tok and University of Alaska. b Number of customers are slightly underreported, and average consumption calculations are slightly overestimated prior to 1970 due to incomplete information prior to 1970 on the following cus- tomer categories--FMUS other (public). * Not available H-26 * GREATER FAIRBANKS AREA COMMERCIAL-INDUSTRIAL-GOVERNMENT ELECTRICITY CONSUMPTION University of Alaska Energy Delivered To Final Consumers (MWh) 1965 * 1966 * 1967 * 1968 * 1969 * 1970 21,768 1971 * 1972 29,567 1973 31,913 1974 27,646 1975 28,259 1976 27,195 1977 25,644 1978 * 1979 * Not available SOURCE: Federal Energy Regulatory Commission, Power System Statement. H-27 GLENNALLEN-VALDEZ AREA COMMERCIAL-INDUSTRIAL-GOVERNMENT ELECTRICITY CONSUMPTION Copper Valley Electric Association (CVEA)a Energy Delivered b Year-End To Final Customers Customers c Consumption/Customer (MWh) (k\fu) 1965 4,188 141 29,702 1966 * * * 1967 * * * 1968 * * * 1969 * * * 1970 7,235 240 30,146 1971 7,657** 249** 30,751 1972. 3,842 122 31,492 1973 8,130 371 21,914 1974 10,193 354 28,794 1975 16,062 426 37,704 1976 22,465 455 49,374 1977 31,307 491 63,762 1978 28,604 495 57,786 1979p 26,917 492 54,709 aGlennallen and Valdez stations. bCommercial, industrial, and other (public buildings). cPublic building customers only since 1970. PPreliminary, based on data from various sources. * ** Not available Valdez only c SOURCE: Federal Energy Regulatory Commission, Power System Statement. H-28 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 TABLE H.4. MISCELLANEOUS ELECTRICITY CONSUMPTION (Primarily Street Lighting) (MWh) GREATER GREATER ANCHORAGE AREA FAIRBANKS 6,907 2,429 5,439 2,585 6,322 1,560a 8,875 2,207 10,257 2,216 11,845 2,289 13,682 11,197 14,086 2,984 15,940 3,361 16,609 3,354 18,619 8,945 18,.542 7,193 17,707 3,467 21,362 5,864 aGolden Valley Electric Association only. * Not available. AREA GLENNALLEN- VALDEZ AREA 0 * * * * 115 136 134 151 97 130 152 171 182 SOURCE: Federal Energy Regulatory Commission, Power System Statement. H-29 TABLE H.5. HISTORICAL RAILBELT EMPLOYMENT H-30 f, {; GREATER ANCHORAGE AREA '(. NONAGRICULTURAL WAGE AND SALARY EMPLOYMENTa ):. (thousand) ~', J: t. I Census Division ,. ~'· ~· "" Matanuska-~:. ~: Year Anchorage Kenai-Cook Inlet Susitna Seward Total ~ ... ,, ::·: 1965 30,704 1,756 1,085 621 34,166 1966 31,519 2,465 1,139 645 35,.768 1967 32,958 3,678 1,07.5 638 38,349 1968 34,021 4,470 988 602 40,081 1969 37,789 4,144 1,002 626 43,561 1970 39,667 3,546 1,142 689 45,044 1971 44,616 3,454 1,415 774 50,259 1972 48,252 3,818 1,447 809 54,326 1973 50,630 4,049 1,610 862 57,151 1974 58,716 4 ,.487 1,786 934 65,923 1975 69,561 5,595 2;149 1,152 78,457 1976 73,019 6,473 2,398 1,137 83,027 1977 76,997 7,340 2,653 1,155 88,145 1918 76,942 6,565 3,083 1,226 87,816 aDoes not include active-duty military or self-employed. SOURCES: 1975 through 1978: Alaska Department of Commerce and Economic Development, Division of Economic Enterprise, "Numbers: Basic Economic Statistics of Alaska Census Divisions," November 1979. 1965 through 1974: Alapka Department of Labor. H-31 Year 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 GREATER FAIRBANKS AREA NONAGRICULTURAL HAGE Al.'iD SALARY EMPLOYMENTa (thousand) Census Division Fairbanks Southeast Fairbanks 28,740 2,044 28,118 2,615 24,868 819 21,662 874 Total 11,511 11,767 11,929 12,383 13,901 14,377 14,648 15,583 15,480 19,749 30,784 30,733 25,687 . 22,536 ~oes not include active-duty military or self-employed. SOURCES: 1975 through 1978: Alaska Department of Commerce and Economic Development, Division of Economic Enterprise, "Numbers: Basic Economic Statistics of Alaska Census Divisions," November 1979. 1965 through 1974: Alaska Department of Labor. H-32 a GLENNALLEN-VALDEZ AREA NONAGRICULTURAL WAGE AND SALARY EMPLOYMENTa (thousand) Year Valdez-Chitina-Whittier 1965 757 1966 826 1967 773 1968 667 1969 786 1970 859 1971 1,087 1972 906 1973 988 1974 1,529 1975 4,763 1976 8,049 1977 4,199 1978 2,043 Does not include active-duty military or self-employed. SOURCES: 1975 through 1978: Alaska Department of Commerce and Economic Development, Division of Economic Enterprise, "Numbers: Basic Economic Statistics of Alaska Census Divisions," November 1979. 1965 through 1974: Alaska Department of Labor. H-33 TABLE H.6. HISTORICAL RAILBELT HEATING DEGREE DAYS H-34 Season 1960-61 1961-62 1962-63 1963-64 1964-65 1965-66 1966-67 1967-68 1968-69 1969-70 1970-71 1971-72 1972-73 1973-74 1974-75 1975:-76 1976-77 1977-78 1978-79 Normal GREATER ANCHORAGE AREA HEATING DEGREE DAYS (Annual) Anchorage Homer 10,261 10,026 11,524 10,696 10,406 9,966 10,781 10,108 11,064 10,577 11,174 10,840 11,384 10,278. 9,997 9,793 10,779 10,410 9,351 9,446 11,670 11,340 12,077 11,471 11,555 11,223 10,807 10,868 10,389 11,115 10,694 9,050 8,901 9,701 9,668 9,395 10,911 10,364 Talkeetna 11,160 12,621 11,304 11,845 11,991 12,499 11,947 11,364 11,851 10,631 13,280 13,406 12,203 12,235 12,081 12,340 10,268 10,984 11,708 SOURCE: "Local Climatological Data," National Oceanic and Atmospheric Administration, Environmental Data and Information Service, National Climatic Center, Asheville, N.C. H-35 GREATER FAIRBANKS AREA HEATING DEGREE DAYS (Annual) Fairbanks Season 1960-61 14,009 1961...;.62 14,786 1962-63 13,692 1963-64 14,912 1964-65 15,009 1965-66 15,688 1966-67 14,544 1967-68 13,671 1968-69 14,664 1969-70 12,939 1970-71 15,215 1971-72 15,141 1972-73 13,547 1973-74 14,391 1974-75 13,808. 1975-76 14,683 1976-77 12,674 "1977-78 13,635 1978-79 13,802 Normal 14,344 SOURCE: "Local Climatological Data," National Oceanic and Atmospheric Administration, Environmental Data and Information Service, National Climatic Center, Asheville, N .~c. H-36 f." ~· ~.·. ~t 1~-•.... . . .. ~·- 1,.· . ~ I ••• I l~ .. ', ' ~- GLENNALLEN-VALDEZ AREA HEATING DEGREE DAYS (Annual) Gulkana Season 1960-61 13~707 1961-62 14~551 1962-63 13~227 1963-64 14,183 1964-65 14,062 1965-66 15,054 1966-67 14,527 1967-68 13,290 1968-69 14~594 1969-70 12,975 1970-71 15,002 1971-72 15,810 1972-73 14,266 1973-74 1974-75 13,768 1975-76 14,061 1976-77 1977-78 13,966 Normal 13,937 Valdez 10,682 9' 911. 10,255 9,130 9,422 10,545 SOURCE: "Local Climatological Data," National Oceanic and Atmospheric Administration, Environmental Data and Information Service, National Climatic Center, Asheville, N.C. H-37 TABLE H.7. SEASONAL VARIATION IN ELECTRICAL ENERGY GENERATION Summer Winter July-August Sales December-January Sales Percent Percent of Percent Percent of of Annual Monthly Average of Annual Monthly Average SOUTH CENTRAL 1964 13.1 78.6 21.1 126.6 1965 13.6 81.6 20.9 125.4 1966 14.2 85.2 21.0 126.0 1967 13.5 81.0 21.1 126.6 1968 13.1 78.6 22.0 132.0 1969 13.5 81.0 21.0 126.0 1970 13.4 80.4 22.1 132.6 1971 12.8 76.8 21.3 127.8 1972 12.5 75.0 21.9 131.4 1973 12.9 77.4 20.8 124.8 1974 13.2 79.2 20.8 124.8 1975 13.0 78.0 21.7 130.2 1976 13.0 78.0 20.2 121.2 1977 12.8 76.8 21.7 130.2 1978 13.6 81.6 INTERIOR 1964 14.4 86.4 21.2 127.2 1965 12.5 75.0 21.4 128.4 1966 13.0 78.0 22.3" 133.8 1967 14~9 89.4 21.4 -128.4 1968 . 14.6 87.6 22.5 135.0 1969 12.0 72.0 22.9 137.4 1970 11.7 70.2 25.1 150.6 1971 11.1 66.6 22.9 137.4 1972 11.4 68.4 23.6 141.6 1973 11.5 69.0 22.2 133.2 1974 11.2 67.2 23.7 142.2 1975 10.8 64.8 24.0 144.0 1976 11.9 71.4 21.8 130.8 1977 12.6 75.6 23.3 139.8 1978 12.9 77.4 SOURCE: Compiled from data supplied by the Alaska Power Administration. H-38 APPENDIX I. A REVIEW AND COMPARISON OF RAILBELT ELECTRIC PmVER REQUIREHENT PROJECTIONS I.l. Susitna Riv.er Basin: A Report on the Potential Development of Water Resources in the Susitna River Basin of Alaska by U.S. Department of Interior, sponsored and prepared by the Bureau of Reclamation, Alaska District Office, 1952. This report did not present a detailed analysis of the electric load beyond the accompanying Figure I.l which depicts a 30-year forecast. Residential space heating was assumed to be 50 percent electric by 1970, resulting in an average residential use of 10,000 KWh annually and total consumption of 635,000 }fifu. The market area (equivalent to the present study minus the Glennallen-Valdez area) was projected to have a popula- tion of 260,000 in 1970. Farm requirements were projected at 66,000 Ml~ based on 3,800 to 4,000 farms by 1970. Commercial and municipal require- ments were projected at 256,000 Ml~ for the same year. Projected large industrial electricity users included an electrified railroad, minerals, petroleum, synthetic fuels from coal, and local industries together con- ! suming 1,732,000 MWh in 1970. ~ ;l 1.2. Devils Canyon Project: Alaska, Feasibility Report by U.S. Department of Interior, Bureau of Reclam~tion, Alaska District, 1960. Gross generation requirements in this study are projected for 22 years for utilities, large industry, and military as shown in Figure I.2 and Table I.l. Simple growth rates ranging from 10 to 12 percent annually were FIGURE 1.1. (/) n:: ~ 0 H I I N 1- 1- <! 