Loading...
HomeMy WebLinkAboutAPA2515OO&fffi(g& 0 §00&®©© Susitna Joint Venture Document Number ~sis_·_ Please Return To DOCUMENT CONTROL I C U T I L i TY S Y S T E M S E N G I N E E R I N G D E P Jl. RT f\/1 E N T ~ ---·-....,..------------·----------~-----""· ---- DESCRIPTIVE HANDBOOK OPTIMIZED GENERATION PLANNING PRO·GRAM FINANCIAL SIMULATION PROGR.AM -·-----..____,....__,...-.·, ---~--........ (PROPRIETARY) GENERAL. ELECTRIC March, 1983 1 ' ,_ } ~ i ! l i 4 ··J' Descriptive Handbook ~"l . ~ f 1 Optimized Generation Planning Program }. 1 Financial Simulation Pr·ogram ' 1· (Proprietary) ! Generation Planning and Economics Elect.'~'•lc Utility Systems Engineering Department General Electric Company Schenectady, New York 12345 l ! ' ' ,. t . r ·c J [ ' ' _,_~ ~-· r [ r: l" [ ll. [. [, .' l. Section --- 1 2 3 4 5 6 7 8 9 10 11 12 13 TABLE OF CONTENTS Topic PREFACE INTRODUCTION • • $ • • • • • • • • • • • • e • • • • • • • • • PREDETERMINED GENERATION o • • • • • • • • • • • • • • • • • • (How existing and committed generating units are represented and how this data can be modified) LOAD MODEL • e • • • • • (How loads are modeled) . . . . . . . . . . . . . ~ . . . . . LOAD MODIFICATIONS • • • • • • • • • (How the Load Model can be modified) • • • • • • • • • • • • • STUDY DATA • • • • • • • • • • • • • .. • • • • (What the expansion study input data includes) DATA PREPARATION • • • • • • • • • • . " . • • (How the data is initialized and modified) ADDITION OPTIMIZATION • • • • • • • • ~ • • • (What the addition optimization process does) RELIABILITY EVALUATION • • • • • • ~ • • • • • (How the reliability measures are calculated) PRODUCTIO~ COSTS • • • G . . . . . . . ~ . . . (How the production cost is determined) • • * • • • • • • • • • • • • • • • • • • • • • • a • • • • • • • • 0 • • • • • E~~IRONMENTAL IMPACTS • • • • • • • • • • • • • • • • e • • • (How the environmental effects are determined) INVESTMENT COSTS • • • • • c • • • • • • ,, • • • • • • • • • • (How the fixed costs are determined) OPTIMIZATION RESULTS • • • • • • • • • • n • • • • • • • • f' 0 (What the addition optimization process has done) EXPANSION OUTPUTS • • • • • • • • • • (What expansion outputs are available) ~ . . . . . . . . . . . i Page 1-1 2-1 3-1 4-1 5-l 6-1 7-1 8-1 9-1 10-1 11-1 12-1 13-1 ~··...,......·----~--~-.. ---------~---··---······--·--·--... · .. -·-·· ·-··-·-·-: -----·-···-· ··-········-· -·--··-····--·· ....... --------.---~--····-----····------··---···v··--......... _ .. _, ..................... ___ ~-·~--~~-----·--·-------,r~-------~ ·-· ,::·· I ll '/! :d ;II H lil il i '1 i [ ; l ! ·1 ·l ! j I \ l .. lt ! l,' Section 14 15 16 TABLE OF CONTENTS (CONTINUED) Topic FINANCIAL DATA • • • • • • • • • • • • • • • • (What is included in the financial study data) • • • • • & • • FINANCIAL SIMULATION • • • • • • e • • (How the financial simulation is done) FINANCIAL RESULTS • • • • • • • • • • (What financial outputs are available) ii • • • • • • • • • • • • • • • • • • • • • • • • Page 14-1 15-1 16-1 I 1 ' I f l ; I, I I ; I ; 1 l I l l 1 l l I ,j ; I [ [ f· ' ~ [ ~-~ I~ [ l l 1 l PREFACE The Optimized Generation Planning (OGP) Program and the Financial Simulation Program (FSP) are being offered to the electric utility industry to assist planners in analyzing alternate patterns of generation additions. The General Electric Company warrants that it has exercised professional competence in writing these programs and in testing them extensively. The Company does not assume responsibility for specific results obtainE:>d from the programs and will not be liable for direct, special or consequential damages arising from decisions based on these results. If a program error is discovered and is reported to the Company within 30 days, the liability of the Company is restricted to the "limitations of liability as stated in the License Agreement or in the Agreement for Computer Services signed with the General Electric Information Services Company (GEISCO). The General Electric Company will not assume liability for incorrect results obtained from incorrect input data. iii l I. ~~jo l! .,. I ;~ v ':t 'i ,. ,. ',; !! j. :1 1 I l ' [ f [, (, I' [ f: I. I [ I_ ,., ~~ -~ I ' •~ I. ' .1' ·~ l INTRODUCTION This handbook is designed to aid users in the application of the Optimized Generation Planning (OGP) Program and Financial Simulation Program (FSP) by providing a general, but comprehensive, explanation of the supporting theory and the composition of OGP and FSP. Additional documentation in various levels of detail is available from the Electric Utility Systems Engineering Department (EUSED)~ This documentation includes a four-page introductory brochure (5204.40A), a twelve-page overview in pamphlet form (GEA-l0390A) and complete User's Manuals, which are continually updated as program enhancements are implemented. For further background on the history of the use of OGP/FSP, a separate Experience List and a compilation of typical studies can also be obtained from EUSED. More detailed information can be obtained on many of the subjects treated in this handbook by ref"erring to the listing of available "Supplementary Information" included at the end of specific sections. These materials can be obtained from EUSED by specific request. This introductory section covers the background and status of OGP /FSP within the context of performing the tasks inherent in the general area of gen·eration system expansion planning. This handbook has been segmented according to the sections illustrated in the schematic flow chart illustrated· in Figure 1-1*. This will allow the reader to more easily locate answer~ to specific functional questions that may arise. Note, however, that the simplified schematic flow chart in Figure 1-1 is not intended to represent the actual computational flow of 'OGP/FSP, but rather to represent the conceptual flow the user is most likely to follow while performing a complete study. Another significant point to be noted is that the user can execute OGP/FSP in total in one integrated ste~, which could involve the entire sequence of 2 through 16 and allow examination of the entire planning spectrum, including the following phases: reliability evaluation, operational simulation, investml3nt costing and financial analysis. Typically, those who use OGP /FSP most often utilize almost all of these automated capabilities and obtain a total system cost evaluation of a thirty-year expansion plan. However, it is perhaps equally important to note that, for certain studies, only portions of the overall analysis may need to be executed, or perhaps a predetermined stream of additions needs to be analyzed. Flexibility is the keynote of OGP/FSP, thus making it a very efficient medium for conducting these types of studies. * Figure 1-1 is printed for easy reference on the ·foldout on the last page of this handbook. Turn to this page now and fold out this art for your reference as you read through the remainder' of this handbook. 1-1 The Spectrum of Generation Planning Activities What is known today as the Electric Utility Systems Engineering Department (EUSED) originated more than half a century ago. Since its origlnation, one of the Department 1 s primary goals has been to foster the development and maintenance of state-of-the-art tools in generation, transmission, distribution, control and automation, to aid planners and decision makers of the electric utility industry with t.heir work. As a result of the effort applied to satisfy this goal, many of the major historical breakthr•oughs in the area of digital computations of syst-em reliability analysis and operational simulation a~1d costing of generation systems have originated from EUSED staff members. Figure 1-2 symbolioally presents a summary of EUSED' s current offerings of computer programs which address all facets of the generation system expansion planning problem. _RELIABILITY ----.1_ PRODUCTION ___j _INVESTMENT _I_ FINANCIAL ____J EVALUATION ~, COST -( COST ---, IMPACT ~ N-AREA LOLP :3 -AREA LOLP 2 -AREA LOLP 1-AREA LOLP N-AREA FUEL MANAGEMENT MONTHLY STANDARD INVESTMENT COSTING ,.------:------·------------------- RELIABILITY OUTAGE PROBABILITY OPTIMIZED GENERATION PLANNING PRODUCTION. COST ENViRONMENTAL IMPACT SIMPLIFIED STANDARD INVESTMENT INVESTMENT COSTING CORPORATE MODELING FINANCIAL SIMULATION FINANCIAL MODELING Figure 1-2. Electric Utility Systems Engineering Department Generation System Planning Programs -' ~ w 0 -' LLJ 0 0 ::!: Because all of the different challenges presented to gen~erat.ion planners today cannot be most efficiently solved with only one tool, EUSED has made a spectrum of models available, ranging, on one hand, from interactive time-sharing programs to the most detailed batch-mode models a~cessible in the 1-2 '·· ~~ II~ ~ ·~ i 'J1 ; ·~ :t l I l j ! ~~ l ! 'I I I. (, ~~ l I ,, ( I~, I . I . I "···' .J I. IJ I., electric utility industry today. As shown in Figure 1-2, OGP/FSP is represented as the program it is--a single, free-standing, integrated tool which spans in one step all four major areas of interest to the generation planner. Users can access OGP/FSP in three ways. Members of the technical staff of EUSED will perform complete studies under the customer's direction, or, with a minimum of assistance by EUSED, users can access OGP/FSP completely from a time-share terminal in their own offices in the "remote batch" mode provided through the General Electric Information Services Company's (GEISCO) MARK III Service. OGP /FSP is also available under a licensing agreement for installation on an in-house computer. It should be noted that, in some respects, OGP/FSP is less detailed than some of the free-standing models shown in the blocks above it in Figure 1-2. OGP/FSP was designed with less detail so it would be easier and less costly to use than the more highly detailed, free-standing models. Highly detailed programs are best suited to short-range studies involving little uncertainty in the data assumptions, while OGP/FSP has been designed for long-range studies which characteristically require many cases involving ranges of parametric assumption. This does not mean, however, that every feature of OGP/FSP is less detailed than those of some of the free-standing models. There are many areas where the level of detail and computational accuracy of OGP/FSP are essentially identical to even the mo~~t detailed programs. This is because the calculational procedures . within OGP/FSP were mostly derived from the algorithms contained within the free-standing programs. No significant deviations in accuracy from these more detailed models were accepted for inclusion in OGP /FSP if it appear·ed that compromise would have rendered OGP/FSP unsuitable for its intended purposes in these areas. Overall Generation Planning and OGP/FSP Regardless of the processes or models generation planners can apply to their work, the elemental driving forces and starting points that feed into any orderly planning analysis must also be considered. For example, before executing the charge from their organizations, planners must deal with many uncertain elements as well as their knowledge of the existing system to determine the impact of these factors upon the plans developed. Specifically, planners must consider the impact of such factors as new generation technologies, the effect of inflation, money market conditions, fuel availability, environmental constraints, etc. Then, planners must use their knowledge of the existing system to develop data estimates and assumptions in areas such as generation, load, operational rules, and economic factors. An organization's charge can be summarized by the following four major questions posed by planners and decision makers of the electric utility industry: l-3 • How much generation needs to be added in the future? • What kind of units will satisfy this need? • What is the total cost of this plan in terms of revenue requirements? • What are the financial implications of pursuing this course of action? In the past, these questions were not that difficult to answer. The "How much?" question could be answered almost by rule of thumb. The "What kind?" question did not involve a multitude of technologies, many of which are characterized by undefined cost and availability. "Total cost" meant something fairly straightforward such as "the sum of fixed cnarges on investment and expenses for fuel and operation and maintenance (O&M)." It did not include any extensive consideration of the costs and performance of pollution controls, the anticipation of operational restrictions due to environmental rulings, oil boycotts, etc. Finally, the "financial implications" factor perhaps would not even have been considered by the system planner because it probably would not have been included in the responsibility and job scope of the system planner. However~ that was before tlle ua~!3 of restricted money markets, tight cash flows and delayed and reduced rate increases. Today, planners and decision makers must have a common basis of discussion. The projected plans developed must not only be attractive in terms of their engineering economics, but they must also be imminently feasible in terms of their technological availability, environmental acceptability and financial practicality. During the last decade, generation planners' needs have expanded significantly, and consequently, their goals have become more difficult to achieve. Because of its foresight, EUSED developed OGP /FSP more than ten years ago to help planners meet their needs. OGP addresses and answers questions posed by the conventional generation planning process including: (1) how much generation needs to be added in the future, (2) what types of units will satisfy this need, and (3) what the total cost of the plan is in terms of revenue requirements. However, OGP also extends far beyond these bounds. The program's optimization capabilities not only can derive a feasible plan, but also can derive the best mix of new generation in terms of minimizing the revenue requirements. OGP also can consider a myriad of environmental restrictions and derive a plan which simultaneously addresses lowest cost and m1n1mum environmental impact. Finally, the financial impacts of any stream of unit additions are presented by FSP to aid the user in further ranking alternative plans. To obtain further perspective and knowledge of the scope and capabilities of OGP, refer to Figure l-3. In a simplified flow chart, Figure 1-3 represents the conventional manually aided iterative process a planner would have to use to derive the total system cost. for one thirty-year expansion plan, using a combination of free-standing computer models. First, a preset list of unit types, along with their sizes· and timing, is tested until the sequence is found that would satisfy the required level of system reliability for each year under study. Then some type of operational simulation is conducted to obtain fuel and O&M ~osts for the system. A stream of investment 1-4 ·~ ,. ' ' I r;;' I I I ' I i ) I: I I' I. I, I. I I I. I " I:, l I ~'< t LIST OF i-UNIT SIZES ----~-..- 8 TYPES RELIABILITY 1,. EVALUATION ~, ---t•..., 30-YR. EXPANSION PLAN I I r--------_ _j I I PRODUCTION COSTING INVESTING CO :sTiNG -- P.W. $ I I I I l P.W. $ I + 30-YR. COST FOR ONE PLAN Figure 1-3. Conventional Generation Planning charges is added to the fuel and O&M charges to determine the annual revenue requirements. Then the cumulative present worth of all revenue requirements is derived to provide one figure of merit or "bottom line" which represents the cost for the one plan. If the impact of other input data on the preferable additions needs to be compared, this sequence would have to be repeated. Iteration could also help planners select a single "optimum'' plan. Even today, the conventional manually aided iterative-process can be a tractable approach for many org:anizations. With the number of alternatives and uncertainties in basic parametric assumptions being so great, however, it is not a very efficient or cctmprehensi ve technique, particularly when one considers that the resources and time available to most .planners are rapidly exhausted under the barrage of remaining questions to be answered in today's 1-5 ;- r'· planning climate. Thus, OGP/FSP was developed not 0nly to allow planners to fulfill their responsibilities, but also to free them to address more topics than previously thought pr-actical. The construction of OGP/FSP, as it exists today, was motivated by the electric utility indu~stry's need to address and accomplish the following goals: • To find the best type of unit that will satisfy future generation system expansion requirements • To be able to study alternative generation system expansion plans • To determine financial and envir.•onmental implications of proposed generation system expansion plans Thus, to help achieve these goalsr the authors imposed the following major objectives to be met in the development of OGP/FSP: • Data and logic simplification • Linking of pt10grams • Optimization of expansions The OGP/FSP Process The OGP /FSP process can be represented in various ways. Sections 2 through 16 of this handbook (illustrated schematically in Figure 1-1) are a segmented treatment of the various facets of OGP/FSP. The information is presented in the handbook in this way merely to facilitate reference and discussion. However, as you read the following overview of the logical structure of OGP/FSP, refer to Figure 1-4, which represents a more simplified breakdown than that shown in Figure 1-1. As you review Figure 1-4, you can see that there are certain points at which the user can furnish input data and there are points where output information can be obtained after appropriate calculations have been executed. In the first block, user-furnished input data describes the given or fixed system, which represents all existing and committed generation (i.e., all in-service, under-construction and pre-planned units that are not likely to change except for scheduled retirement). For more details on this user-furnished input data, see Section 2. Once the user has furnished the necessary generation data, the user must then input an hourly representation of load values for the study period. Supplementary modeling and manipulative capabilities facilitate the automated handling of these hourly loads, particularly when load management studies involving many hourly load changes are being performed. This facet of OGP is described further in Se9tions 3 and 4. 1-6 • , .. r ( . I I I· ,I I' i I, ,I . ' I·· I I I I. I I J I I I I I USER FURNISHED YEARLY OPTIMIZATION USER FURNISHED YEARLY SIMULATION REPRESENT EXISTING GENERATION MODEL FUTURE LOADS PROVIDE INPUT PARAMETERS SATISFY RELJAaiLlTY CRITERIA. SIMULATE SYSTEM OPERATION CALCULATE COSTS AND IMPACTS OUTPUT ACCESS FINANCIAL IN FORMATION CALCULATE FINANCIAL EFFECTS I OUTPUT Figure 1-4. The OGP/FSP Process l GENERATION EXPANSION FINANCIAL IMPACTS Next, as indicated in the second block of Figure 1-4, economic and operational parameters must be supplied by the user. To ease input effort and minimize the possibility of inadvertent errors, an automated process of data by exception, integral to the OGP data assimilation logic, is used. Thus, entire classes of data can be specified and stored for further access by a single number which, for example, c.ould represent a variable that applies to· all of a certain generic type of generating unit. Simple controls direct the l-7 program on how these variables vary according to parameters such as unit size and study year, thus making it unnecessary for the user to perform side calculations to derive unique values for each generating unit. Of course, if atypical values need to be specified for any or all generation, this can also be accomplished. For further information on this subject, refer to Section 5. Section 6 describes how this input data is automatically checked for nominal errors, sorted and then adapted for use within the program. The data is also displayed via formatted outputs for examination and reference by the 'User (see Figure 1-4, Block 3). If some individual items of input data are not readily available to the planner, the technical staff of EUSED usually can, upon request, provide average or suggested types of general industry information to aid in the completion of study data. As discussed in Section 7, OGP then derives a complete expansion plan up to thirty years in length, and documents all the attendant generating unit and system costs associated with it. The user may evaluate a manual or predetermined expansion scenario for any or all of the years under study. Alternatively, if calculated as necessary by the re).iability criterion input [e.g~, either a percentage of installed reserve ot-desired loss-of-load probability (LOLP)], OGP will automatically add enough generation each year t0 satisfy that criterion. The OGP optimization process not only considers and rigorously evaluates the relative economics of each different type of new generating and storage unit shown as available to it for commercial operation, but if the load growth compared to individual unit sizes allows, the program will also identify and compare all reasonable mixtures or combinations of the different types of units. This capability precludes the need for inputting preset limits on the absolute number of each type of unit that is to be evaluated. The OGP optimization logic is not forced to remain within a defined "tunnel" of alternatives. Section 8 explains in detail the computations involved in answering the question of "How much generation is to be added each year'? u If LOLP, either daily or hourly, is the design measure, a very efficient technique is used to convolve the capacity outage table with the load model. Unit maintenance requirements, purchases and sales as well as the known emergency tie capacity to external systems are considered. The effect of load forecasting uncertainty may also be addressed at this point. All of the alternative configurations which satisfy the reliability criterion must then be evaluated in terms of their total system costs and impacts. This begins with the operations analysis discussed in Section g. The simulation strategy is a detailed, six-step process involving the reflection of any contracts that may be specified externally to the system under study, the scheduling of conventional hydroelectric generation and the automatic, unit-by-unit allocation of maintenance (designed to maximize committable reserves in all months). Any energy storage on the sy·stem is also scheduled to minimize cost while observing storage limitations. Then, an hourly based algorithm (not a simple load duration curve approach) commits and dispatches all units through time to simulate an entire year of operation. 1-8 , . J ·. r·J; t :~1 '.l < f I ' j t ·' ' ! l I I I I I I I I I I I I I ,I I I ' . I I ..... : . .. An approximated, or otherwise predetermined, loading order is not used for this simulation because the objective function of minimizing total variable operating costs constantly changes the commitment and dispatch priorities which are based on factors such as individual unit thermal cycling capabilities 1 maintenance schedules, spinning reserve requirements, and fuel and O&M costs. All of these factors can be a function of time. Also, the OGP hourly approach, which recognizes the preceding factors, is of critical importance t-Then the impact of load management or certain new generation technologies are under scrutiny. At this point, the random forced outaging of each unit is simulated and the expected unserved energy calculated. The overall production cost calculation may be based solely on economics or, as described in Section 10, it may also be simultaneously or independently a function of up to seven other environmentally related constraints. Thus, the resultant operational simulation may more accurately, reflect real-world production costs versus idealized values based solely on economicso In addition, absolute energy output or fuel usage constraints on individual thermal units may also be input and considered when calculating production costs. Section 11 documents the resultant calculation of all fixed costs including the major category of revenue requirements due to capital investment based on the fixed-rate charge approach. If present, the demand costs of cont1:•actual agreements are also factored into the fixed costs. The total annual cost of each alternative plan as well as the cumulative present worth of all r~venue requirements is also calculated at this time. This calculation may be all thq.t is needed for most manual addition scenarios, but we have not addressed how the economics of alternative plans are evaluated when the user requests OGF' s optimization capabilities. To meaningfully address this issue, we must slightly retrace our description of the overview of the OGP process and discuss how the program logic anticipates the effects of cost inflation and maturing outage rates. This is accomplished within the program by the use of what has been c~oined the "look-ahead" option. This logic will choose the best selection of units each year based on the lowest total cost alternative that has been derived from future conditions (both in terms of reliability and operations) rather than the lowest current total cost alternative. Section 12 describes the "look-ahead" option in detail. This option calculates capacity needs and production costs, using surrogate values rather than the current decision year's outage rates and economic parameters. The surrogate values are calculated by OGP in response to a single user-supplied input, which is the specification of the "number of years to look ahead." For example, a ten-year "look-ahead" study would cause mature (instead of immature) planned and forced outage rates to be substituted for all alternate or trial units vbeing considered by the comparative decision logic. This results in a more meaningful measure of the reliability worth of new generation to the system. In addition, the complete hourly production cost simulation is calculated with ten-year levelized equivalent fuel and O&M costs for all units . based on the previously input decision-year costs, inflation rates and present worth factor. 1-9 . c-----------------'· J I l ' f ! i • Thus, the OGP logic accumulates a set of all alternative annual scenarios calculated with "look-ahead" values, and is then ready to make a logic.al selection of the "best" or optimum new generation to be added in the decision year in question. OGP then selects the lmvest total cost scenario from among all the alternatives it evaluated. It also performs another j_mportant function by permanently adding the units to the existing system and recalculating all reliability, operations and investment data in order to properly document each figure of merit for the year in question. The next step is to enter the· beginning of the next year under study with the new total system and repeat the process just described. Thus, after this process is followed for thirty years, the study is complete and self contained. t It should be noted that an important outcome of the OGP optimization approach is a lack of end effects. For example, if two cases are started in the same study year and have no differences in input data except that the first OGP study is terminated after ten years and the second after twenty, both cases will yield exactly the same results for the first ten years. Section 13 briefly reviews the output reports available from OGP. These reports contain an abundance of planning information, and are presented in a highly organized, easy-to-use format. Many user-requested outputs have been added during the past years to enhance understandability and minimize additional calculations and tabulation by the user. Detailed results are available for the optimum plan as well as for others which were tried, but rejected. The output can also be saved on a file and later accessed by the user for reformatting or plotting. The Financial Simulation Program (FSP) can be used to evaluate the financial impact of the expansion plan developed by OGP. Much of the data describing the generation system and annual costs associated with the OGP expansion plan can be automatically transferred from OGP to FSP. Alternatively, FSP can be run independently of OGP by using a separately available program to input the necessary data to FSP. FSP is a simplified corporate model which combines a description of initial fina.ncial conditions with a generation expansion scenar:l.·':> in the future to project the annual balance sheets, income statements, cash reports and many other key financial quantities for the system under study. The model is used chiefly to determine the relative desirability of. alternative plans on a measurement basis which goes beyond the "bottom line'' of engineering economics' revenue requirements approach. Section 15 describes the internal processes of FSP. Like OGP, it is a highly automated model that can make the required decisions involving key financial driving forces in an unstructured environment. Use of FSP does not require the user to have knowledge of items such as future rate changes, the timing of new issues, etc. While the generation plant is treated in most detail (the annual expenditures for every new generating unit are separately tracked), other plant, such as transmission and distribution (T&D) and a second business, like gas or steam, are also included as aggregates so the firm can be treated on a consolidated and complete basis. It should also be 1-10 ,( ' . ,, il I I I I I I I I ~I I I I noted that FSP has been mainly written to accommodate the structure and policies of privately financed utilities. Section 16 notes that the variety and level of detail of the outputs available from FSP can best be surveyed by examining an actual case study. In this portion of the introductory material we have discussed the highlights of OGP /FSP based upon the graphic representation shown in Figure 1-4. However, the OGP/FSP model is not useful to planners or decision makers unless it is gainfully applied. The Application of OGP/FSP Before proceeding through the remainder of this handbook, you should be familiar with the numerous applications for wh~ch OGP/FSP is suitable. Table 1-1 represents a descriptive sampling of somj of the actual applications for which users of OGP/FSP have found the programs suitable. Except for the first few items near the top, the actual uses in Table 1-1 are not listed according to frequency or importance. In fact, the most important application might be the next one for· which a planner intends to use the programs. It should also be noted that Table 1-1 was not assembled to suggest applications for which OGP /FSP is best suited, because again, that is an extremely subjeoti ve and variable measurement. TABLE 1-1 OGP/FSP Applications • Optimum Generation Mix • Parametric Sensitivity Tests • Joint Ownership • Long-Range Fuel Supply • Economic Justification • Peaking Capacity Needs • Unit Slippages • System Reliability Design Level • Emission Levels • Company Planning in a Pool • Impact of Forced Outage Rates • Unit Size • Cash Flow Impacts • Purchase/Sale Contracts • Impact of Non-Optimum Additions • Breakeven Costs for Advanced Technologies • Effect of Nucluar Moratorium G Loa,d Management Impacts • New· Financing Requirements EUSED would be the last to ever conjecture or recommend that OGP /FSP can be successfully applied to address all questions that may arise in the realm of generation system expansion planning. Any real test or prediction of the ?.pplic!ability of OGP /FSP to provide useful computed information on a given subject in the broad planning area depends on many physical variables such as the utility's size, present and contemplated generation, complexity of interchange agreements, level of detail of output r:1quired, etc. 1-ll ! f 'j ·~ ' ! ~ l ! j ·l J I 1 t"+ . l \ ! ! 1 j I ' ' i l I Perhaps it is more meaningful to note that, to date, approximately eighty organizations have successfully applied OGP /FSP, and the majority of them continue to do so repeatedly. If there is any question about the appropriateness of a contemplated application, the best policy to follow is to contact the responsible technical staff member within .EUSED and describe the projected study. .EUSED is always prepared to assist users with all phases of their OGP/FSP work. The remainder of the information presented in this Descriptive Handbook will focus on the OGP/FSP process segmentation topics illustrated in Figure 1-1. The discussion, which is organized according to the numerical designations corresponding to the process segments in Figure 1-1, covers the topics listed in Table 1-2. TABLE 1-2 Topic or Information Discussed Section Noo 2 Representing Generation 3&4 Representing Loads 5&6 Parametric Input Data 7 Hhat OGP Does 8 Determine "How Much" Generation to Add 9&10 Determine Oper~ting Costs and Environmental Impacts 11 Calculate Fixed Costs 12 What Has Been Accomplished With OGP 13 Available OGP Output Information 14 Additional Input Data fo:' Financial Analysis 15 What FSP Does 16 Available FSP Output Information 1-12 I I I I II j II I .... . , ·I" l ; f ll l ! I II'. ·'I "i I .! \ I \f I! I (i! >:) I I I I I I I , .... ; ... I I INTRODUCTION SUPPLEMENTARY INFORMATION 1. Electric Utility Planning Models, L.L. Garver, 1975 ORSA/TIMS Meeting. 2. Overvie~-1 of Electric Utility Generation Planning Methods, R.W. Moisan, 1977 National Conference of Regulatory Utility Commission Engineers. 3. Overall General Planning Study, R.W. Moisan, 1974 GE Memorandum. 4. Perspectives on the Design and Application of Generation Planning Programs, W.D. Marsh, R.W. Moisan and T.C. Murrell, 1974 Nuclear Utilities Methods Symposium • 5.. General Electric's OGP Program: The Practical Approach to Generation Planning, R.P. Felak and J.E. Lapsa, 1978 EEI Engineering Computer Forum. 6. OGP and FSP Program Descriptions, February, 1982 GE Memorandum. 1. The Application of OGP to Worldwide Energy Issues, D .L. Dees, B.W. Erickson, R.P. Felak, G.E. Haringa and H-G. Stoll, 1981 Conference on Electric Generating System Expansion Analysis. 1··13 I I I ( I l I~ t I r J 1 I ! I I 1 l l 1 .t I f I ! l I ; f l ) ! l t I $ ! '{ I I I I I L .. I . I I I I PREDETERMINED GENERATION There are two types of data input available to the user to represent present and future committed generating units. The separately on a permanent file, and the second characteristics while a specific OGP case is running. first can be stored can determine unit The former is a Gener•ation Model created for use as a data base representing the user's in-service and on-order generating units. In general, the user should build into the Generation Model all data about the units that is not expected to change during a study or series of expansion planning cases. Such an approach will tend to minimize data preparation time and potential errors, ultimately minimizing the cost of executing a study. There are two options available to users for easily modifying their generation model data. As represented in Figure 2-1, an at~iliary generation modification program is available which will override existing information to create a new file. This file can then be stored separately for use in subsequent expansion planning cases. Or, representation of the Generation Model data can be changed during the course of an OGP run by individual data value entry for any particular units. If that is done, the original Generation Model will not be altered permanently. GENERATION MODEL FOR EXISTING a COMMITTED UNITS , USE MODIFY a MAKE AS I--+ OR • • • A NEW GENERATION IS MODEL CERTAIN UNIT DATA MAY ALSO BE CHANGED DURING THE COURSE .OF ANY OGP RUN AT THE USER'S OPTION I • ~--------,-------------------------J .. 0 G p Figure 2-1. Input Options for Generating Unit Data 2-1 \ '· • The user has the option of specifying all of the data in the Generation Model. Or, if desired, customers may use and modify the available generation model data EUSED maintains for most electric utili ties in the United States. In all cases, the data contained in the EUSED files is obtained solely from public sources such as the Federal Energy Regulatory Commission, Edison Electric Institute, etc. At the Generation Model data level, the information is stored by individual unit. There is no limit to the number of units that may be represented in a Generation Model; however, the maximum number of individual generating unit.s with thermal characteristics that may be separately represented at any point in an OGP study is limited to 250. Also, only one equivalent hydro and three equivalent energy storage units will be used by OGP. Sections 6 and 7 will describe how lumping and equivalencing are directed and accomplished. The following information refers to some of the ways the Generation Model data is used later in the OGP simulation process. It is intended to guide the user in the initial placement of units within one type or the other and to highlight the flexibility of generating unit representation that has been built into this format. Units designated as Types 1 through 6 are basically designed to represent thermal generation. Type 7 units are designed to represent conventional hydroelectric generation. Additions of Type 7 units will not be made automatically by the OGP program; their timing and quantity must be specified by the user. Units designated as Types 8 through 10 represent energy storage units such as pumped storage hydro, compressed air storage and batteries. At the user's command, OGP will choose an optimum generation mix from among the nine types of thermal and energy storage units. The characteristics that can be specified by the user at the Generation Model level for each unit type are listed in Table 2-1 and summarized in Table 2-2. Data not input here can be separately specified later during the OGP run. This is discussed in more detail in Section 5. Note that only the data required for a particular study needs to be entered. For instance, if environmental discharge results are not required, the characteristics for those aspects may be omitted from the data. If a pool-wide study is being conducted, each company or area may input data separately. Automatic merging of the data is then done by the program for regional planning. It is important to note that any individual unit data not input here can be easily input at the Data Preparation level via the same mechanism used for characterizing new units, namely "standard tables." These tables are composed of a few discrete data points which are a function of factors such as unit size or year of installation.. Then, the program automatically interpolates for all sizes, where needed~ thus saving the user from having to specify separate values for all parameters for every unit. This mechanism can save a considerable amount of time and tedious labor. In summary, the user supplies any or all data for those individual units, as desired. For data that is not supplied, the program will access the data found in the standard tables. This p~ocess is described in Section 5o 2-2 '·--~c-------~----------~-----~-----------------.. 'll tl J " 1 j.;., Hi-"> ,, - ):' .. \' l I l ! I I I l l I i ~ l\) I w ~·-~w• _,. ··---••..,......>"-">n" -. -.. "-.•,,-,.-.......... •·-~ .......... ___.. ...... _.,.,........_,_~---· ~ i...--· ~--~--~-~~--....._...._-__.._.,.,....__,..,,....,,~, »...:.....-~--·~·--'-•••~·-·-·,__;~~,.._,_. ___ : .. -..__,_.,.......;..,..__.,._, .____,_,-~ .. ~--c:...-_,*'...;..... ____ -:;......,"-ol ......... -"! ~ ~ Unit Type No. 1 2 3 4 ~ •• -~ 8!!1 111!11 III - TABLE 2-1 General Characteristics Suitable for nuclear units because fixed and variable fuel costs can be entered for this type of unit~ Time variation of these costs to simulate effects such as core equilibrium may also be input by the user via use of per-unit multipliers. A maximum of 100 nuclear units may be separately represented. Suitable for central station types of fossil units. At the user's option, smaller units of this type may be automatically lumped by the program before the first year of the study to save computational expense during the yearly calculations. Fuel type may also be specified by company for these units. Suitable for peaking units, suoh as combustion turbines, because this type of unit is normally used to trim the yearly expansion to prevent overbuilding relative to the reliability criterion. May also be lumped at the user's option. Another possibility for base load or mid-range duty fossil units. Combined cycle units, such as STAG* (Steam and Gas Turbine) plants can also be easily represented here. * Trademark of the General Electric Company IIIli: -.. .. .. l!ll Specifiable Generation Model Data For Unit Types 1 through 6 (OGP-6): Station Name, Unit Type No. and Plant No. Company Assignment and .% Ownership Maximum Net Output, MW Minimum Net Output, MW Net Station Heat Rate, Btu/kWh Fuel Input at Minimum Rating Installation Year and Month Retirement Year and Month Fuel Type No. 1!!!8 Fuel Cost, ~/MBtu (also a $/kW/yr for unit types 1 and 5 or 6) · Fuel Inflation Pattern No. Fixed and Variable O&'M, $/kW/yr and $/MWh Mature Outage Rates Plant Cost, $/kW !he Following Additional Data May Be Specified When Using OGP-6A: Atmospheric Heat Rejection, per unit Sulfur Removal Fraction, per unit Precipitator Efficiency, per unit Carbon Monoxide, Nitrogen Oxide and Particulate Emission Coefficients, Pounds, Pounds/MW/hr and a per-unit Scaling Factor Water Consumption Coefficients, Gallons, and Gallons/MW/hr IIIIi Ill[ 0 ·/ l l ! :-1 l I r l ,, 1\) I ~ ~--< '~ ,_~ -~ -·~'~" ___ ,.._ ... ~~--.. --..... ...-. Unit Type No. 5 6 7 8-10 r r TABLE 2-1 (CONTINUED) General Characteristics Another possible nuclear type of un~t, because, if desired, this unit. can also be given a fixed fuel cost with cost variations independent of inflation similar to Unit Type No. 1. Alternatively, another type of fossil unit can be represented here. Same as Unit Type No. 5. However, a maximum of two unit types may have the characteristics of a fixed and variable fuel cost: Unit Types 1, and 5 or 6. This allows for the representation of up to two separate types of nuclear units, such as light water and fast-breeder reactors. Suitable for conventional hydro installations. Might also be used for the simulation of certain contracts or other unique forms of energy supply. The program will lump all individual units into a single aggregate. During the annual system production cost calculations, the total minimum rating is used as base load generation. Surplus energy available, up to the monthly total maximum rating, is then used to decrease peaking requirements. This type of unit must be derated if the user wishes to take outages into account in the reliability calculations. This is not necessary with Unit Types 1 through 6, which can be assigned forced and planned outage rates for probabilistic reliability and production costing calculations. Suitable for puruped-storage hydro plants, batteries, compressed air, or the representation of other energy storage devices. Will be automatically refilled or ch~t·ged up on an economic basls whenever possible. A percentage of unused stored generation can auto- matically be applied to decrease spinning reserve requirements during OGP system dispatch simulation. These units, which are similar to Unit Type 7, can also be derated for reliability calculations. Specifiable Gef!.eration Model Data For Unit Type 7: Station Name and Unit Type No. Company Assignment and % Ownership Monthly Energy, GWh Monthly Minimum Output, MW Monthly Maximum Output~ MW Installation Year and Month Retirement Year and Month Plant Cost, $/kW For Unit Types 8-10: Station Name and Unit Type No. Company Assignment and % Ownership Maximum Net Output, MW Maximum Charge Rating, MW Maximum Storage Size, MWh Installation Year and Month Retirement Year and Month Plant Cost, $/kW :> II ... v =--~~ ·~-~}-"'" .. ..:..:~ --~ . .-:-~--·------· <~~-~~--.... ·, ;-P>'! ~-:---~~-~r-:-~~~~~:-:-~-~~~~""' :r"!\#ff:~Mfo'£)t: ... )'!~;t~~.;7~~~~~~r-"'_"::" ·<·' "·--.. :..-~~..,..,~~~:~·r··:· "·:>it. ·:·~.- --l , I ' .I -~ I & '~ '11 a •. ~~.· - I I Types 1-6 7 8-10 TABLE 2-2 Generation Moded Data Typical Plants Nuclear Fossil -Base Load -Mid Range -Peaking Conventional Hydro Energy Storage Assignable Unit Identification Ratings Heat Rates Service Period Fuel Data O&M Costs Plant Cost Outage Rates Environmentally Related Data (OGP~6A) Unit Identification Energy Ratings Service Period Plant Cost Unit Identification Ratings Storage Capacity Service Period Plant Cost Before executing an OGP case, the user also has the option of modifying the Generation Model through the use of an auxiliary program. This program allows the user to modify an existing Generation Model and produce a new one while simultaneously retaining the old one. This option is presented schematically in Figure 2-1. In general, two types of modifications can be made to an existing file: (1) change selected characteristics of units already on the file, or (2) add or delete complete units on the file. The attributes of the generating unit that can be changed or initially specified are the same as those tabulated in Table 2-1. A program option also allows the deletion of all units from a particular record by the specification of a single input variable. 