3: 0 _j -X': z 0 - ....J ....J - ::'! -POWER REQUIRMENT L_ -/ 1---BY CLASS OF USE / t--4 800 ACTUAL a ESTIMATED FUTURE / L1 t--_/· lv / v 1---,~ "' / ./ 1---'><v~ / v -4000 TOTAL POTENTIAL / -ENERGY REQUIREMENT <v'?' / v f---' I ~ /~"'v f /<v~CJ/ I--I /~~/ t--3200 t--/ Vc.,<y / / 'V 1---ACTUAL ESTIMATED FUTURE I / / .... ~-t-- -<-_,_ "'~ I-- t--2400 I / 00 I v v ~'> t--' t--J 1/. v 0<v -I I "?-~ I /I 'V -1600 j I v - -I 'I r--./ V; ,......., V, v -f---8 00 v v --"' 1--/ L ~~LII AR'f ,...../ 1---_,..,..v l--c--v 1---~ I I 1---·~ ~ ~ f.---:--OTHER -r--~ 1--~ /_ ..-= 194 0 ·44 48 52. 56 60 64 68 72 76 80 GALAN D E.. Y. E R..S 0 Q) 3,400 3,200 10 "- 3,000 2,800 2,600 2,400 2,200 If) 0::: ::::> 0 I 2,000 t- t- ), c::x: 1,800 3: 0 ..J ~ 1,600 LL.. 0 If) z 1,400 0 ...J ...J 1,200 ~ 1,000 800 co <t 600 <t 400 <t 200 0 <t ()) - FIGURE 1.2 . rz3 . Utility [Sj ·Lorge Industrial ~ 40 Of.:, of Military Recorded Projected Firm Energy Generation With Interim Capacity Needed-Assumed To Be Steam 60 65 CALENDAR I-3 // M i scellaneous Hydro ///////////////// 70 75 YEARS i ! ·I I I ' TABLE I.l. Projection of Total Energy Load Devil Canyon Power Market Area Unit: Million kilo,.mtt-hours .:..:. ita:;:y 'l'o·:~a:f. Calendar Ncnr~.:l.litary lot·.n.s 1o<=,d 1oat't ·-----Sup:Pl:.ed Supplied supplied supplied Year Largo by by by by Utility inO.ustria.l Tot.a.1 .t. ,.1 '+ \!• .• 1.. J. .,y p:c'l),')':!~t prc,i~ct pro.~;~t 1960 287 22 309 309 H 1961 317 22 339 339 I 1962 351 19 370 370 ~ 1963 388 16 404 4ot~ 1964 429 14 443 443 1965 474 19 . 493 1 ~93 1966 524 21 545 545 ' ... 1967 580 22 602 602 1968 6lJ.1 25 666 666 1969 711 42 773 270 503 158 661 1970 789 51 • .8110 2'!0 5?0 158 728 1971 876 6o 936 270 . 665 158 82!+ 1972 972 75 1,01~7 270 777 158 935 1973 1,077 99 1 ;7"' 270 oo6 158 1~06'-~ , ... 0 "' 1974 1,191 137 1,328 270 1,058 15G 1 o~~" 1::..J.O 1975 1,315 184 1,499 270 1,229 158 1,387 1976 1,4!.~9 2h2 1)691 270 1, l~?.l 158 1,579 1977 1 '::;\CI:+ 305 1 ~··::9 270 1,()29 158 1,~(87 s -.. 1 _, 1978 1,75J 384 2,13'7 270 1,867 158 2,025 1979 1,92~~ 463 2,39). 270 2,121 1~:3 2,~279 1980 2 ,:'...21 51~.:1 ,..., 1;-63 2j'O 2J 3~J3 158 2,5,1 '-1-.J 1.981 2,333 . 6l6 2,949 270 2,6'(9 158 2,837 1.982. 2,566 695 3,261 519 2 .~7~2 153 2,900 ------·----- ·0 )(") v\ ' ... . 0J I •• ru r-1 \,;) CIJ " ('I') "' ,(l Ll'\ " C'J TABLE I. 2. Projection of Alaska's Electric Power Requirements, Electric Utility Systems, and Nonutility lnstallaHons Region and type of load Load center number (fig. 4) Northwest. ................................... . Utility ..•................ 1,2,3 ......... . Do.................. (') · Nonutility. . . . . . . . . . . . . . . . (') Southwest. .. · ......•.................•........ Utility ................... 4, 5, 5 ......... . Do.................. (') . Nonutility (1) ••••• 0 ••••••••• 0 <:>outncenrral .................................. Anchorage-Kenai .......... 9 to 13 ......... Utility ............................... Nonutility (military) .............•..... Other areas ............... 7. 8. 14. !.), ..... Utility .............................. . Nonutility ........................... . Utility.. . . . . . . .. . . .. . (1) Nonutility (1) 0 0 •••••••••• Interior ...................................... Fairbanks ................. 15 ............• Utility ............................... Nonutility (military) ................... Utility •...•.......... (1) ••••••••••••• Nonutility ............ (') ............. Southeast ..................................... Utility ..•...........•.... I7 to 24· ........ Nonutility (industrial) ...... 20 and 23 ...... Utility ................... (1) ............. Non utility .........•...... (1) .•........... Total utility requirement. ................ Total nonutility require- ment ................................. Total Alaska ..................... · ....... ····· I Scattered uonload center loads. 1965 Energy mwh. 52,927 8, 219 468 44,2·10 154,293 7, 038 I, 255 146 000 ' 643,473 563, 749 406,604 157, I45 56. 030 22, 917 33, 113 7,494 16 200 ' 368,860 239,669 I05, 857 132,802 2, 191 I27, 000 419,942 155,023 246, 621 2, 298 16,000 720,371 919, 121 1, 639,495 Peak de- mand (mw.) 12.09 1. 86 • I8 10.05 35. I1 I. 55 . 25 33 30 144.07 126. 51 92.65 33.85 11. 75 5.05 6. 70 2. 10 3 70 81. 89 52. 23 25. 16 27.07 • 55 29. II 59.84 33. 75 31. 60 . 78 3. 70 163.92 179.08 343 .00 1975 1985 Energy mwh. Peak de-Energy mand mwh·. Peak de- mand (mw.) 72~ 690 I8, 100 700 53,890 I89, 800 . I2,800 2,000 175 000 ' I, 184,240 1, 364, 720 I, 137, 840 226, 880 88 620 50,660 37,960 ll, 600 19 300 ' 654, 130 500, 110 275,850 2U,260 4,020 I 50,000 609,050 323,370 263,000 3,680 19,000 I, 840,620 I, 169,290 3,009, 9!0 (mw.) I6. 52 4.I2 .30 12. 10 43.I5 2. 85 . 40 39 90 324.48 297.79 249. 79 48.00 19.29 Il. 09 8.20 3.00 4 40 H4. 71 109. 25 54.26 45.00 . 95 34.50 111. 04 70.79 35.00 • 85 4.40 408.40 231.50 639. 90 IlO, 680 44,790 I, 100 64, 790 237,990 24·, 790 3,200 210 000 3, 647, 890 3,412,090 3, 201, 190 240,900 165 000 118,690 46,310 I8,900 2I 900 ' 1, 145,680 967,980 72I, 350 246,G30 6, 700 I71,000 959, 730 668,630 263,000 6,100 22,000 4, 815,440 I , 286,530 6, 101, 970 25. I9 10.24 .35 14.60 51.05 5.5I .55 45 00 784.94 739.49 689.49 50.00 35.85 25.95 9.90 4.60 5 00 256.29 215.89 I 54.69 51. 20 I.40 39.00 183. 24 141. 94 35.00 1. 30 5.00 1, 046.02 254. 70 1, 300. 72 NOTE.-1965 utility actual, non utility partly estimated; 1975 and 1985 estimated. I-5 assumed for the utility load. Large industrial requirements were generally unspecified but were projected to total 695,000 llivh by 1982. Military requirement s were assumed to remain constant at a level of 360,000 MWh annually. 1.3. Alaska Power Survey, Federal Power Commission, 1969. Utility and non-utility power requirements were projected to 1985 for all regions of the state. As shown in Table 1.2, the sum of the Southcentral and Interior requirements (excluding Glennallen-Valdez and large industrial but including military) were projected to be 4,410,000 MWh in 1985. This projection of growth was predicted upon a population estimate of 410,000 in 1975 and 550,000 by 1985. I.4. 1974 Alaska Power Survey, U.S. Department of Interior, Alaska Power Administration, 1974; 1976 Alaska Power Survey, Federal Power Commission, 1976; and Devils Canyon Status Report, U.S. Department of Interior, Alaska Power Adminis- tration. The load estimates of these reports are identical and thus dis- cussed simultaneously. Tables I.3 and I.4 present the utility and total load projections by reg{on through 2000, and Figure I.3 graphically depicts the statewide load growth projected. The Southcentral and Yuko n regions generally conform to the railbelt as presently defined except that the Alaska Power Administration includes Kodiak as part of the Southcentral region. As shown in the tables and figure, this was the I-6 Region Southcentral Yukon (Interior) H I -...I Total South central Yukon (Interior) Total South central Yukon (Interior) Total TABLE 1.3 • ... •. Regional Utility Load Estimates, 1972-2000 Actual Requireme nts 1972 Peak Demand 1000 KW 224 69 293 Annual Energy Million KWH 1,037 307 1,344 Peak Demand 1000 KW 680 200 880 610 180 790 530 160 690 1980 Estimated Future Requirements 1990 2000 Annual Energy Million KWH 2,990 870 3,860 . 2,670 780 3,450 . 2,340 680 3,020 Peak Annual Peak Annual Demand Energy Demand Energy 1000 KW Million KWH . 1000 KW Million KWH --------~--~----~ Higher Rate of Growth 1,640 7,190 3,590 460 2,020 970 -- 2,100 9.210 '4, 560 Likely Mid Range of Growth 1,220 340 1' 560 ' Lower 980 270 1,250 Rate 5,350 1,500 -- 6,850 of Growth 4,290 1, 200 5,490 .· 2,220 600 2,820 . 1,470 390 1,860 15,740 4,230 19,970 9, 710 2,610 12,320 6,430 1, 730 . 8,160 Note: Estimated future peak demand based on 50 percent annual load factor. Source: Alaska Power Survey, Technical Advisory Committee on Economic Analysis and Load Projection. ~--------__________ .. _ ----------· .. H I 00 Region Southcentral Yukon (Interior) Total Southcentral Yukon (Interior Total Southcentral Yukon (Inter;or) Total TABLE 1.4. Regional Total Load Estimate, 1972-2000 Actual Requirements Estimated Future Requirements 1972 1980 1990 Peak Annual Peak Annual Peak ·Annual Peak Demand Energy Demand Energy Demand Energy Demand 1000 KW Million KWH 1000 KW Million KWH 1000 KW Million KWH 1000 KW Higher Rate of Growth 317 11 ·465 990 51020 5,020 30,760 7,190 115 542 ' 330 1,610 760 3,980 1,390 -- 432 2,007 1,320 6,630 , 517 80 34,740 8,580 Likely Mid Range of Growth 790 3,790 1,530 7,400 3,040 280 1,310 470 2,270 910 1,070 51100 2,000 91670 3,950 Lower Rate of Growth 650 3,.040 ·1,160 51430 . 1,790 250 ·1,140 370 ·1,760 530 900 4,180 1,530 .71190 2,320 2000 Annual Energy Million KWH 40,810 7,000 47,810 15,300 4,610 19,910 8,510 2,540 11,050 Note: Assume 80 percent annual load factor for industrial requirements; 50 percent for . utility requirements. Higher estimate includes nuclear enrichment facility in 1980's with requirements of 2. 5 million kilowatts. Sou-rce : A.la.s 'ka. Pov..re-r Su-rvey • Technical Advisory Committee on Economic Analysis and Load Pro'ection. ~ ... ... 11 i3 ... ... ::l J' 11 >. .. ... ... ... ) d 1 .. 1 5 ~ ~ !J ... ) ::l .. .. I) u j j .. ) ... .. .. !) ... ~ j I) I) .. I) ... .. -·-·- 0 "l ~ 0 ~ Q) a 0 0 () ~ ~ (/) )o Cl ~ 0 !!< ~ rl ;1. rl ACT UAL < 1965 1970 .. () 0 )o . ::1 0 (/) FIG URE 1.3. ALASKA TOTAL ENERGY REQUIREMENTS (1965-2000) Larg er energy r eq uireme nt s du e t o l arge in dus tr i a l l oad , suc h cs a nuc l ear fuel enr ich.men t pla nt or ot her very large .en ergy int ensive fac i li t y. PROJEC TED 19 8 0 19 90 1995 YEAR -··---- I-9 60 50 4 0 30 (/) 0::: ::::> 0 :X: 2 0 1- 1-<! ::: <! 0::: w 1- 10 9 8 7 (/) 6 1-z w 5 ::E w 4 0::: ::::> 0 w 3 0::: >-~ 0::: w 2 z w I 2000 ---- first study which attempted to delineate a range of forecasts. The 1980 "likely midrange rate of gro\vth" forecast is somewhat high. The projection for utility load utilized the 1980 load projections of the actual utilities. These are sho\vn, for railbelt utilities, in Table I.S. The high and low estimates for 1980 were 20 percent above and below, respectively, those aggregated estimates. For the subsequent decades, growth rates were assumed in each case based upon unidentified population projections. The growth rates declined over time, reflecting the assumtion of increasing appliance saturation, efficiency, and utili- zation of conservation measures. .In the "likely midrange" case, the growth rates were assumed to be 7 and 6 percent, respectively, for the 1980s and 1990s. Military requirem~ts were assumed to grow at 1.7 percent annually. The industrial load growth estimate was based upon a State of Alaska Department of Commerce and Economic Development study of possible .industrial projects. This estimate is shown in Table 1.6. The assump- tions behind these estimates are: 1. a high probability of major new petroleum, natural gas, coal, and other new mineral production and processing; 2. significant further developments in timber processing; and 3. good possibilities that Alaska energy and other re- sources \vill attract energy intensive industries. The estimates vary basically in the speed with which developments occur except that the high case includes a nuclear enrichment facility I-10 H I I-' I-' .. .:; .. TABLE 1.5. Utility Load Estimate Extended to 1990 Projected Requirements Generation 1972 1980 1990 Location and Utility Symbol Southcentral City of Anchorage (AML&P} ~ Chugach (CEA) ]/ Include: Homer (HEA) ]} Kenai (HEA) 3/ Seldovia-Port Graham (HEA) ]/ Sewat·d (SL&P) ]/ · Tyonek Y. Palmer-Talkeetna (MEA) 1/ [S"Ubtota ;·~-A~~h~rage-Cook In 1 et Yukon (Interior} Fairbanks Area . Peak Annual · Demand Generation 1000 KW Million KWH 71.9 31 o. 1 . 123. 5 600.0 (16. 5) (85.1) (4.7) (22. 1) (0.8) (3.4) Not Available Not Available 15.9 . 73.9 211.3 984.0 ..... ··-···- Peak Demand 1000 KW 123.0 360.0 (36.5) (7.9) ~1.6~ . 5.2 (2. 1 )' 54.0 537.0 City of Fairbanks (FMU) 3J 21.0 90.0 . 51.8 Golden Valley (GVEA) 1( 45.9 211.0 131.0 IS u btota 1 , Fa i rba _n k_s_A_r_e_a ____ ....:..6..::...:..6. 9 __ __.=.c:30 1 • 0 ___ 1-=-=82. 8 . ···-·----~-----~-·--·----.. ---~---·--··-·· .. --·---··-·----·· Annual Peak Annual Generation De mand Generation Million KWH 1000 KW Million KWH 658.0 314.0 1 ,680.0 1,800.0 1,065.0 5,370.0 (190.0 (90~0) (490.0) {37.6) ( 14 .. 9) (71.1) (6.8~ ~3.5~ (16.1l (27.9 8.0 (45.0 (8. 7) (3.0) ( 12. 3 225.0 181 .0 850.0 2,683.0 1 ,560. 0 7,900.0 ........ -·. :- 221.0 161.1 687.3 610.0 379.0 1,790.0 8~31=·=0 ======~40_:_1 ___ ?~~~!._:.: _____ _ Region H I .... N Southeast Southcentral Yukon (Interior) &eutheast Southcentral Yukon __ q_~!~~i9_r) ~~ Southcentral Yu):con (Interior) TABLE I.6. .. Assumed Regional Industrial Power Requirements, 1972-2000 A~tual Requirements Peak Demand 1000 KW 4 58 ··--·-····- 1972 Annual Energy Million KWH -·· Peak Demand 1000 KW ~ 260 70 Estimated Future Requirements 1980 1990 Annual Peak Annual Peak Energy Demand Energy Demand Million KWH 1000 KW Million KWH 1000 KW High Rate of Development Assumed -i,-4-1-0 -4-3-e 1,820 3,330 5-1G- 3,540 350 2000 Annual Energy Million KWH 490 240 ----~~--~---------- ~-{} 23,340 1,680 3-;-9-9-e 24,810 2,450 -<!& 130 40 -5-e 70 30 6-3-0 910 280 -35{) 490 210 Mid Range of Development Assumed -t.l:t) l-;-4-1<) -431) -3fol{) 260 1,820 . 760 5,330 -··,---·· 70 .. _____ 4_99 ______ ?.~_0 ____ 1 _~_680 -- Low Development Assumed · -9& ~ -ti-E} 1-;-4-!7-0 130 910 260 1,820 40 280 70 490 ....... ·--·------·-·-· -------------------·-·----·---------- ' in Southcentral with a peak load of 2.5 million kilowatts. All estimates assume that mlnerals other than petroleum and natural gas could account for the majority of industrial requirements by 2000. The relevant facilites within the railbelt were assumed to be coal mining, the nu- clear fuel enrichment plant, metal mining, metal processing plants, and . metal dredging operations. 1.5. Reassessment Report on Upper Susitna River Hydroelectric Development for the State of Ala ska by Henry J. Kaiser Company, 1974. This study projects net energy for the Anchorage and Fairbanks areas (but not Glennallen-Valdez) through 1990. The forecast is sho'vn in Table 1.7. The forecast used a model which related number of customers in the residential and commercial-industrial sectors to population and employment projections. 1980 energy sales per customer were forecast at 13,000 and 14,000 in Anchorage and Fairbanks in the residential sector and 148,000 and 118,000 in those respective areas for the com- mercial-industrial sector. 1.6. Southcentral Railbelt Area, Alaska: Interim Feasibility Report: Hydroelectric Power and Related Purposes for the Upper Susitna River Basin, Alaska District Corps of Engineers, Department of the Army, 1975. The previous Alaska Power Administration report was updated for this study with no significant modifications in methodology but some changes in assumptions which result in a reduction of the projections. I-13 TABLE 1.7. ~-·o··.;··: ~< :·•: -··· Anchorage-Cook Inlet Fairbanks Total Net Peak Net Peak Net Pea k Energy Load Energy Load Energy Lo a d Year million thousand million thousand million thou s and --kwh kw kwh kw kwh k w Actual 1968 559 . 125 161 43 720 168 1973 1,090 213 318 73 1,408 286 - Forecast 1975 1,400 271 410 87 l, 810 3 58 1980 2,740 513 803 164 3,543 6 77 1985 5,080 928 1,354 266 6,434 1 , 194 1990 9,420 1, 721 2,281 434 11, 701 2 ,1 55 I-14 td In this study, the actual areas that would be served by a railbelt electric faicility were the basis for projections so some small com- ponents of the previously defined Southcentral and Yukon regions are not included. Projections are shown in Table 1.8 and Figure 1.4. The utility estimates are some~11hat lower than earlier projections reflecting this as ~11ell as the analysis of t>-.70 additional years of his- torical data. Beyond 1980, the growth rates are identical to the pre- vious studies. These are as follm11s: high range likely midrange lower range 1980-1990 9% 7% 6% 1990-2000 8% 6% 4% Military load growth is now projected at 1 percent annually rather than 1.7 percent. The most significant change in this study from the previous work is a reduction in the projected industrial load which accounts for the majority of the difference between the total load projected in this and the previous study. The 1980 midrange projection has declined from 1,190,000 ~fiVh to 350,000 MWh. The specific industrial developments assumed in the projections are presented and shown here as Table 1.9. 