2-5 :, \ ,, ·. I I I I I I I I [ [ [ L fJ [; ~~ lc' i J r4 ~ r ~ I J r ~ ; A j ' r· ~ j r-t -" r·· i J f' ' ~ l· ) ~-" i i J t .. ?, t J 1; 1': 1 i ., r ' l. ' ·.~ ' ~ j~ .L.~ • \i k,,.; ' !l • Lot t li ),:j,j r L~ ~ L1 LOAD MODEL The user specifies the system Load Model used by OGP to represent peak and shape characteristics which are projected to occur for the years included in the OGP study. This means the Load Model Program does not have any independent forecasting capabilities integral to it. Figure 3-1 presents an overview of the basic load modeling options. As was the case with the Generation Model, the user has the option of supplying basic load shape data for use by OGP or of using historical data available from EUSED, which is obtained from the Federal Energy Regulatory Commission's or Edison Electric Institute's records for electric utilities. USE LOAD MODEL FOR THE YEARS TO BE STUDIED MODIFY a MAKE AS ~-------. OR • • • e • • . A NEW LOAD IS MODEL THE SYSTEM LOAD MODEL ITSELF MAY NOT BE CHANGED DURING THE EXECUTION OF AN OGP CASE. HOWEVER, THE EFFECTIVE. HOURLY LOAD SHAPE 8 HENCE THE IMPACT OF SYSTEM DEMAND ON RELIABILITY a OPERATIONS MAY BE ALTERED VIA INPUT .DATA ARTIFICES SUCH AS LOAD FORECASTING UNCERTAINTY, CONTRACTUAL PURCHASES OR SALES, ETC. Figure 3-1. Input Options for Load Model Data 3-1 0 G p In addition to the Load Model Program usually used with OGP, EUSED also has available a separate program that will convert a maximum of five years of historical hourly load data, in EEI format, into the input required by the OGP load modeling program. This hourly load data, which can be input by company or load area, will be automatically merged. Although up to 40 separate companies or areas may be input, a maximum of 25 is allowed in the OGP run. Alternatively, the user may supply load data for the combined pool. All data is converted to and stored on a per-unit basis as shown in Figure 3-2. 1.0 0.8 ::.:~ <(< I.IJL&J a.. a.. 0.6 >-..J ..J< I;:, ~ z 0.4 oz :E<( 0.2 1.0 o.s 0.2 MONTHLY RATIOS 2 4 6 8 10 12 MONTH HOURLY RATIOS WEEKDAY WEEKEND DAY 4 8 12 16 20 24 4 8 12 16 20 24 HOUR HOUR Figure 3-2. Examples of Per-Unit Load Models 3-2 ~ " ' l I '.·-r• I n . fl 11 ·~ I I I The resultant Load Nodel pr•oduced by the OGP Load Model Program will actually consist of two distinct models: (1) one model will be used for reliability analysis via loss-of-load probability (LOLP) calculations, and (2) the second model will be used for system production cost simulations. Use of these models is discussed further in Sections 8 and g. The basic time period for the OGP process is the calendar month. The daily LOLP load model consists of a weekday peak load model to be used for risk evaluation, and the system operation load model breaks down each month into hours with a different shape of twenty-four hourly loads for both weekdays and weekend dayso Both of these models may be prepared to cover a forty-year period, thus providing the user with flexibility for producing alternate OGP cases. However, the time period for any single OGP run is limited to thirty years. One of the major factors contributing to the amount of computer running time used for large-scale production simulation programs is the chronological load model~ Experience with both chronological and daily load duration load models indicates that the daily load duration approach is clearly justifiable for most long-range planning studies because of the favorable trade-off in computer processing time and model accuracy. Reductions in computer processing time, which can be achieved with the daily load duration approach are significant, while losses in accuracy are relatively small. However, the daily approach is substantially more accurate than using a weekly, monthly or annual load duration curve. WEEKDAY PEAK (MW) A 8 c D 0 TIME I --+ I --·I--- I 20 40 100 AT THIS LOAD OR GREATER ( 0/o) Figure 3-3. Example of a Load Model for Daily LOLP Calculations The daily LOLP load model is used to represent the distribution of weekday peak loads that can be expected during each month. This distribution is derived from company or pool history. As illustrated in Figure 3-3, the sample load model for daily LOLP calculations contains four values for weekday 3-3 peak load and the probabilities of the weekday peak loads being at this load or greater during the month. The load designated by point A Mv1 is the highest weekday peak to be expected during the month; the load designated by point D MW is the lowest. The conditions illustrated in Figure 3-3 are also represented in Table 3-1~ Load in MW A B c D TABLE 3-1 % of Weekdays When a Peak Load Equal to or Greater Than Amount Shown May Be Expected O% 20% 40% 100% In OGP, information such as that listed in Table 3-1 is assembled on a per-unit basis and then reconverted to MW as required for each year of the study. Each month may be different and may have its own distribution of four peaks. This information may also change from year to year. This formulation of the data can be modified by including a load forecasting uncertainty function to recognize that the forecasted peak may either be exceeded or may fail to materialize. This is accomplished during the execution of an OGP case. Input data required for .this option is described in Section 5o In contrast to the probabilistic load model that was concerned with the weekday peak loads only, the production cost load model must consider all of the hours in a day, both weekdays and weekend days, in order to meaningfully schedule generation and determine the resulting operating costs. Holidays are considered to be weekend days. Each month is different and is characterized by a typical shape for a weekday and a weekend day as shown in Figure 3-4. The number of times each of these shapes occurs is determined from a built-in calendar. Each of these day types is represented by twenty-four, one-hour periods of constant load. In addition to daily LOLP, the user may choose to calculate or design a system based on an hourly LOLP. To do the hourly LOLP calculation, the daily LOLP load model is used to define four weekly peaks for each month at O%, 20%, 40% and 100% of the time. Then the weekday and \veekend day per-unit values from the production cost load model are multiplied by each of the four weekly peaks 1 thus producing a typical weekday and typical weekend day at O%, 20%, 40% and 100% Qf the time. The four typical weekday and weekend day shapes plus the number of 'veekdays and weekend days for each month define the load model used to compute hourly LOLP. 3-4 f . . r- t .. l ~ . f , r· t .\ As was true of the Generation Model, the Load Model may be used repetitively for OGP studies. Alternativ·ely, certain characteristics of the Load Model are easily modifiable when required. This option is discussed in detail in Section 4. During the modification process, alternate Load Models are created and saved for use by OGP. Unlike the Generation Model which could be changed during an OGP run, the Load Model, with one exception, must be modified before OGP is executed. It is possible to input new values for the annual peaks as part of the OGP data, but the per-unit shapes and load factors from the input Load Model will not be modified. WEEKDAY WEEKEND DAY 0 4 8 12 16 20 24 4 8 12 16 20 24 HOUR HOUR Figure 3-4. Example of a Load Model for Production Cost Calculations Certain m~n~mum data must be made available for the Load Model Program to operate including the starting year, the number of years desired, and the number of companies or areas to be represented.. The following data must also be specified: {L 1. Data for the reliability load model a. Month/annual per-unit ratios, by month b. Per-unit ratios associated with the O%, 20%, 40%, and 100% points on the peak load duration curve, by month 3-5 r r·, r .• -~>' ' 2. Data a. b. for the production cost load model Month/annual per-unit ratios, by month Weekday and weekend day per-..unit hourly ratios in descending order, by hour, by month The following data may also be specified on an annual basis: 3. Annual peak MW load, by year and company or area 4. The per-unit ratio of the company peak loads to that of the pool, by compc:my 5.. The per-unit peak load grmvth multipliers, by company If the user does not wish the program to utilize its built-in calendar to determine the number of weekdays and weekend days and holidays for each month, ·the number of each may be separately specified. 3-6 LOAD MODEL SUPPLEMENTARY INFORMATION 1. Load Shape Modeling, T.C. Murrell, 1974 GE Memorandum* 3-7 I' ! ' I I I ' '' I ' i, I ,r ~~ lOAD MODIFICAl.IONS A load modification program is available to facilitate changes in OGP Load t•lodels. Historically, EUSED has used per-unit or normalized load models. The use of pe~-unit load data allows direct conversion of the per-unit load model to a HW load model each year by supplying only the annual peak load.. If load shape~~ do not vary during the course of the study, loads for each year in the study can be created simply from one year of shape data and forecasted annual peaks for each study year. Modification of a monthly/annual peak ratio allows energy in that month to be altered while preserving the daily load shape. Because of the uncertainty of future loads and their shapes, the Load Hodel Pr-ogram provides the user with several convenient methods for altering load shapes by specifying consistent combinations of monthly or annual load factor, load grow-th and/or enet~gy growth. This capability enables the user to evaluate the impacts on the magnitude or shape of the system load due to load management, new rate structures, etc. To minimize input data requirements, the actual load changes are performed automatically. The Load Model Program can be used in several ways to perform a varie't.Y of tasks. The following information is a brief summary of the major permutations that are possible and the input i terns which must be provided to accomplish those changes. 1. The overall yearly Load Model car be changed in the following ways: a.. Input the annual r-nv desired c1r the annual per-unit grm-1th multipliers for the pool annual peak loads. b. c .. d. Input the desired annual load factors in per unit. Specify annual energies by inputting the MWh for the first year and annual per-unit growth multipliers thereafter. Input the per-unit ratio of company-to-9ool peak for each company or area on an annual basis, the individual company peak load growth per-unit multipliers, or the annual MW peak loads for each company. 2. The Load Model can be modified on a monthly basis by specifying certain combinations of the following data.: a. b. c. Input the monthly energies by specifying the desired MWh for each month and year or by specifying initial values and a growth rate for each month of eac~ YGa" thereafter. Specify monthly load factors for each year. In pt..~~ ;t4>:>nthly pool peaks for each year in MW, by specifying an annual pool peak and twelve per-u1nit multipliers to define the monthly peaks, or by inputting monthly growth multipliers for each year. At the user's option, only the production cost load model or both the 4-1 production cost load model and reliability load model monthly pool peaks will be changed. 3. The load model used f'or the reliability calculations within OGP can be changed in one of' the following ways: a. Input the month-to-annual peak ratios annually. b. Change the per-unit ratios for the O%, 20%, 40%, and 100% points of the distribution of monthly weekday peak loads for each year. 4. The load model used for the production simulation calculations within OGP can be changed, using one of the following three options: a. Change the per-unit month-to-annual ratios for each month each year. b. Change the per-unit weekday-hour to peak-hour ratios on an hourly basis each month or annually. c. Similarly, change the per-unit v-reekend-hour to peak-hour ratios on an hourly basis each month or annuallye 5. To account for special holidays or other unique time periods, the following modifications can be made to override the program's built-in calendar logic: a. Input the number of' weekdays in each month annually. b. Specify the number of weekend days and holidays per month annually. 6. The number of years covered by the Load Model can be extended in either of the following ways: a. Use the complete per-unit load model associated with the last year, whether it is user supplied or read from the original Load Model file. b. Manually input the data to be utilized in the extended time period on an annual :.asis. c. ·ro add years to the beginning of the load model, use the per-unit load model associated with the first year on the original Load Model. In all cases, the maximum total time period permissible for Load Models is forty years. However, a single OGP run can cover only thirty years. If the user wants the load modeling program to automatically adjust the load factor, tv-ro options are available: modify the month-to-annual ratios or modify the daily shapes. The option chosen should reflect the changing load patterns that are impacting the load factor. 4-2 - \ . ~1· . ~ . . - ;I . ' ~-1 ~ i I ~]. ' ':' i ' 'i 'I ' . ~ ~··I I, rr-. . ' ' 'I ' I I>' I r( ~ L \;. -- [ [ Figure 4-1 shows an example of modifying the monthly peaks to increase the overall load factor. The program first determines the change in energy to be included on the load models This energy is then allocated to the individual months in proportion to the ttvalley depth." As a result, the lm~1est monthly peak load will be modified the most. The user can specify which months are to be excluded from modification • 0 ~ <( 0: ~ <( IJ..I 0.. _J <;::f :::> z z <( ......... I ~ z 0 :E 1.0 MODIFIED t . 0,8 1---~ ~ VALLEY ____ ~ ----, DEPTH 1 --~ I --1-- 0.6 INITIAL/ I 11r f- OA f- 0.2 I- I 0 JANUARY f 1 I r -_::1 ~-----I ---..., I ----L_ __ 1 I 1 I l I I I DECEMBER MONTH Figure 4-1. Example of Monthly Peak Modification The load factor can also be changed by modifying the daily shapes, as show"ll in Figure 4·-2. The program starts by determining the change in energy on the load model required by the new load factor. This energy is allocated to the individual months in proportion to the original monthly energies. Within a month, the energy is divided between the weekdays and the weekend days according to the original energy for each day typeo It is then allocated to the hours in the day in proportion to the "valley depth" of each hour, with the lowest hour being adjusted the most. The user can specify the months and day types to be modified. The hourly LOLP load model cannot be changed directly. However, any change introduced to either the production cost per-unit shapes or the daily LOLP load model will implicitly change the load model used to calculate hourly LOLP. 4-3 HOUR/MONTHLY PEAK RATIO 1.0~----""-~ 0.8 -......., ---'--...~~-~ ~-,__1 --,__,_ .__, MODIFIED O.G INITIAL/ t... 1_.., j .. .,_ --,__..,__, VALLEY DEPTH 0.4 I 0.2 o~-------------------------------------------------------~ I HOUR 24 Figure 4-2. Example of Daily Shape Modification 4-4 • r . fl ~ r g ~ 1 L l l l t L. 1. LOAD MODIFICATIONS SUPPLEMENTARY INFORJ\ttATION Examples of Load Management Calculations, G.Ae Jordan, 1977 GE Memorandum. 2. The. Impact of Load Factor on Economic Generation Patterns, G.A. Jordan, W.D. Marsh, R.W. Moisan and J.L. Oplinger, 1976 American Power Conference. 3., Load Management and Generation Planning, C.D. GallowaJr, L.K. Kirchmayer and R.W. Moisan, 1976 Conference on Power Syst.em Planning and Operations. 4-5 JJ ' ' -] ,J (ll. ; : iJ I·~ I. ~~ ' . ' L .II· ' ' ' L, ·t· .. L~ STUDY DATA Study data is defined as all information supplied by the user other than that which has already been described in the discussion of the Generation and Load Models. In this section of the handbook, the following OGP study input data are discussed: • Attributes by Unit Type • Alterations of the Effective Load Model • System Reliability Controls • Operationally Related Items • Financial Information Only a few study data input items will alter the effective Load Model. Along with the specification of load forecasting uncertainty and the application of contractual purchases and sales, it is possible to input new values of annual peak load. However, with the use of input study data at this point in the over~ll OGP study process, a user can modify the Generation Model in many ways. For example, override data and manual installations and retirements can be specified. The information in this section provides a complete description of nine types of optimizing units, as well as certain economic and operational characteristics of the Generation Model units. These descriptions are for data that can be applied to describe various types of generating units. 1. Generating Unit Descriptive Data for the Six Types of _Thermal Units a. Kind of Generation b. Each type of thermal unit is assigned to one of four physically descriptive categories. Each category, in turn, is assumed to have certain characteristics which are generically represented as either nuclear, base load fossil, intermediate fossil (such as mid-range steam, peaking steam or combined cycles), or peaking fossil (such as combustion turbines). The user can make a maximum of two nuclear designations, and these must correspond to units designated as Type Nos. 1 and 5 or 6. A maximum of 100 nuclear units is allowable. Permissible·Unit Sizes and Earliest Service Year Allowable Data for permissible unit sizes and earliest service year allowable applies to new thermal units only, and is illustrated in Figure 5-l. For the six thermal types, a total of sixty sizes and the years in which these sizes became available can be specified. OGP uses only discrete sizes and years. The types of units to be included in the optimization each year are at the user's discretion. OGP will choose one unit size from each type from among only those units which have been listed as available for installation in a given year. As an option, when the annual load growth or reliability criterion is sufficient to accommodate more than one new added unit, the user can 5-l FULL LOAD RATING,MW 900 - 800 -.;...,_.--~--- - 700 600 5001 - - ---·· ~1~~--~1--~--~~--~--~~--~--~'--~~~1 1985 1987 1989 1991 1993 YEAR IN WHICH THE RATING JS FIRST AVAILABLE Figure 5-l. Example of Unit Size Availability Input also specify that the program try mixtures of more than one type of unit in the optimizing trials. This is discussed further in Section 1. In any case, the maximum number of units, thermal or otherwise, which can be installed and/or retired in any one year is 100. c. Plant Cost To represent the plant cost as a continuous function of unit size, a set point, (MW 0 ,C0 ) and a "D" (doubling) factor can be specified for each type of unit as shown in Figure 5-2. The "D" factor is calculated so that when MW = 2(MW 0 ), C = (1-D) ( C0 ). This means that when unit size is doubled, th~; cost (expressed in $/kW) is reduced by the fraction D. For intermediate points, where a factor 5-2 . ' r f $/KW :I .1 I. I I I I II d. I I I I \ \ \ \ .· Co ~--------------~------------------------, C=(I-D)Co -----------+-----------~--........_ ......_ ----- ~c-----------t.----- Mwo MW=2MWo MW UNIT SIZE Figure 5-2. "D" Factor Representation for Thermal Generation of two does not apply to the change in unit size, C = C0 (MW 0 (MW)k where k = ln(l.O-D)/ln(0.5). For each type of unit, the user can also specify the per-unit inflation multipliers to be applied to plant costs on an annual basis. In addition, if more than one company or area is represented in the study, the plant cost can be adjusted with another per-unit multiplier for each company. Of course 1 any a typical unit may be assigned its own individual plant cost. ' Fuel Types and Costs A different fuel type can be assigned to each type of unit. The fuel designation given to unit Type No. 2 has the capacity for additional flexibility. Different fuel types can be assigned by the class of Type No. 2 units, which is comprised of automatic addition choices and manually installed units, and the class of units comprised solely of generation model units. The type of fuel assigned can also differ f'or each of the twenty-five companies or areas the user may have chosen to 'represent in the study. Information on a maximum of twenty fuel types plus their attendant cost and yearly inflation 'rates can 5-3 be stored and accessed later. Up to twenty different patterns of inflation can be specified.. Units designated as Type No. 1 and either No. 5 or No. 6 may also be given a fuel cost inputted on a fixed $/kW/yr basis. Both the variable and fixed portions of the fuel costs for those types of units can also be represented as varying factors during the years of the OGP study. This variation can be apcomplished in a maximum of five separate steps inputted as per-unit multipliers as shown in Figure 5-3. This program feature facilitates the more exact treatment of such phenomena as nuclear unit core equilibrium or changes in coal pile storage volume. Atypical units may also be assigned their own specific fuel costs on a unit-by-unit basis. ¢ /MBTU COEFFICIENT, P. U. $ /KW/YR COEFFICIENT, P.U. ..._ ___________ ------- 0.70 0 3 6 9 12 15 YEARS AFTER START-UP 1.00-- 0.90 ---------- I I-------- 0 3 6 9 12 15 YEARS AFTER START-UP Figure 5-3. Example of Nuclear Fuel Cost Variation With Time 5-4 ', I I I I I I I I I I J e. f. O&M Costs The. user can input fixed and variable components of 0&!'1 costs. The fixed O&M cost is input in $/kW/yr via specification of a set point and a "D" factor similar to that used for plant costs as was shown in Figure 5-2. The variable portion may be specified either as $/fired-hour/MW or as $/MWh. Each component, as well as the yearly inflation factor, may differ for each type of unit. O&M costs may also be assigned on an individual unit basis. Heat Rates Heat rate data for each type of unit is input by specifying the continuous full load, net station heat rate in Btu/kWh. In addition, as shown in Figure 5-4, to represent the unit characteristics at minimum load, the minimum load output as a per-unit of full load output (A) and the fuel input as a per-unit of full load fuel input (B) are specifiable for each type of thermal unit. Individual characteristics can be input for atypical unitso 1.0-----------------------,---- FUEL INPUT P. U. OF FULL LOAD MINIMUM s---------ar A MINIMUM POWER OUTPUT-P. U. OF FULL LOAD Figure 5-4. Input-Output Representation for Thermal Generation 5-5 ,, 1.0 g. Commitment Minimum Uptime Rule The Commitment Minimum Uptime Rule can differ for each type of unit. The uptime rule assigned to each type of unit may be overridden on a unit-by-unit basis. Once a unit is committed, it must operate, at least at its minimum output level, for the entire commitment period. The four different categories which may be assigned to a unit are listed in Table 5-l: TABLE 5-l Rule No. Characteristics -1 The unit must be committed all week, at least at its minimum, unless :i:t .is on planned outage. 