1-15 TABLE 1.8. Actual Reguirements Estimated Future Reguirements Type of Load 1974 1980 1990 2000 Peak •Annual Peak Annual Peak Annual Peak Annual Demand Energy Demand Energy Demand Energy Demand Energy Area 1000 kw Million/kwh 1000 kw Million/kwh 1000 kw Million/kwh 1000 kw Million/kwh . Utilities High Rate of Growth H Anchorage 284 1,305 650 2,850 1,570 6,880 3,430 15,020 I ...... 0'\ Fairbanks 83 330 160 700 380 1,660 800 3,500 --Total 367 1, 635 810 3,550 1,950 8,5 40 4,230 18,520 Likely Mid-Range Growth Anchorage 590 2,580 1,190 5, 210 2,150 9,420 Fairbanks 150 660 290 1, 270 510 2,230 ---Total 7 40 3,240 1,480 6,480 2,660 11,650 . Lower Rate of Growth Anchorage 550 2,410 1,010 4,420 1,500 • 6. 570 Fairbanks 140 610 240 1,050 350 1,530 Total 690 3,020 :1.,250 5,470 1,850 8,100 TABLE 1.8. (cont.) Actual Reguil-ements Estimated Future Reguirements Type of Load 1974 1980 1990 2000 Peak Annual Peak Annual Peak Annual Peak Annual Demand Energy Demand Energy Demand Energy Demand Energy Area 1000 kw Million/kwh 1000 kw Million/kwh 1ooo kw Million/kwh 1000 kw Million/kwh National Defense H Anchorage 33 155 35 170 40 190 45 220 I Fairbanks 41 197 45 220 50 240 55 260 I-' --"--1 Total 74 352 80 390 90 430 100 480 bdustrial High Rate of Development Assum e d Anchorage 10 45 100 710 2,910 20,390 2,920 20,460 Fairbanks l/ Mid-Range Development Assumed Anchorage 50 350 100 710 410 2,870 Fairbanks 1/ -- Low Development Assumed Anchorage" 20 140 50 350 100 710 Fairbanks 1/ ··--------··--···-... ·----· -· TABLE 1.8. (concluded) Actual Reguirements Estimated Future Reguirements Type of Load 1974 1980 1990 2000 Peak Annual Peak Annual Peak Annual Peak Annual Demand Energy Demand Energy Demand Energy Demand Energy Area 1000 kw Million/kwh 1000 kw Million/kwh 1000 kw Million/kwh 1000 kw Million/kwh Combined Utility, National Defense, and Industrial Power Reguirements H . I f--1 Higher Growth Rate 00 Anchorage 327 1,505 785 3,730 4,520 27,460 6,395 . 35,700 Fairbanks 124 527 205 920 430 1,900 855 3,760 ·----Total 451 2,302 990 4,650 4,950 29,360 7,250 39,460 Likely Mid-Range Growth Rate Anchorage 675 3,100 1,330 6,110 2,605 12,510 Fairbanlcs 195 880 340 1,510 565 2,490 Total 870 3,980 1,670 7,620 3,170 151000 Lower Growth Rate Anchorage 605 2, 720 1, 100 4,960 . 1,645 7,500 Fairb ank s 185 830 290 1,290 405 1, 790 Total 790 3 ,55 0 1 ,390 6,250 2,050 9,290 ·-· ... ·------·--- -----. ---------· & -----------:----.-. -------·-···. . --.. .Lota.L H I t--' \.0 f'JU ----·------.. -------. :FIGURE 1.4. ESTIMATED FUTURE POWER ·REQUIREMENTS 1974-2000 2. ANNUAL ENERGY REQUIREMENTS 40.ooor----r---rr----r----r----r====F:::::::=l I i I . . . I ·-·--· ·---···· .. -.. ·---~-.. ---·-·. ·---··:··---,-. ··-·-···-···-·····-··-·: ···--~---·-·· -~·-·--· ----~~-·-···. ·-··· .. -··-···--·-I -··--... -j· . . .. ···-···---·---- --------··· .. ·····+-·--·--·------~---+-------·----· --1----···----· ---·· --·· ....... ·-;.f:o-'?J ----~---·· ··--·----·! .. i : i I ~c} I l I I <?J~ I ! , ~~ : ::r: 10000 -· .. ... .. .... ·····. --, --.. ,. _________ i_, ___ .. __ ...................... .,... __ ,. _______ , .. _ --------·-·----1··---.. -· ~ . ! · : I I I t\00qe ! z . I i . I ~\0' I 0 5000 ---~----·-· -·-t------·--· .. --~-----'--------~--------I ----' . o\e ----1 ---------------L ·· :J 4000 -------------.. --1--------~-.. ------:--·-!-.. ___ .. _____ --, -----\..!)"lei <t.s\1~---------~--j-------------.. --r----·--· -------- ~ 3000 ------------r ------------~,-~-~:~:-=-----,--------------·-r ------------------1-------... -!----------- >-, . I ! ?i 2000 ---· ----.............. !··--. . ............ 1 .. -· .. ----------- 1 -----.... ----.... T -. .. .. i W I Note: Includes estimated annual energy requirements Q ! 1 ! for utility, notional defense, and industrial power · ------r--T 1 -----systems in the Anchorage-Cook Inlet and Fairbanks--- _J I I • <r . 1 · Tanana Volley areas. ! ! => I I ! ·. I . I i : ~ · :~~ _:=:~:-: ~-:--F:·: ____ ~--!T-~= --=---r=-=-~:=_-.:·r=~-~---~-==~r~~~ ~ ~:~--~ =-: -- ---------------1 ---------:--~---------~~ ------· -----+----· -----1"----· r -····· ·- ··:·.·:-··-................ : ---.---------.... -~ ---, ....... _. _____ ·--· .... T-.. ·-----··----.. ·-·--~-...... -· --.. --· -·:---.. -.... .. .i.. .... -.- . I ! I j I l 'I I l , 1 ! 00 A.?. A.-September 1975 I L-----~----~~----~--------~-----~-----~~----~~ 1965 70 75 80 85 90 95 2000 YEAR 1.),2 90 TABLE I. 9. Industrial Capacity in ~·iW Rate of Development Lov Range 1·1id Rane;e High Range ·Year 1980 1990 2000 1980 1990 2000 1980 1990 ~ Anchore5e .Area: Kenai Peninsula: Chemical. Plant Y ll 11. 11 12 14 16 1.3 16 20 LNG PLant l./ .4 .4 .4 .4 .5 .6 .5 .6 ·1 New Plent 1.0 10 10 1.0 10 10 10 10 1/. Refinery -2.2 2.2 2.2 2.2 3 4 3 4 5 Timber Y 2 3 5 3 5 5 5 5 5 other Vicinities: Coal Gasification 10 10 250 10 250 250 Mining and Mineral Processing 5 25 -5 25 50 25 50 50 Nuclear Fuel Enrichment 2500 2 500 Timber 5 7 5 7 7 7 7 1 1-l'e'W' Cit;[ 17 30 10 30 70 30 70 70 -TOTAL (rounded) 20 50 100 50 100 410 100 2910 2920 Fairbank3 Area ~/ Source: 1974 Alaska Power Survey Technical Advisory Committee Report on Economic Analysis and Load Productions, pages 81-89. " !I Existing Installations 2/ Timber _processing and oil refinery loads totaled l e ss than 10 MW. I-20 A population growth rate of 3 percent annually is assumed in the study, resulting in estimates of 410,000 in 1980 and rising to 740,000 in 2000. L.7. Electric Power in Alaska, 1976-1995 and ''Alaska Electric Power Requirements," Alaska Review of Business and Economic Conditions, Institute of Social and Economic Research, University of Alaska, 1976. This study estimated growth in electricity requirements for the entire state based upon a model of the Alaskan economy and detailed assumptions concerning customer growth and average consumption rates. Projections were made based upon two sets of economic assumptions and four sets of electricity use assumptions. This resulted in a signifi- cant range of projections, a reflection of the uncertainty surrounding both the future of economic growth of the state, and electricity use patterns. Table 1.10 shows the range of estimates of sales for utili- ties for. the Anchorage, Southcentral, and Fairbanks regions (a larger region than the railbelt). The economic projections assumed growth rates in population for the state of between 3.8 and 4.8 percent with the growth concentrated in Southcentral Alaska. I-21 Anchorage, Southcentra1, & Fairbanks LOWEST HIGHEST TABLE 1.10. 1976 ISER ELECTRIC POHER REQUIREMENTS PROJECTIONS Average Annual Total Energ;y: Sales Grmvth Rates 1975- 1975-1975- 1974 1985 1995 1980 1985 1995 . 1,468 3,697 8,092 9.1 8 .8 8.5 1,468 7,787 20,984 17.6 16.4 13.5 I-22 Military electricity requirements are assumed to remain constant over time. Self-generated industrial electricity requirements are not spe-· cifically modelled. 1.8. Alaskan Electric Power: An Analysis of Future Requirements and Supply Alternatives for the Railbelt Region, Battelle Pacific Northwest Laboratories for Alaska Division of Energy and Power Development and the Alaska Pmver Authority, 1978. This report did not do a load growth analysis but summarized and interpreted several earlier studies, including those of the Institute of Social and Economic Research and the Alaska Power Administration (APA). Individual load growth studies of several railbelt utilities were also utilized in selecting a narrow projection band from the broader band represented by the earlier analyses. The industrial scenarios developed by the APA were somewhat modified as well. The resulting proj ec.tions for the railbelt are shown in Table 1.11 and Figure 1.5. I. 9. Upper Susitna River Project Pmver Harket Analyses, U.S. Department of Energy, Alaska Po\ver Administration, 1979; Southcentral Railbelt Area, Alaska, Upper Susitna River Basin, Supplemental Feasibility Report by the Corps of Engineers, 1979; and Phase I Technical Memorandum: Electric Power Needs Assessment, Southcentral Alaska Water Resources Committee, 1979. This report is a continuation of the work done by the Alaska Power Administration and again provides load requirements projections for I-23 Year 1974 1980 1990 2000 TABLE I. 11. - I Range of Railbelt Annual ·consumption (Includes use by utili ty : and industrial customers likely to be part of an intertied \ . system. Excludes national defense and non-intertied users.) , Annual Consumption -Compound Annual Gro0th Rate I 1.6 B kHh ! I 2.6 to 3.4 B kl-!h 8.4 to 13. 4 ( 197 4-1980) I i 15.3 (1980-1990) I 8.5 to 10.8 B HJh 9.6 to I ! 16.0 to 22.5 8 kHh 4.0 to 10.2 (1990-2000) ' I-24 y \ Vl :r: :s:: ~ Ll.. 0 Vl z 0 -l -l FIGURE I.5. 10, 000 . -···-·--------------------------,- 1976 ACTUAL ~ / , """f.1975 ACTUAL 1974 ACTUAL lOOG ----·-··----------· --------------------1 -----------------------..,.------1 lQQL-----L-----~-----------L~--------~----~ 1975 1980 1990 2000 Most Likely Range of Utility and Industrial Annual Consumption; Intertied Railbelt Area I-25 utility, military , and industry electricity users. The overall projec- tions are summarized in Table 1.12 and Figure 1.6. Utility requirements are projected on the basis of e x plicit popu- lation estimates (provided by the ISER econometric and demographic model) multiplied b y average annual consumption per capita estimates. The estimate of per capita growth in average annual electricity consumption is based upon growth in this indicator for Alaska utilities during the interval of 1973 to 1977. This is presumed to capture the effect of conservation over the period, and, furthermore, the growth rate is assumed to decline over time to reflect additional conservation measures and saturation of appliances, etc. For the three "consumption per capita" scenarios, the following annual growth rates were assumed: Time Period Scenario High Medium Lmv 1980-1985 4.5% 3.5% 2.5% 1985-1990 3.5 3.0 2 .0 1990-1995 3.0 2.5 1.5 1995-2000 2.5 2.0 1.0 2000-2025 2.0 1.0 0.0 These growth rates were applied beginning in 1980. In the interim , growth rates of net generation were assumed to be 12 percent annually for Anchorage and 10.3 percent for Fairbanks. The high and low estimates l-26 TABLE 1.12. Rail belt Area Energy Forecast (Gl.ffi) 1977 .1980 1990 2000 2025 ··--(Historic) Utility: High 3,410 8,200 16,920 38,020 Mid 2,273 3,155 6,110 10,940 17,770 Low 2,920 4,550 7,070 8,110 National Defense: High 348 384 425 544 Mid 338 338 338 338 338 Low 330 299 270 210 Self-Supplied Industry: High 170 2,100 3,590 8,490 Xid 70 170 630 1,460 3,470 Low 141 370 550 1,310 Total: High 3, 928 10,684 20,935 47,054 Mid 2,681 3,663 7,078 12,738 21,578 Low 3,391 5,219 7,890 '9, 630 Trend @ 1973-77 annual/growth: (3,215) (10,270)-(33,000) (601,000) l-27 ,(f) ··o: ·~ ·o ~ ~I N 00 . FIGURE I. 6 • . I 0 0 1000 ,__--------------------~---:-~~ .. ~ .... ::-:: ..• -:-: .... --:_:-::-.: ... -:-. :-:-: .. 7.:.::7., :: .... :-:":" .. ':":'" ..• :::-. ~--:::::--:-:-1 · )O,'ooo .I .·ao,·aao ·TOTAL :RAILBELT AREA . . :7-910~0 ! ~-.· . , . •. ·. : ' . . _:s_o,ooo ENERGY :FORECA.ST ,-5 0,000 ,, ,__tl ...... ·: .Up p er Su s itna Project ~ower .Market Analysi~ '. 6 .: -• • • : : • ' • • • ' • • • • • I · ' ~ . ' • . :: I 4'L~.·.RRPi - I .,,'30,,090· ... I ~. j • • ,. • • ! I .2o,ooa :- .• ..• :,.•• I;', ' .LOW • • ••• ,~ •• • ..... ."~> •.; ~ •• ,. : ·' • .... ··•· .. ··· .. ,I ·.-. ,j. ··:;·.·.:·.:· for 1980 appear to be based on estimates of growth between 1977 and 1980 which are 29 percent higher and 27 percent lower than the mid case re- spectively. Two population estimates are utilized. These are presented in Table I.l3. TABLE 1.13. POPULATION ESTIMATES (thousands) Year Anchorage-Cook Inlet Fairbanks Statewide Low High Low High Lmv High 1980 239 270 60 62 500 514 1985 261 320 68 77 563 641 1990 299 407 75 95 618 790 1995 353 499 82 114 680 947 2000 424 651 90 140 743 1,158 2025 491 904 99 179 820 1,485 The result of t>.;o population projections and three per capita con- sumption projections is six scenarios. The high/high and lmv/lmv see- narios became the high and low projections, respectively, while the average of the high/low and low/high became the midrange final forecast. Military requirements were projected to remain constant in the mid- range estimate, +1 percent in the high case, and -1 percent in the low case. I-29 The self-supplied industrial load projection is based upon the earlier Battelle report which, in turn, is based upon the earlier APA work. The high range forecast is summarized in Table I.l4. The only difference between the cases is that the midrange does not include the aluminum smelter and the low range contains neither the smelter nor the ne'" capital city. TABLE I.l,4. SELF-SUPPLIED INDUSTRIES FORECAST HIGH RANGE Existing refinery (2.4 MH) Existing LNG plant (. 4 to . 6 H\v) Coal gasification (0 to 250 MH) New city (0 to 30 HH) New refinery (0 to 15.5 MH) New LNG plant (0 to 17 MIV) Mining and mineral plants (5 to 50 M\v) Timber (2 to 12 M\v) Existing chemical plant (22 to 26 M\V) Aluminum smelter or other energy intensive industry (0 to 280 M\v) A comparison of the estimates developed for this report with earlier studies by-APA was done and is presented as Table I.l5. It indicates a slight downward shift in the projections. Finally, a compilation and extrapolation of forecasts done by the utilities themselves was compared to the study results. As Table I .l6 shows, the aggregated utility forecasts are considerably above those of the study. I-30 TABLE 1.15. CONPARISON OF UTILITY ENERGY ESTUIATES 1976 MARKETABILITY REPORT, UPDATE OF 1976, AND 1978 ANALYSIS Upper Susitna Project Power Market Analysis Anchorage-Cook Inlet Fairbanks-Tanana Valley Total Railbelt •I Forecast· 1976 Update 1978 1976 Update 1978 1976 Update 197 8 I ~ Range .··r Re~ort of 1976 Forecast ~eeort of 1976 For e cast Report of 1976 Forecast . Year 1974 Historic 1,305 1./ 1,189.7 ll 330 353.8 1,635 1,543.5 1975 High 1, '•89 377 1,866 ·· Mid 1,467 371 1,838 Low 1,450 367 1, 816 Historic 1,413.0 450.8 1,863.8 1976 High 1,699 430 2,129 H Hid 1,649 lfl7 2,066 I UJ Low 1, 611 407 2,018 f-' Historic 1,615.3 \ 468.5 2,083.8 \.; ' 1977 High 1,939 t.9o 2,429 Mid '· 1,853 469 2,322 Lm-1 1,790 453 2,242 Hi s toric 1,790.1 1,790.1 482.9 482.9 2,273.0 2,273.0 1980 High 2,850 2,660 2, 720 700 720 690 3,550 3,380 3,410 Mid 2,580 2,540 2,500 660 690 655 3,240 3,230 3, 155 Low 2,410 2,460 2,300 610 660 620 3 ,020 3,1 2 0 ·2, no 1990 Hi gh 6,880 6,300 6,630 1,660 1,700 1,570 8,540 8,000 8,200 Mid 5,210 5,000 4,880 ; 1, 270 1,360 1,230 6,480 6,360 6,110 Lo\-1 4,'•20 4,410 3,590 1,050 1,180 960 5,470 5,590 4,550 . -)· .. -~ooo·-..-nigh 15,020 13,600 13,920 3,500 3,670 3,000 18,520 17,270 16,920 Mid 9,LI20 8,950 8,960 2,230 2,440 1,980 11,650 11,390 10,940 LoH 6,570 6,530 5, 770 1,530 1,750 1,300 8,100 8,280 7,070 lJ 1974 historic data revised between 1975 and 1978. APA. 11/78 -----------·-·------------···------------~-------------------------------------------- TABLE 1.16. Utility Forecasts 1978 Susitna Forecasts Energy (G'i·TH) High ~fid Low 1980 3,344 3,410 3,155 2, 920 1985 6, 277 5,460 4,455 3,630 1990 10,965 8,200 6,110 4,550 1995 17,748 11,600 8,140 5,690 2000 26,550 16,920 10,940 7,070 Peak (H'i-1) 1980 725 778 720 667 1985 1,377 1,244 1,021 830 1990 2,986 1,873 1,396 1,039 1995 3,835 2,645 1,858 1,293 2000 5,641 3,865 2,497 1,617 I-32 Glennallen-Valdez projections were made using the 1976 Copper Valley Electric Association Power Requirements Study as a base. Total energy was projected to be 74,000 HWh in 1980, 134,000 11'\ffi in 1990, and 240,000 Mlffi in 2000. I.lO. Differences Between Present and Previous Studies I.lO .A. ECONOHIC PROJECTIONS Differences exist in the economic projections among the various studies because of both different predictions about what large scale projects will be undertaken within the state and when they will occur (the gas pipeline construction is -an example) and different assumptions apout how the support sector of the economy and the population responds to economic development. In both these areas, the present study has some benefit of hind- sight which earlier studies have not. The 1979 APA study, for example, utilized a 1980 statewide population estimate ranging between 500,000 and 514,000. Present projections place the number closer to 420,000. Part of the error was the result of overly optimistic projections of large project activity which have not yet materialized but which to a large extent the present study expects