1 If committed, the unit must remain committed for the entire week. 2 If committed for a weekday, the unit must remain committed for all weekdays; if committed for any of the daytime hours of a weekend day, the unit must also be committed !'or at least the night hours of that weekend day. 3 If commit ted, the unit must remain on line for all of the hours in the commitment zone. ~ue to program logic constraints, the final economic priority list developed for each month of the production simulation must have the unit uptime rule values listed in monotonically increasing order. Occasionally, this requirement will automatically cause the program to change this variable for some units. If the user does not want the input uptime rules to be changed by the OGP program, the user can select the option to develop the priority list based on the input uptime rules and sort, by economics, the units with the same uptime rule. h. Mature Unit Forced Outage Rates i. Mature unit forced outage rates can be specified separately for each type of unit in per unit as a function of unit size. As illustrated ir:t Figure 5-5, when points (X,A), (Y,B) and (Z,C) are input, the OGP program will interpolate linearly between the three points and assume outage rates at a constant value for units with capacities greater than Z MW or less than X ~~. This variable may also be specified on an individual, unit-by-unit basis. Mature Unit Planned Outage Rates Mature unit planned outage rates are handled similarly to the forced outage rates just discussed, and they also are uniquely specifiable by unit. 5-6 1 1 '!' .1 .. 1 " ;·, ., ' I ·I • J ·I ~-' . I • I' . £ ' l .. ,, I ' ,; OUTAGE RATE, PER UNIT 1.0 MAXIMUM c~--------------------------~--------~• 0 ~~----~~--~--------------~------------X Y Z FULL LOAD RATING, MW Figure 5-5. Mature Outage Rate Representation for Thermal Generation Alternatively, the user can input the specific months during which maintenance is to be scheduled for particular units. A maximum of 25 manual maintenance patterns can be defined, and these can be assigned to any or all of the units on the system. If a manual maintenance pattern is specified, the unit's planned outage rate will be ignored. If a unit is installed or retired in mid-year (i.e., during any month except January), the program, at the user's option, will prorate the planned maintenance for the year, based on the number of months the unit is in service. j. Immature Unit Outage Rates To obtain immature unit outage rates, a different per-unit multiplier may be input for forced and planned outage rates for each type of unit. The number of years for which the adjusted rate is to apply is also input; however, the same number of years must apply to both the forced and planned outage rates-., The alternative method is to input yearly multipliers for the first ten years after a unit has been placed in service. An example of this approach is iliustrated in Figure 5-6. If the unit is a mid-year installation, the first immaturity multiplier is applied to that portion of the first calendar year in which the unit is in service. 20.- 18 16 1- 14 1- OUTAGE RATE 0/o 12 ~ 10 ~ 8 i- ....... OL 0 2 3 4 5 6 7 8 9 10 YEARS AFTER START-UP Figure 5-6. Example of Multi-Step Immature Outage Rate Treatment for Thermal Generation k. Environmental Discharge Data If environmental discharge calculations are desired, ~ the characteristics can be input by type of unit and fuel. For $ummarization purposes, individual units may be assigned to one of 100 possible plants and the plants subsequently assigned to one of 25 regions. Other units which may be added during the course .of the study may be assigned to different plants by unit type. If desired, 5-8 ') .I 1 . I I. ' .\ unit commitment and dispatch can be simulated on a basis which biases operations to minimize the calculated environmental impact. (1) Characteristics Assigned by Unit Type The following data may be specified on a per-unit basis in OGP-6A for each of the six types of thermal units. • Waste heat rejected to the atmosphere • Sulfur removal efficiency • Precipitator efficiency (2) Fuel Type Assignations The following data may be specified in OGP-6A for the maximum of 20 types of fuel. • Heating value expressed as Btu/unit of fuel (OGP-6 also) • Sulfur content in pounds/unit of fuel and as a percentage by weight • Garbon monoxide in pounds, pounds/MW/hr and a scalar per-unit quantity to reflect the actual amount released with combustion • Nitrogen oxides in pounds and pounds/MW/hr along with a unique scalar • Particulates in pounds and pounds/HW/hr along with a unique scalar ., Water consumption in gallons and gallons/l-1W/hr (3) Unit Commitment Weighting Coefficients . At the user's option, per-unit weighting coefficients, assigned by region, can be used to alter or modify the priority list developed from the combination of economic and environmental factors. The objective function to be minimized is N L I=l M E WJ [ EIJ ( Pr ) ] , J=l where N is the number of units commit ted, M is the number of economic and environmental factors (M = 1 for OGP-6, M = 8 for OGP-6A), ~1 is the weighting coefficient for each factor, 5-9 ,, and ErJ is the hourly type J emissions or the hourly cost for unit I as a f'unction of Pr, the powe~r output of unit I. The following per-unit weighting coefficients can be used in OGP-6A: • Fuel plus variable O&M, $ (OGP-6 also) • Atmospheric heat rejection, MBtu • Cooling medium heat rejection, MBtu • so2, tons • NOx, tons • CO, tons • Particulates, tons • Water, thousands of gallons Fuel plus variable O&M is initialized to 1.0; all other coefficients have been initialized to 0.0, resulting in a unit commitment based strictly on economics. (4) Unit·Dispatch Weighting Coefficients The unit dispatch weighting coefficients that may be used are similar to the unit commitment weighting coefficients. However, the values need not be the same. Wher'e the commitment considered full load emissions and costs, the dispatch uses incrementa.·~ values. 1. Energy and Fuel Usage Constraints (OGP-6A only) OGP-6A has an option that enables the user to limit the energy output of certain units. This option is useful to model units with a limited supply of fuel or environmental restri<~tions. To use this option, all units within a specific thermal type designation are assigned a maximum monthly capacity factor or a maximum monthly energy output in MWh. Alternatively, a maximum fuel usage may be specified to limit the operation of units assigned to the designated fuel type. The energy or fuel limits may also be input on a unit-by-unit basis. Details describing how the commitment and dispatch procedures change when the energy or fuel limitations are in effect are presented in Section 9. All new thermal optimizing selections must 'be made from among the six types of generating units just described. The three types of energy storage units to be considered for automatic addition ar~ described in part four of this section. 5-10 J I j J J I J J I I I . l .J f L 2. Generating Unit Descriptive Data for Unit Types 7-10 (Conventional Hydro and Energy Storage) a. Deration b .. c. To simulate unavailability for reliability calculations, the user can specify a per-unit multiplier to reduce total apparent maximum output. Plant·cost The user can sp~e:ify a ~t/kW and inflation rate which may vary nnnually for each type of unit. O&t.f Cost For each type of unit, the user can specify a fixed component in $/kW/yr, as well as its yearly inflation rate. 3. Generating Unit Descriptive Data for Unit Type 7 Only (Conventional Hydro) a. b. c • Scheduling --- 'the user 'has the option of specifying that hydro generation be scheduled either all week or on l<Ieekdays only. The amount of hydroelectric (i.'nergy available for use in the reliability calculations may be specified separately from the production cost energy.. This is done on a monthLy basis. If a separate r·~liability energy is not specified, OGP will use the production cost energy with the run .. of-·rivero portion derated, E.'xcess ·Energy If there is more hydl'O energy available in a month than can 7Je Jsed, the user has the option of spilling the excess energy or carrying it forward into the next month. The user can also speoify a maxim::tm aru.ot:.at .of energy to be carried t'orwe r-rl., 4.. Generating Unit Descriptive Data for Unit Types 8-10 Only (Energy Storage) a. Efficiency One overall efficiency value is input for all units of each type by month and year. The efficiency is the ratio of the energy generated to the energy stored. If the unit also burns fuel (as with c1ompressed air energy storage) , the efficiency should include the energy from the fuel, resulting i;:;. an efficiency that might exceed 100%!' 5-11 ,, b. Char-ge/Discharge Rating~ For each energy storage unit, o sepapate maximum charge and maximum discharge rating is specified on a monthly basis. c. Heat Rate Heat rate data 9an be specified for compressed air storage units. d. Scheduling Order The order in which the three types of energy storage are scheduled can be specified. e. Cost Biasin~ The OGP energy storage algorithm schedules energy storage units to minimize production costs.. However 7 to operate the energy storage unit more or less than the economically desirable amount, the charging cost may be biased with a per-unit multiplier. This biasing is done separately for each type of the three types of energy storage. f. Operational Mode Each type of energy storage must be assig.Jed an operational mode from among the following possibilities: Option 1: weekday charge -weekday generate weekend charge -weekday generate Option 2: weekday charge -weekday generate weekend charge -weekday generate weekend charge -weekend generate Option 3: weekday charge -weekday generate weekend charge -weekend generate Note that Option 1 specifically excludes generation on the weekend, regardless of economics. Option 2, however, permits weekend generation. The third optior:, unlike the previous two, results in a daily charge/discharge cycle, which means the energy storage device is fully chai"ged at. the beginning of each day. g. Free Energy For each type of energy storage, ext~a energy that was not generated by the thermal system may be included for dispatch. This Dpecification is done on a monthly basis, and can ~epresent rainfall, melting snow, or ~u:ty other inputs to a pumped-storage pond. The variable can ba n()gat.ive t.o represent such things as evaporation from t;he puraped-storage pond or air loss in compressed air caverns. · , ' , )'l·, • r , , :$' 1 , '"f£4L. ~ ,J•; I I I I I I I 1 .. I. jt·' \ . ~~ The three types of energy storage units to be includ€d in the optimization lvill have the characteristics just described for Unit Types 8-10. 5. Generating Unit Descriptive Data for Unit Types 1-10 6 .. a. Fixed Charge Rate b. The levelized annual fixed charge rate may differ for each type of unit as lV'ell as for each of the companies or areas represented in the OGP study. This rate may also vary each year of the study. Retirement Policy For each type of unit, the number of years from the initial installation date to retirement is entered. This data can be input separately for existing units and for the units automatically added by the program. Alternatively, any existing or manually installed unit may be assigned a particular retirement year and month. A maximum of 100 units may be retired and/or installed each year of the OGP study. Units that have been added automatically during the study may also be retired during the study period. Data Concerning Effective load Model Modification a. Purchases and Sales A maximum of ten individual contractual commitments for purchases and sales may be represented, and all the data in the following list may be changed each year of the OGP study. To simulate the impact on the reliability calculations of emergency purchase capability from neighboring systems, zero-hour contracts can be specified. • Contract name • Number of hours/day for which the contract is in effect • All vi"eek or vteekday only operation • Demand charge, $/kW/yr • Demand charge inflation rate, per unit e MW used for calculating the demand charge, if different than the monthly contractual demand • Energy charge, $/MWh • Energy charge inflation rate, per• unit • Monthly demand, MW 5·-13 b.. Load Forecasting Uncertaint.z As shown in Figure 5-7, the user can specify a distribution of peak loads in per unit of the forecast peak load on the Load Model along with the per-unit probability of each of the peak loads occurring. A maximum of ten points on the distributlon may be input. Thus, a maximum of ten new peak load forecast values is calculated from each point on the original Load Model for use in the reliability calculations only. The LOLP is calculated at each load level and weighted by the probability to produce one expected value for LOLP$ However, one should note that the use of load forecast uncertainty with the hourly LOLP option may incur substantial computational expense. PROBABIL ( P.U.) !TY I 0.21 0.04 -s 0.50 0.21 0.04 l 0 +3 PERCENT ERROR IN FORECAST PEAK ( 0/0 ) Figure 5-7. Example of Load Forecasting Deviation Input for LOLP Calculations c. Overriding the Annual Pool Peak New values for the annual pool peak on the Load Model can be input. All of the per-unit information on the Load Model will remain unchanged, and the MW loads will increase or decrease in proportion to the new pool peak. 5-14 (:; ' I I 7. I I I I I I I I I I I fl L !I be. fl L,, '~ '~ Other Reliability Related Data Input Items a. Unit Size Guide The user can control the unit size selected by OGP for optimizing comparisons each year if the type of unit is available in more than one size in a given year. Different guidelines can be input for the three kinds of units: base load, intermediate or peaking. To control unit size selection for base load units, the user must state the approximate number of base load units that should be added to meet a certain number of the years of load growth. The program will use that input while maintaining the percent reserve margins or reliability of the system to determine the unit size to consider. For intermediate-sized thermal units, the user can control siz~ selection by specifying a per-unit multiplier which is applied to the current year's load growth. To select the size for thermal peaking units 7 the user can specify a per-unit multiplier which is applied to the system installed capacity. Both of these multipiers may be changed annually. As an optimization alternative, a single. generator size (MW), maximum charging rate (MW) and storage capacity (MWh) may be designated for each of the three types of energy storage. These three quant.ities may be changed annually. b. Using a Unit Type to Trim the Yearly Expansion To minimize possible overbuilding of a system in any year while optimizing, the .1~;er can state that when the system's capacity deficiency for any ~ype of unit under consideration is at or below a specified percentage of that unit's size, the difference can be compensated for with an appropriate amount of capacity of trim units. Any one of the unit Type Nos. 1-6 can be designated as the trim type. If not otherwise specified, the program will trim with unit· Type No. 3. Implicit in the use of this option are two assumptions: (1) the type of unit selected as the trim type has an economically desirable peaking capacity and (2) the unit's size is not greater than that of the type of generating unit it has been allowed to supplant. A different percentage is specifiable for each of the six types of units. c. lfew Capacity Installation Criterion The criterion upon which new capacity installation is to be based may be specified as either a probabili3tic risk ind(;!X of days per year (daily LOLP), hours per year (hourly LOLP), or an installed percentage reserve requiremen1i. There are several ways in which the percent I'eserve margins of system reliability can be specified. Normally, it is based on the month of each year in which the annual peak load occurs. Al ternatj_vely, the user can specify any othe.r month during which th~.s test should be conducted. 5-15 () ·~t::~ .... ·-·-···-····-.. ' ~ r__, ~ .• , ... ,, . ,, ···~' The actual calculation of the percentage of system reserve can be specified in three different w~ys. The program uses the first method listed unless the user specifies one of the others. Method 1: [(Capacity + Contr~cts) -Load]/Load Hethod 2: [Capacity + Other Contracts (Load 0 Hour Contracts)]/(Load -0 Hour Contracts) Method 3: [Capacity-(Load-Contracts)]/(Load -Contracts) Additionally, the user may specify any month other than January in which all automatically added units are to be installed. d. $conomic Reserve Margins The new capacity installation can also be based on economic reserve margins. The program will automatically add new units beyond those needed for system reliability, if the savings in production costs, resulting from the displacement of higher cost generation, is enough to offset the additional investment cost. The user specifies the years in which this economic overbuilding is to be evaluated and the unit types to be used. A maximum percent reserve margin can also be input. e. Mature Outage Rates in Optimizing Decisions In the optimization process, a decision-biasing option, termed the "look-ahead" feature, may be utilized. To do so, the user specifies that the relative economic merits of new generating units will be evaluated with their mature planned and forced outage rates \·lhich are anticipated to be in effect for most of their service lives. Thus, by the use of this option, new generating units are not penalized by their immature characteristics. Instead, a total system evaluation is made of the new generating units based on their anticipated future outage patterns. After the preferred new generation alternative has been decided, the OGP process then recalculates the year in question, utilizing the immature outage rates that have been specified. This recalculation is necessary to obtain an accurate estimate of the actual total system costs for each year of the study. Since the decision to add capacity is based on mature outage rates, the recalculation with immature outage rates may indicate a capacity shortage. When this occurs, the program adds enough thermal units designated for trimming (usually gas turbines) to satisfy the reliability criterion. 5-16 ,, :I ,_ .I I I .I .I :J ;I -I I I . I ~-' _I I . ~j J J J .J 89 Other Operationally Related Data Input Items a. Levelized Future Fuel and Variable O&M Costs in Optimizing De·cisions At the user's command, anticipated variable operational costs may be simulated and taken into account similar t0 the procedure just discussed for using mature outage rates to make new generating unit evaluations. In this case, however, when the cost inflation of these operational factors is defined, OGP automati ~lly calculates and substitutes levelized values for use in the ye&rly production cost calculations. The user inputs the number of years following the decision year for which the program is to calculate these levelized values. Also, as is done for the outage rates, after the best type of generating unit is selected using these levelized values, the current year's values are retrieved, and a complete production simulation is calculated to supply the correct values for the record. be Planned Maintenance An outage due to planned maintenance may be disallowed during the month in which the annual system peak load occur3. Alternatively, a maximum o.f five months may be specified in place of' that m.onth. The user can also define a maximum of 25 maintenance patterns, which can be changed annually. Specific units may then be assigned to one of these patterns. In order to minimize the annual system risk, OGP will select maintenance months for all units not assigned to a specific pattern. The maintenance schedule developed by OGP can be saved and used for subsequent OGP runs. The system must have the same units in all runs using the same maintenance schedule • c. Unplanned Maintenance Treatment By extending a unit's planned outage period, the. user can simulate forced outages. Alternatively, the user can select a stochastic treatment in all the year.,ly decision calculations or only i.n the calculation of the final yearly costs. This technique calculates the effect of random forced outages after unit commitment and will r·esult in the simulation of emergency energy purchases. d. Spinning Reserve Specifications The spinning reserve may be expressed on a yearly basis in one of three ways: (1) as a percentage of the monthly pe1ak load, (2) as a specific M~i requirsment, or (3) as a per-unit rat.io of the largest unit that is not on maintenance during the month under study. Further, the maximum percentage of unused pumped-storage hydro generation that may be considered ·available for credit toward the spinning reserve required can ~lso be input separately. 5-17 l '1 J ,i • e. Sale of Excess Energy In cases where generating unit cycling limitations and m1.n1.mum output specifications result in a system operating condition where the load is less than the sum of the minimum rating of the committed units, OGP will terminate its calculations.. The user has the option of allowing the case to proceed automatically. To do this, the user must specif'y that some of the excess generation be sold by stating the maximum percentage of the minimum unit loadings that may be sold. The $/MWh and annual inflation rate which is to be associated with this transaction may also be specified. f. Purc?ase of Emergency Energy In instances where manual expansion or random f'orced outage rate calculations are utilized in the production simulation, shortfalls of available generation may be indicated for some periods of time. An indication of a shortfall may be interpreted as the expected value of energy not served. In such cases, the user can specify the cost of such an unexpected tie energy purchase in $/MWh along with its own yearly inflation rate. g. Commitment Zone Specifications A maximum of six commitment zones may be defined, and these must be apportioned between the average weekday and average weekend day. The number of hours in each of these zones must also be defined, as shown in Figure 5-8. The unit commitment will remain constant throughout each zone. Commitment changes between zones depend upon the minimum uptime rules associated with the generating units. 9.. Other Financially Related Data Input Items o c.r a. Initial Plant Investment Accou~ts To account for the capital costs of the existing system, the user can input one number which represents annual fixed charges for the units on the system at the start of the OGP study. b. Chan~es in Investment Cost If a unit is added either manually or automatically, its fixed charges on investment may be deleted at the user's option when the unit is retired before the snd of the OGP study. c. Cost Basis Year All cost data 1nput to OGP must be referenced to one speciftc year. The various annual inflation rates input separately will adjust the costs from this reference year to the year being studied. I} ~· "1tl 5-.~:.o • \... •.• .;! L~J ,. .. I I I I I I I I I 0 J d., .I .I J .I J I I I WEEKDAYS WEEKEND DAYS COMMITMENT ZONES I COMMITMENT ZONES 2 I 3 4 5 I 6 I I 4 8 12 16 zo 24 4 8 12 16 20 24 HOUR HOUR Figure 5-8. Example of Thermal Commitment Zone Specification J Present Worth Calculations The user specifies the preJsent worth interest rate and the reference year for doing the present worth calculation on the annual system ·costs. It is assumed that the costs for each year occur either at the end of the year in December, or at the beginning of the year in January. 5-19 STUDY DATA SUPPLEMENTARY INFORMATION 1. "D" Factor, GE Memorandum. 2. Typical Generation Planning Inputs, HaH.. Heiges, Power Generation Report No. 162. 5-20 • i) I I I I I I I I I I J I I I DATA PREPARATION The Data Preparation portion of OGP reads the user's input data, adapts it as needed for use by the OGP program, and then displays it in a printed output format which enable~ the user to easily examine and refer to the· data~ Input data is accepted from three sources: the Generation Model, Load Model, and study data. These topics are discussed in Sections 2, 3 and 5. The following major automatic data preparation processes are pert'ormed by OGP: • Checking of' the user's input data for nominal errors • Lumping of certain fossil units according to the user's speeifications • Aggregation of conventional hydroelectric and other types of energy storage • Initialization of the system for future planning This section of the Descriptive Handbook covers a number of the key functions that are initiated in the Data Preparation portion of OGP. First, the user's input data is checked for nominal errors. If errors are found, a summary of the errors detected is displayed on the output, and program execution is stopped. Next, the data is merged, sorted 1 and reduced for further use. The reduction logic used is designed to minimize the cost of computation in areaa that have little effect on model accuracy. For example, the program routines process individual station data for conventional hydro plants (i.e., generating unit Type No. 7) and lump these individual plants into one aggregate conventional hydro plant. This single plant is characterized by a monthly maximum output in. MW, a monthly minimum output in MW, and a monthly energy output in MWh. All values are determined by summing individual unit characteristics. The OGP Data Preparation routines also lump individual energy storage units within generation unit Type Nos. 8, 9, and 10 into a single energy storage plant for each type~ The cycle efficiency of the lumped energy storage plant is specified by the user. A weighted average of the individual unit cycle efficiencies is the preferred approach for estimating the lumped unit's cycle efficiency. Caution is advised when the user chooses to lump two plants of very dissimilar characteristics such as one with large storagE! and small generation capacity and one with small storage and large generation capacitys The user can exercise morE~ control over the lumping procedure for thermal generating unit Type Nos. 2 and/or 3 than that used for hydroelectric generating. units. The user specifies the largest unit capacity to be lumped as well as the maximum lump size allowable. Lumping is recommended only for the smaller units in relatively large systems. When utilized judiciously, lumping can reduce program computation expense both in the reliability calculation and production costing areas ~tithout resulting in a prohibiti v~; loss in calculation accuracy. 6-1 I Characteristics are assigned to each lumped gener~ting unit as indicated in Table 6-1. The standard table data values used are based on the size of the resultant lump. l''or example, if five 10 MW units are lumped together, they will be assigned the characteristics of a 50 MW unit. Maintenance patterns, energy limits, and/or fuel limits should not be specified manually for the individual units which comprise a lump; however, the lumped units together may be assigned a specific maintenance pattern, energy limit, and/or fuel limit. If' no maintenance pattern is specified for the lumped units, OGP automatically will schedule maintenance, using the standard table planned outage rate. Unless specified by the user, all lumped units are assumed to have no energy or fuel limits, even when some or all of the units comprising a lump have these limits. Unit Characteristic Fuel Cost Fuel Input at Minimum Output Fuel Input at Maximum Output Heat Rate Inflation Pattern Installation Year (Type 2) Installation Year (Type 3) Minimum Output Rating O&M Costs Planned Outage Rate Forced Outage Rate Environmental Emissions (OGP-6A) Retirement Date Station Name (Type 2) Station Name (Type 3) Plant Identification No. 6-2 Numerical Value Weighted MW Average Based on Lump Components Standard Table Standard Table Weighted MW Average Based on Lump Components Pattern Assigned to Thermal Type Zero Weighted MW Average Based on Lump Components (used to deter·mine the lump's retirement year) Standard Table Standard Table Standard Table Standard Table Standard Table Standard Table "Equivalent No. " "G.T. Lump No.n Plant 100 I I I I ,. I li I 'I I I I I I I I I ! I. I ' I ,. I i I ADDITION OPTIJ\11ZATION The next portion of the program with which the user becomes involved is the OGP optimization logic, which compares the total system cost of thermal and energy storage alternatives.. The generation and loads are modeled, and all other input data has been initialized to reflect the conditions for the first year to be studied. The user can direct and control OGP's optimization process. Usually, the optimum expansion is the one that minimizes the present worth of future revenue requirements. In OGP-6A~ the economics used to determine the optimum expansion can be biased by other factors. The impact of restrictions on unit operation due to fuel or energy limitations can be included in the decision-making process. By judicious data specification, OGP can be biased toward expanding the. system to produce the least total impact on the environment. The OGP process automatically spans time, year by year, to determine answers to the following questions for each year of the study: (1) How much generation should be added to the system? and (2) What kind of generation should be added to the system? Then OGP tabulates the total system costs or impacts of that stream of additions. After the user has input into the program a list of available units and their characteristics, the OGP process chooses from this list the optimum combination of unit additions for each of the years under study. This selection is made in a logical manner which satisfies the input constraints. OGP produces a year-by-year plan that also has been optimized through time. OGP's addition optimization process differs in several ways from a "screening curve" approach. One difference is that screening curves for a system change after each new unit is added and also each time the load and generating unit availability change. Thus screening curves are a very gross approximation of system operation. As a result, the process does not lend itself to studying a mixture of new additions or accurately calculating resultant system costs. The OGP method, on the other hand, actually performs for every new type of generation being considered a complete hourly system commitment and dispatch in the context of the total system's operation. The costs calculated via the OGP method are a very accurate simulation of how the new units being considered would actually impact on the total system's operation. Thus they aid the planner in the decision-making process by providing an excellent guide for the selection of the type of unit with the lowest overall cost. The use!'' may also study the merits of a manually fitted expansion and obtain meaningful documentation of the costs and impacts of those additions. The use of sepa:rate computer programs to determine generation reserve rE!quirements and perform production simulation does not lend itself to minimizing the sum of the investment and operating costs for a thir•ty-year study. Thus to help synthesize system expansions, in 1968, the reserve and simulation calculations were packaged into one computer code with an optimization procedure. Dynamic programming was considered, but sir..ce the Markov property does not apply (i.e., factors such as generating unit outage probabilities and nuclear fuel costs depend on the maturity of the unit), the 7-1 l t, r, i' l number of combinations to be considered in a thirty-year expansion becomes prohibitive. Linear programming also has been applied,.but it is difficult to model the nonlinear load carrying capability of the system with unit additions or the many individual operating rules of utilities such as minimum run or shutdown times. OGP's inherent optimum-seeking capabilities are enhanced by several user-controlled features l<lhich are internal to the program. These include logic (e.g .. , the "look-ahead" feature) which allows for the anticipation of future operating costs and outage rates, the addition of a mixture of types of new units in any given year, and the use of decision-making aids which seek to closely fit unit sizes to yearly load growths. Two additional points regarding the accuracy of the OGP optimization process should also be emphasized. Because the OGP optimization process is continuous, there are no end e·ffects to complicate the interpretation of the study results. In other words, if two studies were conducted and there were no changes in the input data except f'or the number of years studied, the results from the two OGP cases for the initial overlapping time period would be the same. For example, if case A used the same input data as case B, except that in case A the years 1990 through 1999 were studied, and in case B the years 1990 through 2009 were studied, both cases would display the same optimum expansion plans and output quantities through the year 1999. Thus, the OGP optimum stream of additions, as produced by applicati.on of the 11 look-ahead 11 feature, is continuously self-correcting and can effectively answer both of the following questions: (1) What is the best unit to add to my system next? and (2) What should my system mix be through the end of the study? The second major advantage of the OGP optimization logic is that the user is never required to deterministically or otherwise artificially restrict the absolute number of any type of future generating unit candidate that may be optimally evaluat~d or added by the program, either on an annual basis or as a total for the entire time period under study. The only exceptions are the current built-in limitations of 100 units added or retired in any one year and 250 units in total on the system in one year. These limitations are dictated solely by an effort to minimize core storage requirements, and could be increased, if the uservs needs dictate it. Because OGP does not require arbitrary estimates or restrictions on the number of unit additions per year, the program yields two significant benef'its for the user. First the program cannot find itself halted during the study by a preset tunnel or wall. This means all cases will continue through the final year of the study and yield complete results, thus obviating restarts and iteration solely to obtain f'ull-term outputs. The other benefit of this feature is that it precludes spurious or "local" optimums. In any case, it is only logical that the following question should arise~ How can a yearly decision...;making process be utilized to produce a stream of optimized results when, in retrospect, future con,ditions will almost always prove past assumptions to have been faulty? For example, fuel cost inflation may affect the relative economic desirability of operating certain types of 7-2 '!.. ...... ,,.; I ' I -1 I I I I ' I .1.,. ' t . ·' units, or the maturation of outage rates may increase the effectiveness of certain kinds of base load generation and, hence, strengthen their usefulness. To address such possibilities, the user has the option to "look ahead" at each unit in the yearly decision-making process-When each type of generation is considered by the program for comparison with other types to choose the "best" one, levelized values of fuel and O&M costs and mature outage rates can be utilized in all calculations. The num~er of years for which the "look-ahead" feature is to be implemented is specified by the user .. Typically, the number specified will be a function of the study length and the planning philosophy of the company, thus reflecting the critical payoff period for projects and the uncertainty present in the·input data •. Thus, through use of the "look-ahead" feature, the potential anomaly of "smart n decision.~ made year-by-year proving incorrect in time is avoided by anticipating the effect of future changes in system conditions. Naturally, if the input data is time invariant, there is no need for ''looking ahead" in the OGP optimization process, and the optimizing loop will proceed directly to the solution. In any case, for cost and impact documentation and for advancement to the next decision year of the study, after the optimum type of unit has been chosen with the use of "look-aheadlf calculations, a final complete system dispatch and costing process is then conducted using the input data specified for the year under consideration. The addition logic, which searches to install the best combination of new units each year, operates only during the years in which system load growth and the reliability criterion will accommodate the addition of more than one unit. Briefly, the program begins by considering each of the available types alone, adding as many units as may be required. The total levelized operating and capital charges \·Thich accrue from each type by itseJ.f for the first year are stored. This provides the program with a relative ranking of the economic desirability of the unit types available. Beginning with the type that gave the lowest total cost, the program starts mixing. One or more of these lowest cost units is replaced (as required by their relative size and availability) by one or more units of the next cheapest type of available generation. This levelized cost is then computed and stored. Next, a comparison is made to determine if the cost of the second combination is less than the cost of the first. If it is, the logic replaces more units of the first type with the required number of additional units of the second type, and then computes the total levelized costs which are compared once again to the lowest previously obtained cost. When a mixture of units is obtained that yields a cost greater than that from a. previous trial, the program no longer will attempt to increase the number of Jnits of the particular type of generation it had most recently added.. Instead, OGP recalls the lowest cost combinat.ion obtained thus far and tries to beneficially replace some of these units with the next cheapest type of unit allowed. Computation costs are thus conserved by avoiding the calculation of patently unprofitable combinations. 7-3 ' ' ' •• If the user chooses to allow the program to overbuild the system based on economics, the process begins as described earlier. OGP considers each of the available types alone, adding sufficient capacity to the system to satisfy the reliability criterion. Then, for the types of units designated as available for overbuilding, the program continues to add units beyond those necessary for reliability, as long as the total system costa decrease. At the user's option, the program will then take the lowest cost. overbuilt expansion from the types· of units labeled "base load" and attempt c'idditional overbuilding with the lowest cost energy sto":"age option. The best expansion evaluat.ed thus far." now becomes the starting expansion for the mix logic, where combinations of unit types are evaluated. It is important to note that in the preceeding description it has been assumed that the OGP program preselected a single available unit size for each of the available types of generation. This unit size selection was based on the u.ser~specified size availability definit.ion, described in Section 5, and the unit size guidelines input for the three kinds of generation--base load, intermed.iate, and pea.king. Although OGP si.mul taneously observes other factoz~s, such as maintaining an appropriate amourit of' installed reserve, it is the user who has controlled the sizes on which to optimize. When factors such as negative economies of scale or dramatic changes in availability as a function of size oi" time are represented, caution must be exercised in apecifying the optimization choices. In addition to the yearly unit mix and size selection logic, it is also possible to trim the expansion to prevent an excessive amount of capacity overbuilding in any one year. This input item also was described in Section s. As discussed in that section, trimming may not be advisable when the relative system economic desirability of' the unit used for trimming purposes would. not normally be exhibited, or when the only size of trim unit available in a given decision-making year is greater than that of the unit being supplanted. The net effect of the OGP optimization proce~s is discussed further• in Section 12. At this p•Jint, however, it is sufficient to remember that the program logic perfo~.:m8 the t.mit selections via a comparative, iterative simulation process to obtain an expansion which recognizes the realities of the generating units~ physical and economic attributes and restraints. Thus, the OGP optimization program logic represents the actions ordinarily performed by a 3ystem planner. 7-4 r! . ~~~- ~ , . < f : I ,I .. f ' ' . I ' ~ ~ ·~r:, t'! •' l ~ . . ' . ' L . ' . . ~> ~ i ' "' ADDITION OPTIMIZATION SUPPLEMENTARY INFORMATION 1. How Are Optimization Methods Used In the Optimized Generation Planning Approach?, L.L. Garver, 1978 GE Memorandum. 2. Unit Size Selection, G.E. Haringa, 1977 GE Memorandum. 3. OGP-5 Unit Mix Logic, G.E. Haringa, 1978 GE Memorandum • 4. OGP-5 Siting Logic, G.E. Haringa, 1978 GE Memorandum. 7-5 I'' ' -' ~~ 1 ... : ,J r· ,, ' " t~ ~~ 0 ~~ /'\ [, ' .. ~· l~ L$ ~~ I . .. lJ t.J tj LJ t~ t~ L RELIABILITY EVALUATION Based on a user-specified reliability criterion, OGP will automatically determine how much new generation is needed each year by analyzing the system loads and generation. One of three possible reliability criteria may be specified: (1) daily loss-of-load probability (LOLP), (2) hourly LOLP, or (3) percent reserve margin. The system can also be expanded to the. economic reserve margin. After <=J.dding new capacity to satisfy the reli,ability criterion specified, the program will install additional capacity if it is economical to do so. The maximum number of individual generating units that may be added and/or retired in any one year is limited to 100. The amount of generating capacity required to serve a specified sequence of load demands for a giveri year may be computed using a probability model of generating unit availability termed the loss-of-load probability (LOLP) method. Since its introduction in 1946, the LOLP method has gained wide acceptance in the electric utility industry. Currentlyt OGP will calculate a daily LOLP and/or an hourly LOLP. Historically, utility system planners measured generation system reliability with a percentage of generation reserve index. This planning design criterion only measured the difference betwean total installed generating capacity and annual peak load demand. However, this approach proved to be a relatively insensitive indicator of system reliability, particularly when new alternative units with varying sizes and forced outage rates were compared. Today, LOLP is the accepted measure of generation system reliability. The LOLP technique is a probabilistic measurement of the expected number of days per year on which the available capacity cannot. meet the load demand. The LOLP index provides a consistent and sensitive measure of generation system reliability, although its name is somewhat misleading in two respects. First, the Lidex is not a probability; it is an expected value of the number of days per year of capacity deficiency. Second, it is not a loss of load, but rather a deficiency of installed available capacity. Despite the misnomer, the LOLP approach is well accepted in the utility industry today. It should also be noted that, in gen6ral, daily LOLP is not related to hourly LOLP by a factor of 24 hours; i.e., daily LOLP does not equal hourly LOLP divided by 24. The following discussion refers mainly to daily LOLP. Similar program processes are used when hourly risk is desired. The process of calculating the OGP system's reliability index involves the following steps: 1. Choose an index. 