to occur in the early 1980s. A second component of error ~vas the fact that the entire cyclical pat.tern of economic behavior in response to the construction of the oil pipeline was not captured in the economic projection technique. I-33 Only the data from the expansion portion of the cycle was internalized into the model but not the contraction portion of the cycle. As a result, the economic model itself contained some upward bias . I.lO.B. ELECTRICITY CONSUMPTION PROJECTIONS The recent APA study as well as the earlier ISER study developed projections based upon the product of population and consumption per capita. The present study has generally lower growth rate assumptions for per capita consumption based upon explicit estimates of saturation, use patterns, and conservation measures. Another distinguishing characteristic of the present study is that conservation measures re- duce consumption growth rates until they have been completely implemented. Subsequently, growth may actually accelerate. The 1979 APA study assumed that the reduction in the rate of growth of electricity consumption after 1973 was the result of the implementation of conservation efforts grow- ing out of the 1973 oil embargo and higher energy prices. This study conclu~es that the reduction in the growth rate was not conservation- induced because, particularly in Anchorage, there was no, and even today is not, price incentive for the conservation of electricity. I.lO. C. SELF-SUPPLIED INDUSTRIAL PROJECTIONS Earlier studies done by APA and Battelle all trace their projections of self-supplied industrial electricity consumption to a list of pro- jects compiled in the early 1970s in a State of Alaska Department of Commerce and Economic Development, Division of Economic Enterprise pub- lication entitled "Power Demand Estimates, Summary and Assumptions for I-34 the Alaska Situation." Rather than a projection, the report is simply a list of potential projects with their related energy requirements. Such a "list" is not felt to be appropriate as the basis for sound electric pm.;er requirements planning. I.lO.D. DEFINITIONS The present study projects total sales of energy to final consumers. This is a smaller quantity than net energy for system \vhich is the con- cept used in some earlier studies. The difference is transmission and distribution losses and energy unaccounted for. I-35 APPENDIX J BIBLIOGRAPHY Alaska Department of Commerce and Economic Development, Division of Economic Enterprise. The Alaska Economic Information and Reporting System. Quarterly Report. January 1980; April 1980. "The Alaska Economy, Year-End Performance Report 1978." Vol. 7. "Numbers: Basic Economic Statistics of Alaska Census Division." Novemb2r 1979. Alaska Department of Commerce and Economic Development, Division of Energy and Power Development. "1979 Community Energy Survey." Alaska Department of Community and Regional Affairs. "Selected 1970 Census Data for Alaska Communities." Part V Southcentral Alaska. March 1974. Alaska Department of Labor. "Alaska's Economic Outlook to 1985." Juneau, Alaska. July 1978. Statistical Quarterly. Second Quarter 1979. Alaska Department of Labor, Employment Security Division. Alaska Labor Force Estimates by_ Industry and Area 1974. Juneau, Alaska. October 1975. Alaska 1970 Census Atlas: Population by Election Districts. Juneau~ Alaska. n.d. Alaska Department of Revenue. "Fuel, Conservation, and Individual Credits Relative to the Individual Income Tax." Juneau, Alaska. January 1980. "Revenue Sources FY-1979-81." Quarterly Update. March 1980. Alaska Department of Revenue, Petroleum Revenue Division. Petroleum Production Revenue Forecast. Quarterly Report. March 1980. Alaska Division of Planning and Research. Alaska Statewide Housing Study 1971. Vol. II. December 1971. "Alaska Power Survey." 1969 and 1976 by Federal Power Commission. 1974 by Alaska Power Administration. Alaska Public Interest Research Group. Funding Proposals for Conservation/ Renewable Energy Research, Development, Demonstration, Education and Outreach. Draft. March 12, 1980. Alaska Public Utilities Commission. Annual Reports to the Legislature. Alaska State Legislature, Budget and Audit Committee. Energy Related Budget Items. March 12, 1980. American Gas Association, Department of Statistics. 1972 Gas Facts. Arlington, Virginia. 1973. "Anchorage Real Estate Research Report." Volume III. Anchorage Real Estate Research Committee. Anchorage. Fall 1979. Andreassen, Arthur. "Changing Patterns of Demand: BLS Projections to 1990;11 Monthly Labor Revietv. December 1978. Applied Economics Associates, Inc.; R & M Consultants, Inc.; and Science Applications, Inc. "The Role of Electric Potver in the Southeast Alaska Energy Economy." Prepared for the Alaska Pmver Administration. Juneau, Alaska. March 1979. Arlon R. Tussing and Associates, Inc. "Background Information on the Electric Power Industry in Alaska and Its Regulation with a History of the Proposed Susitna River Hydroelectric Project." For the Alaska State Legislature, House of Representatives, Power Alternatives Study Committee. For Committee Use Only. December 31, 1979. "Susitna Hydropmver: A Review of the Issues." Prepared for the Alaska State Legislature. April 15, 1980. Arthur D. Little, Inc. Alaska's Mineral Resources as a Base for Industrial Development. Report to the State of Alaska, 1962. Potential for Use of Alaska's Energy Resources. Report to State of Alaska, 1962. Baring-Gould, Michael, Marsha EnTin Bennett, Peg Hargis, and Joyce Taylor. "Valdez City Census." University of Alaska. Anchorage. August 1978. Baughman, Martin, and Paul Joskow. "The Effects of Fuel Prices on Residen- tial Appliance Choice in the United States." Land Economics. February 1975. "Energy Consumption and Fuel Choice by Residential and Commercial Consumers in the United States." Energy Systems and Policy, Vol. 1 No. 4. 1976. Baumgartner, Mark, and Charles Backus. "Energy End Use: Alaska's Northern Railbelt ." For the House Pmver Alterhatives Study. Alaska Federation for Community Self Reliance, Inc. February 1980. Baumgartner, Mark, and Samuel Skaggs. "Energy Conservation Potential for Alaska's Railbelt." Draft prepared for Alaska Center for Policy Studies by the Alaska Federation for Community S~lf-Reliance, Inc. Harch 1980. J-2 Bowling, S.A., and C.S. Benson. "Study of the Subarctic Heat Island at Fairbanks, Alaska." Geophy sical Institute. Fairbanks, Alaska. June 1978. Bracken, E.O. "Power Demand Estimators, Summary and Assumptions for the Alaska Situation." Alaska Department of Economic Development. April 1973. Breeding, L.E. State of Alaska Energy Conservation Program: Phase III Evaluation. Anchorage, Alaska. December 15, 1976. Brown, R.V., A.S. Kahr, and C. Peterson. Decision Analysis for the Manager. Ne\v York: Holt, Rinehart and Winston. 1974. CH2M-Hill. "First Phase of Feasibility Assessment for Hydroelectric Development at Grant and Crescent Lakes." Anchorage, Alaska. October 1979. California Energy Commission, Assessments Division. "California Energy Demand 1978-2000, A Preliminary Assessment." (Staff Draft) Sacramento, CA. August 1979. "Appendix A: Analysis of Residential Energy Uses," 300-76-034, 1976. Carroll, T; Owen, Robert Nathans, Philip F. Palmedo, and Robert Stern. "The Planner's Energy Workbook." Brookhaven National Laboratory and State University of New York Land Use and Energy Utilization Project. Sponsored by the Office of Conservation and Environment of the Federal Energy Administration.