2. Deterministically modify loads to reflect contracts and zero-hour contracts. 3. Schedule conventional hydro (derated) to minimize LOLP. 8-1 r l [ [ L f f .1L f t L Schedule energy storage (derated) to minimize LOLP. Schedule maintenance to minimize risk. 6. Build a system cumulative capacity outage.table. 7. Convolve the capacity outage table with the load model. 8. 9. Add new units as required to satisfy the specified reliability criterion. Determine resultant risk(s) and effective load carrying capability. Generation system reliability is affected by several factors such as load characteristics, unit size, and planned and forced outage rates. A generating unit's planned outage rate is a measure of the time required each year to provide for planned maintenance during scheduled outage periods.. Typically, these planned unit outages are scheduled in the spring and fall when peak loads are reduced from summer or winter. In OGP, unit maintenance periods are automatically scheduled to minimize risk. Despite the scheduling of maintenance to minimize the effect on system reliability, adequate generating reserves must frequently be installed to maintain system reliability during unit maintenance periods. The forced outage rate of a generating unit is also important in assessing a unit's effect on generation system reliability. While unit size determines the magnitude of the outage, the forced outage rate indicates the total duration of failure or unplanned downtime. The effect on system reliability will vary with the type of generating unit as well as with the unit's maturity, its design, and the effectiveness of its maintenance program. After the individual unit forced outage rates are known, the cumulative capacity outage table is developed. Basically, this requires the identification of all possible outage events (e.g., in a system with N units, this means ;::N events) and a determination of the probability of the outage occurring.. Hm<JE:Yer, since the LOLP approach is more concerned with system capacity outages than with particular unit outages, the probability of a given total amount of capacity being on outage must be calculated. This information is presented as a cumulative capacity outage table as described in the upcoming example. OGP uses a highly efficient recursive computer technique to dir·ectly cal.culate the cumulative capacity outage table from a list of unit ratings and forced outage rates. For example, consider a small system comprised of only three units. The thermal system is represented by a cumulative capacity outage table which answers the following question: Given the three unit sample system characteristics listed in Table 8-1, what is the probability of hav·ing X MW of capacity or more on outage? 8-2 r,;· r- fr z ' ; [t t I LL Unit Capability (MW~ A 100 B 150 c 200 TABLE 8-1 Forced Outage Rate (P.U.) 0.01 0.02 0.03 (1 -F.O.R.) Innage Rate (P. U.) o. 99 0.98 0.97 The probabilities of all possible combinations of uniL3 being in or out are calculated as shown in Table 8-..2,. The cumulative column, which gives the probability of X HW or more on outage, is obtained by starting with the value at the bottom of the probability column and adding upwards. For example, a cumulative value of 0.000600 is obtained for X MW = 350 by adding exact probabilities of Oe000006 and 0.000594. TABLE 8-2 Units Probability of X MW On·Outage X M\ol Probability or More on Outage None 0 (0.99)(0e98){0o97) = 0.941094 1.000000 A 100 (0.01)(0.98)(0.97) = O.D09506 0.058906 B 150 (0.99)(0.02)(0.97) = 0.019206 0.049400 c 200 (0.99)(0.98)(0.03) :: 0.029106 0.030194 A,B 250 (0.01)(0.02)(0.97) = o. 000194 0.001088 A,C 300 (0.01)(0.98)(0.03) = 0.000294 0.000894 B,C 350 (0.99)(0.02)(0.03) = 0.000594 0.000600 A,B,C 450 (0.01)(0.02)(0.03) = 0.000006 0.000006 The cumulative capacity outage table must be recalculated each time there are any changes in unit rating, forced outage rate, unit retirements, or new unit additions. This requirement i,s a significant factor that should be considered in the writing of LOLP computer codes, if computer running times are to be maintaine0 at reasonable levels without sacrificing accuracy. An example of a system with a larger number of' units, and hence a fairly smooth cumulative capacity outage probability characteristic, is shown in Figure 8-1. 8-3 " 1.0 0.1 CUMULATIVE PROBABILITY OF MW OR MORE ON 0.01 OUTAGE 0.001 0.0001 MW CAPACITY OR MORE ON OUTAGE Figure 8-1. Example of a System's Cumulative Outage Probability If the load demand is known for a particular hour and the installed capacity is known, the LOLP can be calculated. As shown in Figure 8-2, the reserves are obtained by subtracting load from capacity. On this basis, a deficiency in available capacity (i.e., loss of. load) occurs if the capacity on outage exceeds the reserves. Ti1e probability of this outage is read directly from the cumulative capacity outage table, and is the LOLP for one hour. The annual LOLP is the summation, which thereby becomes an expected value, of the hourly probabilities. Conventional utility practices analyze the weekday peak hourly loads only (260 in all). Although, in the past, computer running time was a major factor considered in the selection of this approach, the current use of only the weekday peak hours for calculating daily LOLP is based upon several technical considerations. First, the probabilities vary exponentially with load changes. Off-peak loads of less than 90% of the daily peak load will generally add less than one percent to the LOLP risk. Second, generation outages usually tend to persist for at least one day. Third, the interpretations of other utility personnel, particularly system operators, are more meaningful when expressed in terms of days/year rather than in hours/year of expected problems. Simply dividing the hours/year by 24 will seriously misstate the actual number of days/year. 8-4 • r· l i.l t .. ~ ' L CUMULATIVE OUTAGE PROBABILITY OR GREATER I. 0000 u MW w;wl 0. 6342 10 INSTALLED CAPACI7Y 0.3 7 19 20 0.2463 30 .,.\ 0.1986 4·0 RESERVES 0 0 ..,--- 0 0 I 0 0 I I 0 0 I CNE' 0 0 YEAFt I I DAYS ~ DAYS/ YEAR = L DAILY OUTAGE PROBABILITIES Figure 8-2. Example of the Daily LOLP Calculation Procedure The preceding discussion is based on the assumption that the hourly demand was specified deterministically. The loads, which are convolved with the cumulative capacity outage tacle 1 result from modifying the original system loads to tteflect contracts, conventional hydro and energy storage. Contracts are assumed to be firm; that is, purchases reduce the load and sales add to the load. After contract modif5.cation, conventional hydro is scheduled in a peak shaving mode recognizing the derated capacity and energy limitations. Finally, energy storage units are scheduled to minimize system LOLP, using a derated generator rating, derated cha~ge rating and a derated maximum storage capability. The inclusion of load forecasting uncertainty is easily integrated into the computational procedure. First, the LOLP is calculated at each demand point in the uncertainty distribution. The equivalent is then determined by weighting the LOLP result at each demand point with the probability distribution value. 8-5 By utilizing the LOLP technique, syste:m planners can design the generation system to a specified level of reliability. As the demand increase;s with time, generation additions are automatically scheduled by OGP so the LOLP doss not exceed the design criterion. Figure 8-3 illustrates LOLP plotted versus the annual peak load for a specific generation system. Since the graph is almost a straight line on a semi-log basis, one can see that LOLP varies exponentially with load changes. The design criterion used in this example is 0.1. Based on the peak load for 1985 indicated on the graph, the generation system is able to meet the 1985 load at a reliability level better than 0.1. Therefore, no additional capacity is required. 10 ORIGINAL ysYSTEM / +I NEW UNIT / + 2 NEW UNITS 1.0 I I I I / + 3 NEW UNITS '/ +4 NEW UNI1"S LO.LP. 0,1 / -~ f CRITERION I I I I I I I I I : I I I I I II I I NEW UNIT I I 11 1 EFFECTIVE J LOAD CARRYING .01 I I CAPABILITY I I I I I I I + + I ~ I~ I I I 1986 1987 ANNUAL PEAK LOAD(MW) Figure 8-3. Example of the Automatic Unit Addition Process 8-6 • ' ' ' r ,, ~ . ' ·' I I j t .t ... 1: • I; ~ l f I ! I E t ~ • ' In 1986, the annual peak load growth has increased the peak demand to a point where the generation system cannot maintain the desired LOLP. In anticipation of this, OGP would schedule a unit addition for 1986. 'I'he MW excess of load, indicated by the bracket on the graph, is the difference between the 1986 peak load and the system's load carrying capability at the desired LOLP before any new units are added. With the installation of the additional new unit, the curve shifts to the right. In 1986, the LOLP has decreased with the new unit addition~ but has not yet fallen below the example's design criterion of 0.1. Thus, a second unit is required. As indicated on the graph, the addition of the second unit causes the LOLP to fall below the desired level. A similar occurrence is exhibited for the planning process in 1987. It is also interesting to note how the effective load carrying capability of each unit is measured. As shown by the brackets between the curves for 1986 and 1987, this capability is the: difference in MW, measured at 0.1 LOLP, between the annual peak loads that can be supported with and without the additon of each new unit. If the user is designing the system to meet a certain percentage of installed generation reserve, the automat.._ c addition process proceeds straightforwardly based only on the ratings of the units on the system and the load model specified. The calculations of the percentage of installed generation reserve are based on the peak load of a specified month. They are referenced to the maximum specified ratings on all generating units. The percentage of reserve can be calculated via one of the following three approaches: • [(Capacity + Contracts) -Load ]/Load • [(Capacity + Other Contracts -(Load -0 Hour Contracts)]/(Load -0 Hour Contracts) • [Capacity -(Load -Contracts)]/(Load -Contracts) Percentage of installed gene..1ration reserve is also calculated for reference, even if LOLP ls the design criterign being used. When mul ti-oompany or regional studies are being conducted in which the j,dentity of more than one area has been retained within OGP by the tlSer t s data L.J specifications, automatic additions will be distributed among all of the different companies as smoothly as possible. The only exception is for areas which may be restricted to having no generation. Automatic addition units are assigned to a specific company to maintain an approximately equal pereentage of installed reserves and percentage of peaking capacity for each company. t .J li L; yJhen these two items are a tradeoff between the companies, an economic comparison is made. 8-7 Wnen using OGP-6A, there may be energy-or fuel-limited units. For daily LOLP calculation, limited units are checked each month to determine if they could operate at full load for all the weekdays in the month. Units which pass this test are treated as usual. The units which fail this test are not included in the daily LOLP calculation for the month in question. Limited units must have enough energy to operate at full load during all the hours in the month in order to reduee the hourly LOLP for that month. 8-8 r~ l . i.! ,; r, '. f l,.> r: l\.;j •• 3 rr " t, rr· l1 t, f(' ' ~ '·' 11 *•A t h IL~ ~.' l ; ~J .1 ,.. it ~ L ~ # L ~ il \,; t, l L l !I l. I 'r p t· 1~ l 1 ~ I l . . I L I ~ ' '*' RELIABILITY EVALUATION SUPPLEMENTARY INFORMATION 1. Load Shape Modeling for LOLP Calculations, R.W. Moisan, 1972 GE Memorandum. 2. Calculation Procedure for LOLP, L.L. Garver, R.W. Moisan, 1972 GE Memorandum. 3. Adjusting Maintenance Schedules to Levelize Risk, L.L. Garver, 1972 Winter Power Meeting. 4o Analysis of Partial Outages, R.W. Moisan, 1974 GE Memorandum. 5. Computing the Loss-of-Load Probability, 1972 GE Memorandum. 6. Effective Load Carrying Capability of Generating Units, L.L. Garver 1 1966 Winter Power Meeting. 7. Loss-of-Load Probability (LOLP), G.A. Jordan, 1977 GE Memorandum. 8. Generation Reserve Value of Interconnections, A.M. Adamson, A.L. Desell, L.L. Garver, 1976 IEEE Summer Power Meeting. 9. Determining Interconnection Benefits Using Single-Area, Loss-of-Load Probabilit-y Studies, L.L·~ Garver, 1974 GE Memorandum. 10.; The Cost Benefits of Alternative Generation Reserve Levels, M.H. Bensky, H.G. Stoll, R.S. Szczepanski, R.E. Usher, 1978 American Power Conference. 8-9 c ;1 ......... _: ' ' I I I I I I. I ,',.. t','. PRODUCTION COSTS In OGP, the fuel and related operating and maintenanc~ costs are determined by an hourly simulation of the system's operation. Until 1971, deterministic models were used almost exclusively to estimate overall fuel costs by simulating the operation of individual units. Today, with the availability of large, efficient. digital computers, utility system planners can utilize production costing programs which simulate the actual system operation on an hourly basis. Although this calculation involves a significant amount of computer time, these simulation programs allow planners to completely investigate unit performance and various system operating strategies. The basic production simulation model performs various analytical functions required to simulate generation system operations during the OGP study. Although the production simulation is performed on an hourly basis, the routines ar·e designed to determine monthly and annual electric power generation operation expenses consisting of fuel and operating and maintenance expenses. In addition, OGP-6A can determine the operational characteristics of the generating system with respect to various environmental effects. The user has the option of biasing or overriding the normal, unconstrained, economically determined unit commitment and dispatch. This is accomplished by specifying weighting factors for various environmentally related quantities which will direct the program to operate units such that their environmental impact will be minimized. This capability is addressed separately in Section 10. The operational simulation for both OGP-6 and OGP-6A first accesses the Load Model. For each month, the number of weekdays and weekend days within that month is specified. As previously described in Section 3, the Load Model contains twenty-four hourly loads for each typical weekday and weekend day of every monthe (Refer to Figure 3-4 for an example.) The basic sequential functions of the operational simulation strategy are outlined in the following six steps: • Determine load modification based on recognition of contractual purchases and sales (i.e., reflect firm contracts). • Schedule conventional hydro. • Schedule monthly thermal unit maintenance based on planned outage rates or input manual maintenance. • Schedule pumped storage hydro or other types of energy storage. • Commit thermal generating units to serve the remaining loads based on economics or environmental factors, spiiming reserve rules, and unit cycling capabilities. • Dispatch the generation based on relative production costs and environmental emissions specified by the user. 9-1 • The production simulation performed is for a total utility system or pool commitment and dispatch assumed to have an unlimited power transfer .capability between areas or companies internal to the pool represented. Since the user is not required to input or otherwise predetermine a loading order or sequence of unit commitment and/or dispatch, the user is relieved of the responsibility for this complex and error prone calculation. OGP automatically determines the ideal loading order for every commitment and dispatch period of the study at the time it is first needed. This section describes how OGP follows the six steps outlined above to determine production costs. It also discusses the commitment and dispatch of units with fuel or energy limits. 1. Purchases and Sales The hourly loads are initially modified by OGP to consider the firm purchases and sales that exist between the area being studied and entities outside that area. A purchase is subtracted from the Load Model for the number of' hours specified in the input. A sale adds to the Load Model. This concept is illustrated in Figure 9-1. The specified schedule and cost of the purchases and sales may diff'er for each contract. The demand and energy charges will be determined separately. Also, before proceeding to the next step, OGP resorts the resulting Load Model. c <t 0 -I 0 4 WEEKDAY 8 12 16 20 24 HOUR 4 WEEKEND DAY MODIFIED LOAD 8 12 16 HOUR 20 24 Figure 9-1. Example of Load Modification Based on Firm Contracts 9-2 .l 2. I I I I ·1:, 1 .. I ' . I. I I I ~ ' •."''·· Conventional Hydro Scheduling Hydroelectric energy is assumed to have an incremental fuel cost of zero, and is scheduled to maximize its beneficial effect upon system operating costs. There are generally t\-JO types of conventional hydro. The first, run-of-river hydro, is typically an installation which has minimal storag~ and probably a low head. Units in this type of installation tend to be base loaded, because the river flow requi~ements and dam characteristics dictate that tne unit must be operating mosb of the time. The second form of conventional hydro is the pondage or si.'llple storage hydro. Units in these installations are usually scheduled during peak load time periods because the system's incremental fuel cost i~ the highest at these times. If the pondage hydro is scheduled to shave peaks, it maximizes its effect on system operating costs~ A sample schedule of both run-of-river and pond age hydro is provided in Figure 9-2. The rur.-of-river energy that must be produced by this type of hydro unit is accounted for by subtracting a constant capacity from every hourly load in the month as shown on the graph. This capacity value is referred to as the plant minimum rating and is provided as input data. After run-of-river energy is used, there may be remaining energy, \vhich can be used for peak shaving. In such situations, the program uses the remaining capacity and energy of the hydro unit to reduce the peak loads as much as possible. If any excess energy exists at the end of a month, a user-specified maximum storage amount can be carried forward into the next month. p2-l T WEEKDAY MODIFIED LOAD HOUR INITIAL LOAD WEEKEND DAY P 1 = MINIMUM RATING (M W) P _1 _ MAXIMUM MINUS 2.., -MINIMUM RATING (M W) HOUR Figure 9-2. Example of Conventi.onal Hydro Operations f 0 3. Thermal Unit Maintenance Maintenance schedules designed to account for planned downtime, due t.o activities such as repairs or refueling, are developed by OGP for eac:h generating unit based on user-specified planned outage rates (PORI. Increased maintenance levels, which might be required during the fir:3t several years of a unit's operation or during its shakedown period, a !"e modeled using an immature multiplier [e.g., Immature POR = (1.15) (Matur-e POR)]. Often the planned maintenance of individual thermal units on a utility system is scheduled on a monthly basis. During these scheduled maintenance periods, an individual generating unit is unavailable for energy production. Planned maintenance is normally scheduled to minimi.ze its effect on both system reliability and system operating costs. The levelized available reserves approach is one strategy commonly used to schedQle maintenance. With this approach, the peak loads are examir1ed throughout the year, and individual generating units are scheduled in an attempt to levelize the peak load plus capacity on maintena·1ce throughout the year. The starting point for implementation of this approach is during the months where peaks are at their lowest (i.e., valley month,:;). The ·user can specify a maximum of five months during which maintenance is not to be allowed. The illustrat.ion in Figure 9-3 represents an annual OGP-derived maintenance schedule for a particular utility system. The shaded area indicates the total amount of capacity on maintenance for each month. Thus the generating units available for service are identified for each point in time. If a prespecified maintenance pattern has been input for any or all units, those will be scheduled first. Any remaining units ~ill automatically be scheduled by the OGP program. Thermal generating units are scheduled for maintenance by OGP for an integer number of months. This assumption is reasonable for large base load capacity, but tends to be less accurate for smaller sized mid-range and peaking .capacity. Based on this assumption, and the user-specif'ied planned outage rate t'or each generating unit, a target megawatt-months of planned maintenance is calcul·ated for each unit, and the units nre maintained for the nearest whole number of months. The actual megawatt-months of maintenance may differ from the target level; i.e. , a fractional megawatt-month of residual maintenance may exist. Ttis residual can be either positive (not enough maintenance was done) or negative (too much was done). When this occur·s, the program applies that residual to the next unit scheduled for maintenance, and includes the residual in its maintenance calculation. Residuals are carried over onLy for units of the same type of generation. The residual maintenance f.)r the last unit in each type of generation is used to derate that unit. The overall effect of the residual calculation is to ensure that, for ea<~h t:y;:'e of generation, the correct amount of megawatt-months of maintenanc:e is so:-:...,.duled, even though the scheduled maintenance for an indi vi'dual unj t may vary slightly from that actually desired. 9-4 ( I I I I I I I I I INSTALLED CAPACITY MONTHLY PEAK LOADS ~--L-~,1 JANUARY JUNE JULY MONTH DECEMBER Figure 9-3. Example of Maintenance Scheduling 'rhe user may specify that a maintenance schedule from a. previously run OGP simulation be used. It is necessary that the first run save this schedule on a file, and that both cases be manual expansions (i.e., there are no automatic additions) with the same generating units. It should be noted that the maintenance scheduling algorithm used for production costing differs somewhat from that which is used for the system reliability analysis discussed in Section 8. Although~ levelization is still the criterion, risk, rather than reserves, is levelized when the LOLP calculation is performed. 4. Pumped-Storage Hydro Scheduling The system operating conditions involved when pumped-storage hydro or other energy storage devices exist on the system must also be considered •. Energy storage scheduling algorithms have been included in production costing programs for some time. Although the devices studied are usually referred to as pumped-storage hydro, these algorithms have been utilized to study other ene~"gy storage devices on electric utility systems such as batteries, thermal storage, etc~ The dispatch of energy, storage units is scheduled to minimize the total sy·stem fuel costs during a specified time period. OGP recognizes losses in the cycle as the program schedules generation and charging energy to maximize the system fuel cost savings. The user can specify that the scheduling be done on a daily or weekly basis. Energy storage units are assumed to be fully charged at. the start of a week, and incremental 9·-5 1' n , i~ 0. ('· amounts of generation are balanced by enough charging to fully recharge the unit before the start of the next week. Since system fuel cost tradeoffs are an integral part of energy storage scheduling, a specification of the system fuel cost in $/hr as a function of the system's hourly thermal megawatt output is required. This specification is derived from the cost characteristics of the individual thermal generating units. Because of the nonlinearity of system operating costs, operation of the pumped-storage hydro unit can save fuel dollars, despite a cycle efficiency of less than 100 percent. 0 WEEKDAYS \wEEKDAY / GENERATING 4 -~ WEEKDAY/ PUMPING a 12. ts HOUR 2.0 2.4 WEEKEND DAYS WEEKEND GENERATiNG / 4 WEEKEND/ PUMPING 8 12 IS HOUR 20 24 Figure 9-4. Example of Energy Storage Scheduling on a Weekly Basis As shown in Figure 9-4, OGP's basic scheduling approach for energy storage devices is to do the following: 1. Start with the highest load, i.e., the load which is the costliest to serve during the week. 2. Schedule one megawatt-houi"' of generation. 3. During the lowest load in thE~ day (if on a daily refill cycle) or week (if on a weekly refill C:!ycle), schedule enough charging energy to replace the one megat.ratt-hc,ur of generation plus the losses in the cycle. Thus, for ever-y megawatt-hour of generation, there must be a correspondingly greater· number• of megawatt-hours of pumping. 9-6 . 6-~---c--~-------""7·---------------------------·----------------------------------------------- I ' r ,. I :: I I I. 'I· ~ : ' ' -'/It'_-...,., ;f_ :P:. 4. The fuel cost savings provided by the megawatt-hours of generation are then compared with the increased cost of the megawatt-hours of pumping. The fuel and variable O&M costs of the storage device are included in the pumping costs. If the savings exceed the costs, the process is continued. During the scheduling of the energy storage devices, one of two conditions will limit the amount of energy storage operation. The first is when the incremental savings balance the increased cost, causing additional operation to be no longer economically beneficial. The second occurs when the physical limits of the storage reservoir are reached. The storage reservoir conditions are being monitored while the iterative scheduling of the storage is in prog~ess. The schedUling will not violate the minimum or maximum reservoir level of the unit anytime during the week. 5. Thermal Unit Commitment After modifications for contracts, hydro, and energy storage operation have been made, the remaining loads must be served by the thermal units on the system. The cost characteristics of thermal generating units are modeled, using a single incremental heat rate. This yields a single incremental cost curve as illustrated in Figure 9-5. Specific unit operating costs are determined by the fuel input curve, fuel cost and variable O&M cost. In order to minimize the thermal generating unit operating expense of a power system, two fundamental objectives must be met: (1) the number of units committed each hour should be minimized, subject to the commitment policy and operating constraints of the power system, and (2) the generating units in each commitment, as determined for the first objective, should be dispatched on an equal incremental cost basis. Since system production costs are exti"emely sensitive to variations in unit commitment, it is essential that the unit commitment policy of the power system be fully considered¢ In addition, most, if not all, production costing algorithms used in the electric utility industry dispatch generating units on an equal incremental cost basis. However, dispatching generating units on an equal incremental cost basis w1~a1n a zone of constant commitment will minimize production costs only with respect to the units included in the commitment. If the zone commitment has not been minimized with regard to the commitment policy, the zone production cost will not be minimized. Based on discussions with utility system planners and experience with large-scale production costing programs, three commitment conditions have been found to prevail. Night time periods generally have one commitment because the cycling of units during the night is avoided, if possible. Second, a generating unit committed to peaking service for a specified hour generally remains on line for at least four hours. Finally, commitment variations during weekends tend to be minimal. 9-7 i FUEL INPUT (Mbtu) · hr i INCREMENTAL FUEL INPUT (~) MWh MINIMUM '\,_MAXIMUM MW OUTPUT i FUEL INPUT ( ~r ) I i INCREMENTAL FUEL INPUT $ ( MWh) I Figure 9-5. Example of Thermal Unit Cost Characteristics Based on these three observations, OGP was developed to accommodate six zones. As shown in Figure 9-6, each weekday has four zones of constant commitment and, each lveekend has two zones of constant commitment. The user can define the duration of these six zones. Typically for weekdays, there will be three four-hour zones during the day and a twelve-hour zone during the night, whereas for weekend days there will be two twelve-hour zones. 9-8 • i ra r \ ., t ( J < ' l I 1 I ' t [ l t 'I f I t ! l ; t ,, t L ~ t. ~ C-,. ll L n L t~ !I b l~ 0 c:t g ~ ~ 0 WEEKDAYS I I I I COMMITMENT ZONES I 2 .3 4 4 a 12 16 HOUR 24 WEEKEND DAYS FINAL COMMITMENT --------, PRELIMINARY IV'coMMITMENT 4 L------- COMMITMENT ZONES I s 1 s a I 12 HOUR Figure 9-6~ Example of the Thermal Unit Commitment Process 24 Unit commitment determines how many units will be on line each hour, and attempts to provide an adequate level of operating reliability, while at the same time, minimizing system operating costs. The reliability requirement is addressed by committing enough generation on line to meet the load plus a spinning reserve margin. This spinning reserve margin protects the system from units suddenly tripping off line or from t.te lines opening. The economic aspect of the reliability requirement is addressed by committing units in the order of their full-load energy costs. For example, the least expensive units are committed first, and additional units are scheduled on line until enough generating capacity is available to meet the load and spinning 1reserve margin. Once generation has been scheduled to meet the load plus spinning reserve margin, the preliminary commitment is complete. This commitment is preliminary because it requires specific generating units to be shut down in the middle of the night and turned on again the next day. During peak hours, some units are shut down on an hourly basis. Yet, as discussed earlier, the generating units may not have the ability to cycle as required by the preliminary commitment. The preliminary commitment is then reviewed to determine if any unit's cycling rules or capabilities have been violated. If there is a violation, the preliminary commitment will be increased in order to keep the unit on line 9-9 l·, during the problem hours. This observation of the minimum downtime rules for individual generating units allows the preliminary commitment to become the final commitment. The user also has the option of basing the commitment order on both econoreics and minimum uptime rulesG When this option is used, the Rule 1 units are committed first, then the Rule 2 units and finally, the Rule 3 units. Within each uptime rule, the commitment order is based on full-load fuel costs plus variable operation and maintenance costs. 6. Thermal Unit Dispatch If a unit is committed, its output must be equal to or greater than its minimum loading level. When the final commitment has been established, all the units' minimum loads will be scheduled first. Typically, the sum of the minimums does not equal the load. The remaining load will be served by the units' incremental loading sections as shown in Figure 9-7. PRIORITY FINAL LIST COMMITMENT DETERMINISTIC DISPATCH FULL OUTPUT 4 2 Fi.gure 9-7. Example of Thermal Unit Dispatch SYSTEM LOAD The dispatching function in a production costj.ng program loads the incremental sections of the committed units in order to serve the demand at minimum system fuel cost. This dispatch technique is referred to as the equal incremental cost approach (or m~n~mum incremental cost approach). The incremental loading sections are dispatched beginning with the least expensive unit. When enough incremental loading sections have been scheduled so the load is served, the remaining unloaded incremental sections will be the most expensive.. Thus, the system spinning reserve margin is allocated to the generating units so system fuel costs are minimized. 9-10 . :~1-~"' -~----------------------------. -------- ··· r L .. · 1: I I:, l I~ 1; __ I 1:" r f: I I . ~ I .. I \:i 1: . I t· ,, l -: ,. ,f --. ~·~ Jt J. ~~ ~: ..:..~ At this point, loading levels on the individual generating units are established. The hourly energy disposition is scheduled, and the hourly production cost is determined for each unit. The thermal dispatching function for the system utilizes the incremental heat rate curve, an additional piece of performance information available from the input-output curve. As shown in Figure 9-5, the incremental heat rate curve is ~ partial derivative of the input-output curve with respect to power output. As with the input-output curve, the incremental heat rate is transformed to an incremental fuel cost curve when it is multiplied by the fuel price • The random forced outage option illustrated in Figure 9-8 is a technique for simulating the effects of forced outages on system operation. The program commits units and arranges them according to their incremental cost, beginning with the least expensive unit.. The committed units are sequentially dispatched until each load has been met. However, as the units are dispatched, the technique recognizes that each unit will be out of service for a period of time proportional to its forced outage rate .. During these outages, load is transferred to more expensive generating units. With full consideration of all possible combinations of forced outages in the system, via modified recursive convolution, the program then computes the expected dispatch for each generating unit. STOCHASTIC DISPATCH ORIGINAL OISPA1CH -~---.----4 3 2 .....__-~----- TIE 5 4 3 2 I REPLACEMENT //ENERGY ONE HOUR ORIGINAL LOAD Figure 9-8. Example of the Effect of Random Forced Outages Also considered in these calculations, with respect to the generating units that originally were not committed, are operating policy constraints, relative incremental costs and spinning reserve requirements. Once a combination of forced outages requires a unit not originally committed to be placed into service, the unit. remains in service for the number of hours it is needed to overcome the capacity 9-11 shortage. Likewise, when a combination of forced outages creates a spinning reserve violation, the program will bring additio;'lal units into service to provide a sufficient amount of additional capacity to remove the violation and maintain spinning reserve. Another option for including forced outage rates in production costing is based on a deterministic technique in which the period of forced outage is added to planned maintenance. When systems are large enough to permit maintenance throughout the year, this procedure of extending each unit's maintenance in proportion to the unit's forced outage rate yields production cost results that are very close to those yielded by the stochastic option just described. Since the extended maintenance approach is a less complicated method of treating forced outages, it has the advantage of requiring less computer processing time.. It is recommended that the user specify the application of the stochastic technique only in the final calculations for each year rather than for every decision triale Through the execution of the production simulation, all hourly fuel and O&M costs for the individual units, energy and demand charges from purchases and sales~ and nuclear fuel inventory charges are accumulated and totaled on a monthly basis. The user has the option of specifying monthly or annual output for the optimum system only or for all trials evaluated. Other quantities which are also available unit by unit include maintenance months, energy output, hours on line, capacity factors, and $/MWh for the total of the fuel and O&M ,costs. Capacity factors are calculated on an annual basis when annual results are printed, and on a monthly basis when monthly output is obtained. Other output quantities, such as fuel consumption, are summarized by type of generation and fuel as well as on a system-wide basis. Energy from contracts, hydro, and energy storage devices is also shown. All production simulation results may be stored on a separate file. The user can later access this file to perform additional calculations based on the production simulation results or to reformat the output. 7. Commitment and Dispatch with Fuel and/or Energy limitations The OGP-6A program has an option which allows the user to specify that all units within a thermal type or all units that burn a specific kind of fuel must operate in a limited mode. The monthly energy limitation may be input as MWh or as capacity factor. The fuel limitation is specified in physipal units of fuel sqch as barrels of oil or tons of coal. If a unit has both a fuel and energy restriction, a comparison of the two limits is made, based on the assumption that the unit is operating at full load. The more severe limit is used to commit and dispatch the unit. A limited unit assigned to minimum uptime Rule 3 (refer to Table 5-1 for an explanation of each uptime rule) is tested to its limit before being committed to each commitment zone. If the fuel or energy limitation is not a factor, assuming full-load operation for commitment zone one, the unit is committed and dispatched in the usual manner. Before a unit is 9-12 ·r.=·.~.--. ~·~-------·--·~----~·--····-··-··-----·-···------.. -~ .. I ! ' J ~t" . ..... '"' ·~"~· (' r f ~ ! ~ f, i "' I I. .. · ~. ~ I I 1 ~; 'ir I . ·• I '.5 I. I . .. I, J I -: I ~· -~ -·! l.· J; ~~ t committed and dispatched to zone two, the monthly fuel or energy allotment for the unit is decreased by the amount of energy or fuel actually used in commitment zone one, rather than according to the assumption of full-load operation for commitment zone one. This process is repeated for subsequent commitment zones until there is insufficient fuel or energy for the unit to operate at full load for the commitment zone under consideration. Units that have been assigned to uptime Rule 2 must have sufficient energy or fuel to operate at full load for all weekdays during the month, or they are excluded from further consideration. Those units which pass this test are committed and dispatched for the weekdays, as long as no fuel or energy limitations apply. The monthly fuel or energy limit is then reduced by the actual weekday usage. Next, these units are tested to determine if there is sufficient fuel or energy remaining to operate at full load during all weekend days in the month. If so, the units are committed and dispatched in the usual manner. If not, the units are prohibited from weekend operation. Finally, units assigned to uptime Rule 1 must have sufficient energy or fuel to operate at full output for all days in the month.. Units that do not satisfy this condition are prohibited from operating at all during the. month • If one assumes that weekday commitment zones contribute more to total production costs than weekend zones, this procedure will maximize the economic benefit of a limite~ unit. Furthermore, although assuming full-load operation for a unit may be a poor method of estimating its actual operation, there is an option which allows each month's unused fuel or energy to be carried forward for use the next month. Energy or fuel residuals may not be carried forward to the next year. 9-13 PRODUCTION COSTS SUPPlEtv1ENTARY INFORMATIOI""~ 1. Commitment and Minimum Uptime Rules, 1913 GE Memorandum. 2. OGP-5 Modified Reoursive Convolution Eff'aetsl D.L. Dees, 1978 GE. Memorandum. 9-14 i " ' ·0 'I I . ' 'I. j ' t "' I· t : I l : ~ -· I 1-· I . . I I J l J I ~ ' - ~; ~ ' I!; ~i I ·' ENVIRONMENTAL IMPACTS Section 9 discussed the program logic that simulates the operation of the generation system. As stated earlier, this simulation is done to minimize the total system cost of Serving the lO.ad. In the envj,ronmental option of the OGP-6A version of the program~ the factors listed in Table 10-1 may also be considered in the logic. In addition to calculating the environmental quantities in physical units, OGP also has the flexibility to commit and dispatch units to minimize operating costs, emissions, or a weighted sum of costs a~'.ld emissions. TABLE 10-1 Environme~tal Factors Units Heat rejection into the atmosphere MBtu Heat rejection into the cooling m~~diuni MBtu so2 emissions Tons NOx emissions Tons CO emissions Tons Particulate emissions Tons Water consumption Thousands of gallons Through use of the fuel and/or energy limiting option for generating units in the OGP production simulation, the absolute level of all the quantities listed in Table 10-1 can also be controlled. This control can be accomplished by precalculating the hours of operation or the amount of fuel to be consumed based on the quantity in Table 10-1 being limited, and then inputting that value into the program. Eight weighting coefficients to be applied to unit economics plus the seven factors listed in Table 10-1, can be input into the program. A unique commitment number will then be calculated for each unit in the system. This commitment number is the sum of the weighting coefficients times their corresponding quantities. The units are then committed on the basis of lowest commitment number, subject, as stated earlier, to their minimum uptime rules. Therefore, to achieve an accurate environmental commitment of units, the original emission factors should accurately predict the emissions at full load. In the environmental dispatch logic, the same eight factors may be considered, but the incremental values are used rather than the full-load values. The incremental value is the slope of the line passing through the curve at the minimum and maximum rating of the unit. The dispatch logic is similar to the commitment logic. Thus a set of eight weighting coefficients 10-1 1,; I i: l' I I! is input into the program, and a unique incremental dispatch number is calculated for each unit. The units are then loaded to their maximum rating on 'the basis of lowest dispatch numbere In the areas of emission calculations, etlvironmental commitment of units and environmental dispatch of units, the greatest accuracy will be obtained when the emission curve is approximated by a straight line which passes through the actual emission curve at the minimum and maximum rating of the unit as shown in Figure 10-1. The values input into the program are the slope and intercept of this straight line o They are determined using the following equations: SLOPE (lbs/MWh) = (lbs/hr)max -(lbs/hr) min MW -MW . max mJ.n INTERCEPT (lbs/hr) = (lbs/hr)min -(SLOPE)(MWmin) EMISSIQ\JS (LBS/HR) CORRECT AT END POINTS MIN OUTPUT ( MW) MAX ACTUAL EMISSIONS Figure 10-1. Unit Emissions Characteristics and Representation Since a single step incremental representation of the units is used in OGP' s dispatch logic, all of the units committed during a given hour except the "swing unit" will be operating either at their minimum or maximum'rating. The.refore, it is more important to accurately predict the emissions at the end. points rather than throughout the length of the curve. For a typical system, the one swing unit during each hour represents only a small part of the total 10-2 , .. r 'I r rl I ' L " 11 I . . I system output, so an error in the emissions of this unit will cause only a slight effect on the total emissions of the system. For larger systems, the effect of the swing unit will be negligible. Therefore, only the end points of the emission curve, not the entire curve 1 are required to obtain a high degree of accuracy in the environmental output. The user may assign indiyidual generating units to plants, and subsequently may assign plants to regions within the total company or pool-wide area being. studied. Thus, by judiciously assigning specific weighting coefficients, which are input by region, the operation of specific groups of generating units can be controlled. Regardless of whether the unit commitment and dispatch is being biased by the environmental factors, the user may also obtain summaries on fuel consumption and environmental emissions in physical units such as barrels of oil or tons of S02. These. summaries are available by unit, plant, region, and/or fuel type. For consistency, the user also can r"3flect the cost of various emission control equipment via other inputs, such as O&M and plant cost data. 10-3 ENVIRONMENTAL IMPAGS SUPPlEMENTARY INFORMATION 1. Fossil Fuel Emissions, G.Ao Jordan, 1975 GE Memorandum. 2. Calculation of Emission Factors, G.A. Jordan and R.B. Roginska, 1975 GE Memorandum. 3~ Environmental Input Data, G.A. Jordan, 1975 GE Memorandum. 4~ Cost.s of' Meeting Clean Air Requirements, D.R. Vierath, w.w. Walkley, September, 1976 Power Engineering. 10 .. 