· June 1977. Casper, David A. "A Less-Electric Future?" Energy Policy, pp. 191-211. September 1976. Charles River Associates, Inc. Analysis of Household Appliance Choice. June 1979. Prepared for EPRI-EA-110. "Development of a Commercial Sector Energy Use Model for New York State." Prepared for New York State Energy Office, Albany, New York. August 1979. Christian, Jennifer H. "Alaska's Oil and Gas Industry, 1959-71." Alaska Review of Business and Economic Conditions; Vol. IX, No. 2. April 1972. Code of Federal Regulations. Title 10, Part 430. Cohn, Steve, Eric Hirst, and Jerry Jackson. "Econometric Analyses of Household Fuel Demands." Oak Ridge National Laboratory. Oak Ridge, Tennessee. March 1977. Conopask, J.V. for Small Economic Service. "Data Pooling Approach to Estimate Employment Multipliers Regional Economies." U.S. Department of Agriculture, Development Division, Economics, Statistics, and Cooperative 1978. J-3 Conway, Christopher J. "The Impact of Conservation and Interfuel Sub- stitution on Electrical Load Growth 1974-1993." Presented to the Select Committee on Hydro Affairs. Toronto, Ontario. February 1979. Cro-v1, Robert E., James H. Hars, and Christopher J. Conway. "A Pre- liminary Evaluation of the I.S.E.R. Electricity Demand Forecast." -Working Paper #1. Toronto, Canada. deLeemv, Frank. Institute. The Distribution of Housing Services. Washington, D.C. 1974 . The Urban Demand and Conservation Panel of the Committee on Nuclear and Alterna- tive Energy Systems. "U.S. Energy Demand: Some Low Energy Futures." Science, pp. 142-152. April 1978. Dolezal, Patricia L., and Richard L. Ender. Municipality of Anchorage." Anchorage Anchorage, Alaska. September 1976. "1976 Population Profile: Urban Observatory Program. Edison Electric Institute. Economic Growth in the Future. New York: McGr mv-Hill. Electric Power Research Institute. "How Electric Utilities Forecast: EPRI Symposium Proceedings," R. T. Crm.; (ed.) EPRI EA-1035-SR. March 1978. Ender, Richard L. "The Opinions of the Anchorage Citizen on Local Public Policy Issues." Anchorage Urban Observatory. December 1977. England, Gerald M. "Cogeneration at the University of Alaska Heating Plant." The Northern Engineer. Summer 1978. Erickson, Gregg. "The Natural Gas Industry in Alaska." Alaska Review of Business and Economic Conditions. February 1967. Fairbanks North Star Borough, Community Information Center. "1978 Fairbanks Energy Inventory." Special Report No. 4. Fairbanks, Alaska. July 1979. Community Information Quarterly. Fairbanks, Alaska. Winter 1979. Community Information Quarterly. Fairbanks, Alaska. Spring 1980. Fairbanks North Star Borough, Impact Information Center Final Report. Chapter 7: ENERGY. Fairbanks, Alaska. 1977. Federal Energy Administration. "Energy Consumption in Corrnnercial Industries by Census Division--1974." National Energy Information Center. Harch 1977. J-4 "Residential and Commercial Energy Use Patterns, 1970-1990." Project Independence Blueprint. Final Task Force Report. Vol. 1, November 1974. Federal Energy Administration, Energy Research and Development Adminis- tration. "Tips for Energy Savers." August 1977. Federal Energy Administration, National Energy Information Center. Energy Consumption in Commercial Industries by Census Division-- 1974 .. Washington, D.C. March 1977. Federal Po\,er Commission. Form 12 annual reports of utilities, Annual. Federal Register. Vol. 42, p. 36648, July 15, 1977. "Energy Conservation Program for Appliance." Flaim, Paul 0., and Howard N. Fullerton. "Labor Force Projections to 1990: Three Possible Paths." Honthly Labor Review. December 1978. Fryer : Pressley : Elliott, Inc. "Alternative Energy Resources of Alaska's Railbelt." Prepared for Alaska Center for Policy Studies. April 1980. Goldsmith, Scott. "Alaska Electric Power Requirements: A Revie\.;r and Projection." Alaska Review of Business and Economic Conditions, Vol. XIV, No. 2. June 1977. "Alaska's Revenue Forecasts and Expenditure Options." Alaska Review of Social and Economic Conditions, Vol. XV, No. 2. June 1978. Man-In-The-Arctic-Program Model Documentation. Institute of Social and Economic Research. 1979. Goldsmith, Scott, and Lee Huskey. An Economic Impact Analysis." lature by Institute of Social Alaska. April 1978. "The Alpetco Petrochemical Proposal: A report to the Alaska State Legis- and Economi :: Research. Anchorage, "Structural Change in the Alaskan Economy: the Alyeska Experience." Paper presented at the 29th Alaska Science Con- ference. August 15, 1978. Goldsmith, Scott, and Tom Lane. Oil and Gas Consumption in Alaska 1976-2000. For the Alaska Royalty Oil and Gas Development Advisory Board and the 1978 Alaska State Legislature by the Institute of Social and Economic Research. Anchorage, Alaska. January 1978. J-5 Goldsmith, Scott, and Kristina O'Connor. "Historic and Projected Oil and Gas Consumption." For the Alaska Royalty Oil and Gas Develop- ment Advisory Board. January 1980. Halvorsen, Robert. Econometric Models of U.S. Energy Demand. Lexington, Massachusetts: Lexington Books. 1978. Hausman, Jerry A. "Individual Discount Rates and the Purchase and Utilization of Energy-Using Durables." Bell Journal of Economics. Spring 1979. Heberlein, Thomas Electricity and Policy. A. "Conservation Information: The Energy Crisis and Consumption in an Apartment Complex." Energy Systems 1975, Vol. 1, No. 2. Henry J. Kaiser Company. Upper Susitna River Hydroelectric Development. Report to the State of Alaska, 1974. Hirst, Eric, and Janet Carney. "The ORNL Residential Energy-Use Model: Structure and Results." Land Economics. August 1979. "Residential Energy Conservation." Energy Policy. September 1977. Hirst, Eric, and Bruce Hannon. "Effects of Energy Conservation in Residential and Commercial Buildings." Science, pp. 656-661. August 17, 1979. Hirst, Eric, William Linn, and Jane Cope. "A Residential Energy Use Model Sensitive to Demographic, Economic, and Technological-Factors." Quarterly Review of Economics and Business. Summer 1977. Haag, Rush. The Promise of Power. Potential Economic Development in Southeast Alaska. Alaska State Yukon-TAIYA Commission and Alaska Department of Economic Development. Juneau, Alaska. January 1972. "Housing Units Authorized by Building Permits." Compiled by Department of Housing and Urban Development. Hitchens, Diddy, Richard L. Ender, and G. Hayden Green. Anchorage Municipal Housing Study. Anchorage Urban Observatory Program. Anchorage, Alaska. 1976. Hitchens, Diddy R., Richard L. Ender, G. Hayden Green, and Marsha Bennett. "A Profile of Five Kenai Peninsula Tmvns." Bureau Hanagement and Urban Affairs and Anchorage Urban Observatory. of 1977. Ide, Ed~vard A. "Estimating Land and Floor Area Implicit in Employment Projections.'' Prepared for U.S. Department of Transportation. July 1970. J-6 Jackson, Jerry R. "An Economic-Engineering Analysis of Federal Energy Conservation Programs in the Commercial Sector." Oak Ridge National Laboratory, Oak Ridge, Tennessee . January 1979. Jackson, Jerry R., Steve Cohn, Jane Cope, and WilliamS. Johnson. "The Commercial Demand for Energy: A Disaggregated Approach." Oak Ridge National Laboratory. Oak Ridge, Tennessee. April 1978. Jackson, Jerry R., and WilliamS. Johnson. "Commercial Energy Use: A Disaggregation by Fuel, Building Type, and End Use." Oak Ridge National Laboratory. Oak Ridge, Tennessee. February 1978. Kirkwood, Craig and Sioshansi, R. Perry. "Review of ISER Draft Report." Woodward-Clyde Consultants. April 1980. Kreinheder, Jack. Memorandum to Representative Bill Miles regarding Pacific LNG Project, Research Request No. 30. House Research Agency, Juneau, Alaska. February 29, 1980. Kruse, Jack. "Fairbanks Community Survey." Economic Research. Fairbanks, Alaska. Institute of Social and 1977. "Fairbanks Petrochemical Study." Community Information Center. Fairbanks, Alaska. March 31, 1978. "Research Notes: Fairbanks Community Survey." Institute of Social and Economic Research. Fairbanks, Alaska. December 1976. Lane, Ted. Lower Cook Inlet Petroleum Development Scenario: Economic and Demographic Impact Analysis. Alaska OCS Studies Program. 1980. Lin, William, Eric Hirst, and Steve Cohn. "Fuel Choices in the House- hold Sector." Oak Ridge National Laboratory. Oak Ridge, Tennessee. October 1976. Loll, Leo M. A Study of Technical and Economic Problems--State of Alaska. For Alaska Department of Economic Development by the Institute of Social, Economic and Government Research. Fairbanks, Alaska. August 1967. Love, James, and Eric Meyers. "Draft Legislation prepared for the Alaska Center for Policy Studies." January 20, 1980. "Matanuska-Susitna Borough 1976 Census." Matanuska-Susitna Borough. Palmer, Alaska. December 1976. Midwest Research Institute. "Patterns of Energy Use by Electrical Appli- ances," Final Report. Prepared for EPRI, EA-682. January 1979. Miller, Kent, and Scott Goldsmith. Energy Consumption in Alaska: Estimate and Forecast. Prepared for Alaska Department of Commerce and Economic Development, Division of Energy and Power by the Institute of Social and Economic Research. Anchorage, Alaska. January 1977. J-7 New England Power Pool (NEPOOL) and Battle Columbus Laboratories. "Mod- els for Long Range Forecasting of Electric Energy and Demand." June 30, 1977. (Available from Ne,.;r England Power Planning, 175 Brush Hill Avenue, West Springfield, MA 01089; revised and updated version forthcoming.) Norgard, Jorgen. "Improved Efficiency in Domestic Electricity Use." Energy Policy, pp. 43-56. March 1979. Olsen, Marvin, et al. Beluga Coal Field Development: Social Effects and Management Alternatives. Prepared for Alaska Division of Energy and Power Development and U.S. Department of Energy. May 1979. Parishi, Anthony J; "Electricity Use No Longer Soaring: Nation Expected to Benefit in SO's." New York Times, 6 April 1980. Personick, Valerie A. "Industry Output and Employment: BLS Projections to 1990." Monthly Labor Review. April 1979. Porter, Edward. Alaska OCS Socioeconomic Studies Program. Bering- .Norton Statewide and Regional Economic and Demographic Systems Impact Analysis. Institute of Social and Economic Research. Anchorage, Alaska. January 1980. "The Economic Impact of Federal Energy Development on the State of Alaska." Harvard University. March 1, 1977. Reaume, David M. Speech drafted for Chuck \.Jebber, Commission · of Alaska Department of Commerce and Economic Development. February 1980. Robert W. Retherford Associates. "Management.Audit: Electrical System Performance-Engineering Aspects Evaluation." Matanuska Electric Association, Inc. Palmer, Alaska. September 1979. "Power Cost Study 1979-1993." Copper Valley Electric Association, Inc. Anchorage, Alaska. Harch 1979. Rogers, George W. "Analysis of the Alaska Economy and Its Future Outlook." Financing Alaska's Employment Security Program, Vol. 2. Alaska Employment Security Commission. Juneau, Alaska. October 1958. The Future of Alaska: Economic Consequences of Statehood. Baltimore: The John Hopkins Press. 1962. Rural Electrification Association. "Power Requirements 2 Hatanuska, Matanuska Electric Association, Inc. Hay .1978. Study." Alaska Palmer, Alaska. "Power Requirements Study." Association, Inc. Homer, Alaska. Alaska 5 Kenai, Homer Electric Hay 1978. J-8 "Power Requirements Study." Alaska 8 Chugach, Chugach Electric Association, Inc. Anchorage, Alaska. May 1976. "Power Requirements Study." Valley Electric Association, Inc. Alaska 18 Copper Valley, Copper Glennallen, Alaska. March 1979. "Pmver Requirements Study 1980-1985." Alaska 6 Golden Valley, Golden Valley Electric Association, Inc. Fairbanks, Alaska. Saunders, Norman C. "The U.S. Economy to 1990: Two Projections for Growth." Monthly Labor Revietv. December 19 7 8. Schlaifer, R. Analysis of Decisions Under Uncertainty. Netv York: McGraw-Hill. 1969. Scott, Michael J. "Prospects for Bottomfish Industry in Alaska." Alaska Review of Social and Economic Conditions. April 1980. Southcentral Alaska's Economy and Population, 1965-2025: A Base Study and Projection. Report of the Economics Task Force, Southcentral Alaska Water Resources Study (Level B) to the Alaska Water Study Committee by the Institute of Social and Economic Research. Anchorage, Alaska. January 31, 1979. "Southcentral Alaska Water Resources Study (Level B)." Phase I. Technical Memorandum. Electric Power. Electric Power 1-lork Plan Committee and Alaska Hater Study Committee. March 27, 1979. Spetzler, CarlS., and Carl-Axel S. Stael Von Holstein. "Probability Encoding in Decision Analysis." Management Science. November 1975. Spurr, Stephen, et al. Rampart Dam and the Economic Development of Alaska. School of Natural Resources, University of Michigan, Ann Arbor. 1966. "Status Report, House Power Alternatives Study Committee." Alaska State Legislature. January 30, 1980. Sweeney, Mark P., Joseph E. Feline, Katherine A. Miller, and Robert E. Fullen. "Natural Gas Demand and Supply to the Year 2000 in the Cook Inlet Basin of South-Central Alaska." Prepared for Pacific Alaska LNG Company, Los Angeles, California. November 1977. Swift, W.H. et al. Alaskan Electric Power: An Analysis of Future Re- quirements and Supply Alternatives for the Railbelt Region. For Alaska Department of Commerce and Economic Development, Division of Energy and Power Development, and the Alaska Power Authority by Battelle Pacific Northwest Laboratories, Richland, Washington. Harch 1978. J-9 Alaskan North Slope Ro y alty Natural Gas: An Analysis of Needs and Opportunities for In-State Use. For Alaska Department of Commerce and Economic Development, Division of Energy and Power Development by Battelle-North~vest, Richland, Washington. September 1977. Taylor, L.D. "The Demand for Electricity: a Survey." Bell Journal of Economics. Spring 1975. Tuck, Bradford. "A Review of Electric Power Demand Forecasts and Suggestions for Improving Future Forecasts.'' Draft Sections. Prepared for The House Power Alternatives Study Committee. February 1980. U.S. Department of Agriculture, Forest Service. _E_I_S ____ v~s~·-=B~o~r~a=x~M=i=n=i~n~g~ Access Road for Quartz Hill Proposal. 1977. U.S. Department of the Army, Corps of Engineers, Alaska District. South- central Railbelt Area, Alaska. Interim Feasibility Report. Hydro- electric Power and Related Purposes for the Upper Susitna River Basin. December 12, 1975. U.S. Department of Commerce, Bureau of Census. Annual Housing Survey: 1977. "General Housing Characteristics, Part A." Current Housing Reports, Series H-150-77. 1979. Annual Survey of Manufactures 1976. Fuels and Electric Energy Consumed: Industry Group and Industries; States by Industry Group; Standard Metropolitan Statistical Areas by Major Industry Group. Census of Housing: 1970 Detailed Housing Characteristics. Final Re p ort HC(l)-B3 Alaska. Washington, D.C.: Government Printing Office. 1971. Census of Housing: 1970. Final Report HC(l)-A3 Alaska. Printing Office. 1971. General Housing Characteristics. \.J'ashington, D.C.: Government Census of Population: 1970 Detailed Characteristics. Final Report PC(l)-D3 Alaska. Washington, D.C.: Government Printing Office. 1972. Census of the Population: 1970 General Social and Economic Characteristics. Final Report PC(l)-C3 Alaska. Washington, D.C.: Government Printing Office. 1971. "Demographic, Social, and Economic Profile of States: Spring 1976.'' Current Population Reports, Series P-20, No. 334. Washington, D.C.: U.S. Government Printing Office. 1979. Housing Vacancies: Fourth Quarter 1979. Current Housing Reports. February 1980. J-10 Projections of the Number of Households and Families: 1979 to 1995. May 1979. Statistical Abstract of the United States. Washington, D.C. Various years. U.S. Department of Commerce, National Oceanic and Atmospheric Adminis- tration, Environmental Data Service. "Local Climatological Data." Asheville, North Carolina. 1978. Various issues. U.S. Department of Energy. Information Kit: The National Energy Act. Office of Public Affairs, Washington, D.C. November 1979. "A New Start: The National Energy Act." Washington, D. C.: Government Printing Office. 1979. U.S. Department of Energy, Alaska Pmver Administration. "Potential for Residential Heating Energy Conservation and Rene\vable Resource Utilization In Southeast Alaska." Mimeograph. January 1980. Upper Susitna River Project Power Market Analysis. March 1979. U.S. Department of Energy, Energy Information Administration. End Use Energy Consumption Data Base: Series 1 Tables. Washington, D.C. June 1978. U.S. Department of Interior. Alaska Natural Resources and the Rampart Project. Washington, D.C.: Government Printing Office. 1967. U.S. Department of Interior, Alaska Power Administration. "Devil Canyon Project, Alaska." Status Report. May 1974. U.S. Department of Interior, Bureau of Land Management. Proposed Five- year OCS Oil and Gas Lease Sale Schedule, March 1980-February 1985. Final Environmental Statement. U.S. Department of Interior, Bureau of Reclamation, Alaska District. "Devil Canyon Project-Alaska." Feasibility Report. May 1960. "Susitna River Basin. A Report on the Potential Development of Water Resources in the Susitna River Basin of Alaska." August 1952. U.S. Department of Interior, Office of Minerals Policy and Research Analysis. "Final Report of the lOS(b) Economic and Policy Analysis." December 15, 1979. U.S. Environmental Protection Agency. Alaska Petrochemical Company, Refining and Petrochemical Facility Valdez, Alaska. Appendix, Vol. II. 1979. J-11 Wellen, Paula H. "Interior Alaska Energy Requirements." Fairbanks North Star Borough, Community Information Center. Fairbanks, Alaska. February 1980. Wertz, Russ. "Housing Survey." Draft. Fairbanks North Star Borough. Fairbanks, Alaska. 1980. J-12