4 r. r ,, ~'.1 ~ : ·' ' '- :I ~ I !I [~ Pt INVESTMENT COSTS The investment cost routine computes the total capitalized investment costs at the time of start-up or :i.nitial commercial operation for units added to the system. Then, based on a levelized fixed charge rate (FCR), the routine calculates the annual investment cost in terms of carrying charges on investment. Installation costs will vary according to unit size via use of the "D" factor, company. Ol" area location, percentage ownership, type of unit, and, because of inflation, year of installation~ The levelized FCR used for each unit addition may also vary, depending on the type of unit and company ownership. The annual capital investment cost portion of the total system costs for each generating unit is determined with the following equation: Capital, $/yr = ($/kW)(kW)(FCR)(% Ownership)(% of Year in Service) This equation represents the revenue requirements approach for determining the impact of capital expenditures on the total system costs. Also included with the investment costs are the demand charges associated · with the contracts specified to OGP. These charges are calculated for each contract with the following equation: 12 Demand Charges, $/yr = L (kW ratin~ in month i)(S/kW/yr)/12 i=l At this point, the following costs are available on an annual basis: fuel (both variable and inventory), O&M (both fixed and variable), and contracts (both demand and energy). The capital component is added to these costs. Then the cumulative present worth of all revenue requirements is obtained directly to enable alternative generation expansion plans to be compared. 11-1 I i ' .,. :;- I . ' I. . . I OPTIMIZATION RESUlTS This section includes a discussion of the development of an optimum generation expansion plan regarding the mixture of generation types. It also provides a more detailed description of tha optimization methods used in OGP. In this section, the normal connotation of the term "optimum" is used (i.e., an optimum plan has the lowest total cumulative present worth of the sum of annual charges on investment, fuel, and operation and maintenance costs during the expansion period). It should be noted that other objective and subjective criteria also can be used to reach generation expansion decisions. For example~ one of those alternative ytt:'dsticks could focus on environmental effects.. Others could disregard any impacts of the expansion plan on the system operation and costs that may occur after the period under study. OGP addresses non-economic aspects, and can be used to ensure that the future has not peen mortgaged solely to optimize a single addition • The cost characteristics of the generating units on a utility system make it po~sible to cost effectively serve a spectrum of loads, ranging from base loads, which have an annual duration of 100 percent, to peak loads, which have a duration approaching zero. There are infinite gradations between these two basic types of loads which, for convenience, are referred to as mid-range loads. Each MW of base load requires the generation of & maximum amount of energy; thus fuel cost per MWh is of ma.jor importance to the total system costs. Conversely, peak load MW' s require minimum amounts of energy, which means fuel cost is of minimum importance. In contrast to fuel cost, the capital cost of generating units affects the total system cost equally, whether the units are used for base, mid-range or peaking loads. This contrast provides the opportunity for OGP to minimize the system's total sum of fixed cha...,ges and fuel cost if types of ger..erating units with varying capital costs and fuel costs per MWh can be selected to meet the requirements of the load, resulting in a mixed pattern generation system. In recent years, types of generation have been developed which make mixed patterns possible as well as economical on a total syst.em cost basis. Thus the OGP optimization process essentially is one of determining the lowest cost mix of future units. The OGP program normally is not used to select optimum unit sizes.. From the allowable sizes of each of the alternate ty·pes of units available for selection in a given year, the program, based on the unit size guidelines input for each kind of gene~ation, chooses the size of each type of unit to be considered in the optimization that year. If the user wishes to more closely investigate unit size, alternate sizes may be considered for selection based on eoonomics by appropriate use of the six types of new thermal generation which may be represented. For example 1 one type of base load unit could be specified as 600 MW and another as 750 MW, thus allowing the OGP program to conduct a head-to-head economic comparison of the two ~~zes. For the sak~ of simplicity, the first situation that wi 11 be used to illustrate the OGP optimization process is one in which conditions are fixed. This means there is no inflation of fuel or capital costs; nuclear fuel costs do not change ·as core equilibrium is approached; and there is no immature 12-1 period of planned or forced unit outage rates. Under these conditions the OGP program will directly select, each year, the type. of generation that results in the lowest system cost, and this will be the optimum thirty-year expansion. To understand why this selection yields the optimum thirty-year expansion, one must rirst consider the cost ·trade-offs involved in selecting between the extremes of a base load or peak load generating unit to satisfy the load growth requirement of a system in a specified year. The amount of new generation needed is determined by a reliability calculation.. This calculation takes into consideration unit size1 as well as forced and planned outage rates. Since the increment of new load has a base component as well as peaking and mid-range components, and since it is known that some kind of mixed pattern will be the most economic;al alternative during the thir•ty-year study period, it is not obviol,!s which type of generation should be added to the system for the specified yea~. If a base load unit is added, its lower production cost will make it economical to operate, perhaps even to the unit's maximum availability. The amount of low-cost energy generated will be much more than that required by the new load increment and, therefore, will result in a significant decrease in the system's average fuel cost per MV1h compared to the previous year. On the other·band, if peaking units rather than base load units are added, the higher production cost of the peaking units will make it desirable to run them as little as possible and perhaps not at all, if the total amount of' peaking capacity on the system is less than the installed reserve. However, the addition of peaking units will resGlt in an increase in the system's average fuel cost per M"vlh compared t.o the previous year. This increase is due to the added increment of load energy that will have to be supplied by the existing types of generation at the higher production cost end of the spectrum. The amount of decrease or increase in the system's average fuel cost obtained with the addition of either type of generation in a given year depends on the mix of units in the existing system. For example, if, in the preceding year, the system was composed entirely of base load units, le·ss of a fuel cost benefit will accrue from the addition of another base load unit, and less of a fuel cos~ penalty will result from added peaking capacity. Balanced with these shifting system fuel costs is the difference in capital costs of the base load and peaking units, which may be assumed to be fixed, i.e., independent of system compositicm. The ·net effect of the fixed and variable cost components may favor ba.se load or peaking generation, depending on the composition of the system before the additi.on is made. Figure 12-1 illustrates this concept by showing how·the combination of capital cost difference and system average fuel cost determines the economic choice. of unit additions. In reality, this unit-selecting mechanism is a kind of economic control system equipped with negative feedback to prevent instability. For example, assume that point nAn represents the situation in the first year of a study. Since it is to the right of the dividing line, the decision that year would be to add base load generation. However, as a result of that decision, the system average fuel cost will have moved to the left in the following year. More than one year of base load additions may be r-equired, but eventually the system average cost will move across the dividing line to point '~B." At that point, the decision will be changed, and peaking 12-2 r .. '·' ,.I I .I I !I I .I ·t I I I ~ (/) 0 u ..... 1-<!l zz <t~ .....~-a.<t LIJ za. --wl u-zo wet a:o w...J t..W 14(1) -<t OQj >--~ ~ ' -<!> 600f I I 0 8 A _____ .....,._ _____ --·--------·------ 15 SYSTEM AVERAGE FUEL COST MILLS/KWH Figure 12-1. The Mechanism of Economic Choice 30 generation will be added to the system. Each year peaking generation is added, the system fuel cost moves toward the right, until the dividing line is again crossed. The system average fuel cost is novT high enough so the next decision to be made will be to add base load generation again. This simplified example illustrates an aspect of year-by-year optimization, but the OGP program does not calculate the economic merit of each alternative in this fashion. It rigorously performs a complete simulation each year for each type of generation to be tested. The decision regarding the type of generation to be added is based on the lowest total system cost rather than on the approximation a screening curve analysis would yield. Although the sum of a series of minimal annual costs should produce a minimum total for the expansion, planners must still contend with the question of whether a decision in an early year might compromise the future design in such a way as to force unnecessarily high costs in the later years of the expansion. In effect, planners must answer the question, "In 1995, would we wish we had done something differently in 1985?" 12-3 Eaoh year, enough new capacity is added to a system to accommodate the new load plus ~eserve* Thus as long as the units are properly maintained, their contr-ibutiou to the system's ability to reliably serve the load never diminishes~ The decision to add a particular type of generating ~nit one year does not mean planners are committed to add that same type of unit in future years. Furthermore, there is relatively little under-or overbuilding of the system. The requirement for continuous annual additions to the generating system means that there is also a continuous opportunity to adjust the mix of unit types without ever departing significantly from the absolute optimum in any particular year.. Y~.!ar-by-year optimization produces an optimum cumulative expansion, provided future cha:1ges in costs or outage rates do not occur. This statement has been demonstrated by numerous unsuccessful attempts to manually d~vise a less expensive expansion than that pt.,oduced by the OGP program .. Under conditions where assumptions regarding future changes in outage rates or costs, such as escalating fuel prices~ have been made, a similar s.ituation of not cleparting significantly from the optimum mix of units in any particular year will exist if OGP' s "look-ahead n feature is utilized in the optimization. The mechanisms by which levelized system costs and mature outage rates are utilized in yearly decisions have been described earl.ier in Section 7.. Figur~ 12-2 is a graphic description of the effects of the "look-ahead" process on yearly system costs. Experience with the OGP program indicates that, where inflation. and immature forced outage rates exist, the "look-ahead" feature produces a lower present worth of expansion costs. This result is obtained by permitting somewhat higher system costs in the early expansion years. The resultant savings in later years are more than enough to compensate for these higher system costs in the early years. Figure 12-2 plots the cost differences between a "non-look-ahead" and a "look-ahead" expansion case. Note the annual deficits for the "look-ahead" case for approximately the first five years, followed by a substantial savings, which was anticipated by the "look-ahead" decision logic. Of course, as time progresses, the "look-ahead" case also becomes less costly in terms of cumulative present worth. In summary, in the absence of changing cost and outage parameters, the inherent nature of the economic framework of the generation system makes year-by-year optimization feasible and correct. However, with parameters that change with time, the annual decision regarding the type of generation to be added should be biased by levelized fuel costs and mature outage rates in order to anticipate these changing parameters and produce an optimum expansion. 12-4 I I I I I I + P.W. SAVINGS OVER MINIMIZING YEARLY COST 0 ---YEARLY ----CUMULATIVE /' ------ / I I / /10 15 YEAR OF STUDY 20 Figure 12-2. Example of the Effects of a Ten-Year "Look-Aheadn Period 12-5 '· I I I I 'I· I I J . i EXPANSION OUTPUTS Output options have been designed and included in OGP to provide the user with flexibility in the level of detail and volume of documentation. received. Complete batch output reports as well as summary outputs are available. In addition to being included in the bulk output, the summary outputs may be obtalned at the user's time-sharing terminal. These remote summaries usually contain sufficient information to enable the user to make decisions and/or proceed to run the next case, when execution of the next case depends on the results of the previous one. Most of the OGP output can also be written to a file and stored for future analysis. This enables the user to reformat the output to meet specific needs, plot particular results, or compare or combine the results of several OGP runs. The output available from the OGP program includes the following information: 1. 2. 4. 5. Listing of the input data. Standard tables, as defined by the user, for various unit characteristics. Listing of the unit types and sizes available for optimization and their characteristics. Listing of the Load Model for the study period. Listing of the generating units on the system and their characteristicse 6. Year-by-year summary of the firm contracts input by the user. 8. Production simulation summaries, listing all of the generating units of the system with their energy output, fuel and O&M costs, fuel consumption, and environmental emissions. These summaries can be obtained on a monthly or annual basis, for all the decision passes or just the optimum system. Summary of all of the expansion alternatives, with their associated costs and reliability measures, evaluated during the optimization. 9. Summaries of the final system expansion through time and the associated costs. The "bottom line" result from the OGP program is the annual summary of additions.. Figures 13-1 and 13-2 present the annual capacity additions by type of generating plant (e.g., nuclear, coal, gas turbine, etc.). As shown in Figure 13-1, in the year 1995~ the OGP program added in this sample run one 1300 MW nuclear unit and one 300 MW block of gas turbines as well as 600 MW of pumped storage hydro. The generating units indicated with an asterisk (*) are those units which have been previously committed for service. For example, in 1986, a 500 MW compressed air energy storage unit was committed for service. 13-1 • ..._, .;: t'.•• At the bottom of the Annual Capacity Additions by Type report, a summary is provided$ The first row is the sum of megawatt additions (~rn ADD) during the period. The second row is the capacity in service in 2014 (end of the study). The third row is the MW additions that were added automatically (AUTO) by the OGP program (total additions less committed additions). Other summaries are also provided by the OGP program. Figure 13-3 presents the load, capacity, reserve, LOLP and cost summary. Figure 13-4 presents a more detailed cost summary both on a yearly basis and also on a cumulative present worth basis. OGP makes available more detailed yearly and monthly results as illustrated in Figure 13-5. This is the annual production cost summary and shows the annual history of each generating un~t's maintenance period, hours on line, capa~ity factor, fuel cost, etc. At the bottom of the report, the energy output, capacity fa.ctor, and fuel cost results are summarized by type of generating plant (e.g., nuclear, coal, gas turbine, etc.). Annual fuel consumption and environmen.tal reports are shown in Figure 13-6. A complete sample of the OGP output is included in the OGP Program User's Manual. OGP • s basic structure was designed to maintain a consistent level of detail among three items: the user input, program logic and output format. The level of detail in the program and the computer processing time are intertwined. Adjuncts of these two factors are, of course, data. gathering and the results of the analysis effort. In addition, as the study progresses into the future, the inherent accuracy and confidence represented by the input data diminishes as a result of greater uncertainty in input assumptions. The references included at the end of this section address some of the uses for which OGP vlas written, namely as a long-range generation expansion system planning tool using conventional engineering economics analysis and revenue requirements. Sections 14, 15, and 16 describe the extension of OGP via the Financial Simulation Program (FSP) into th.= realm of financial analysis. 13-2 • ['!''•' .• -,_ f ' I __ o'!J I I . I l i ' I ; • ":! • Jli [ ' ' \..-.: i ~ ' - ,_ GENERAL ELECTRIC COMPANY OGP-6A GENERATION PLANNING PROGRAM V6.10-SUMMARY OUTPUT *~******~********************************************~**** OGP ELECTRIC SYSTEM USERS MANUAL EXAMPLE JOB NUMBER 2526NT 01/13/82 16.886 *****~****************************************************** GENERATION SYSTEM NUCL. F-COAL G.T. STAG C-COAL F-OIL TYPES TYPE 1 2 3 4 5 6 7-10 OPTMZING1991 1990 1985 1987 0 0 lkliCliC PCT TRIM 25 25 0 25 25 25 1984 MW 9805 5449 1752 1000 300 4424 1434 SUM= 24164 *********************************************************************** TOTAL CAPAS. YR y E A R L Y M w A D D I T I 0 N S + TIES ** ******* ******* ******* ******* ******* ******* ***** ****** **** 85 2X 300 25684 86 2X 300 500* 26784 ~7 1200 27809 as 1200 28878 89 1200 30078 90 4X 400 31563 9] ]~]300 2~ 300 l~ 4QO 33748 92 4X 400 35248 93 1X1300 1X 300 1X 400 300 37402 94 1X1300 2X 300 2X 400 39810 i§ ]~J~QQ l~ 300 2QQ !:ll~!:IZ 96 1X1300 1X1300 2X 300 44727 97 1X1300 1X 300 600 46777 98 2X1300 2X 300 49661 9ra ]~]~QQ l~l~QQ 2~ 3QQ ~2!;2~~ 0 2X1300 2X 300 55501 1 1X1300 1X 300 900 57876 2 2X1300 2X 300 600 61361 ~ 2~]~QQ l~l~QQ 2~ ~QQ QQ2~J 4 1X1300 1X1300 1X 300 900 68521 5 2X1300 2X 300 600 72112 6 1X1300 2X1300 1200 76076 z l~l~QQ 2~:J3QQ l~ 3QQ gQQ 803!:19 8 2X1300 1X1300 2X 300 1200 84773 9 3X1300 3X 300 1X 400 300 89374 10 1Xi300 2X1300 2X 300 900 93814 ll 1~1300 2~1300 2~ 300 600 966l~ 12 3X1300 2X1300 4X 300 300 104779 13 3X1300 2X 300 900 109969 14 2X1300 2X1300 2X 300 600 115819 *********************************************************************** MW ADD 26000 42900 13500 5200 0 0 15200 SUM= 102800 MW RET -2845 -3724 ·702 0 0 -4424 -500 SUM= -12195 ****** ****** ****** *1.'**** ****** ****** ****** **** *********** 2014 32960 44625 14550 6200 300 0 16134 SUM= 114769 PCT TOT 28,7 38,9 12.7 5.4 0,3 0. 14.1 SUM=lOO PCT *********************************************************************** AUTO 26000 42900 13500 5200 0 0 14700 $11M= 102300 PCT TOT 25. 4 41 . 9 13 I 2 5. 1 0. 0. 14 '4 SUM= 1 00 PCT Figure 13-1. Annual Capacity Additions by 1ype 13-3 ~--c.· .. ----~-;;'-~~-------~--------------------- • t .. ·, ,, > . '"-~·-. ···-,. '.,. .. 1 j J J I I I I I I I I I' . ,_, I I ~ < ', • ';t 14 ... " GENERAL ELECTRIC COMPANY OGP-6A GENERATION PLANNING PROGRAM V6.10-SUMMARY OUTPUT ********************************************************** OGP ELECTRIC SYSTEM USERS MANUAL EXAMPLE JOB NUMBER 2526NT 01/13/82 16.886 ************************************************************ GENERATION SYSTEM THERMAL HYDRO PSH BATRES COM PAR TY?E 1-6 7 8 9 10 OPTMZING *** 1987 0 0 PCT TRIM 0 0 0 1984 MW 22730 _310 624 500 0 SUM= 24164 ******************~:**************************~************************* TOTAL CAPAB. LOAD LOLP YR Y E A R L Y M W A 0 D I T I 0 N S +TIES MW D/Y ** ****** ****** ****** ****** ****** ****** ***** ******* 85 600 25684 19429 0.4384 86 600 500l!C 26784 20498 0.3606 87 4X 300 27809 21625 0.3904 88 4X 300 28878 22814 0.4489 89 4X 300 30078 24069 0.4720 90 1600 31563 25393 0.4962 91 2300 33748 26790 0.4363 92 1600 35248 28263 0.4829 93 2000 1X 300 37402 29818 0.4334 94 2700 39810 31458 0. 3611 95 1600 2X 300 41847 33188 0.4040 96 3200 44727 35013 0.3728 97 1600 2X 300 46777 36939 0.4323 98 3200 49661 38970 0.4184 99 3200 52535 41114 0.4404 0 3200 55501 43375 0.4533 1 1600 3X 300 57876 45761 0.4739 2 3200 2X 300 61361 48278 0.3987 3 ~6QQ 6522] 50933 Q.40JQ 4 2900 3X 300 68521 53734 0.4600 5 3200 2X 300 72112 56690 0.4085 6 3900 4X 300 76076 59807 0.4590 7 4200 2X 3QQ 80349 63097 0. 4881 8 4500 4X 300 84773 66567 0.4444 9 5200 1X 300 89374 70228 0.4385 10 4500 3X 300 93814 74091 0.4589 1 1 4500 2X 300 98814 78166 0.4862 12 7700 1X 300 104779 82465 0.4414 13 4500 3X 300 109969 87001 0.4836 14 5800 2X 300 115819 91786 0.4926 ****~¥***************************************~***~*~******************* *********************************************************************** MW ADD 87600 0 14700 0 500 SUM= 102800 MW RET -11695 0 0 -500 0 SUM= -12196 l!!l!Cl!Cl!Cl!Clit lll llOIOI()!Sllt ****** ****** ****** ****** ************ 2014 98635 310 15324 0 500 SUM= 114769 PCT TOT 85,9 0,3 13.4 0, 0.4 SUM= 100 PCT *********************************************************************** _AjJTC'l 87600 14700 0 0 SlJM= 102300 POT TOT 85.6 14.4 o. o. SUM= 100 PCT Figure 13-2. Annual Capacity Additions by Type 13-4 Jf. H I : t r '....t .. ·' J r tJ I r·, IS i :1 I r··-i ~ I r. . -J l,,; I f' .,, l~~ I L I ~ L ::''~k\. ·.· '~""""# "''' .. , '" GENERAL ELECTRIC COMPANY OGP-6A GENERATION PLANNING PROGRAM V6.10-SUMMARY OUTPUT ***********~*******~*************~************~*********** OGP ELECTRIC SYSTEM USERS MANUAL EXAMPLE JOB NUMBER 2526NT 01/13/82 16.886 ****~*~***************************************************** TOTAL CAPABILITY (INCLUDING TIES> YEAR TIME OF PCT. YEAR LOAD END PEAK RES. LOSS OF LOAD PROBABILITY D/Y H/Y COST IN MiLLION $ YEARLY CUM. PW COST TOTAL **** ***** ***** ***** **** ****** ****** ******* ******* _1~9=8=5~~1~9~4~2=9--~2~5~7~2~4~~2~5~6~8~4~-~3~2~·~2~~0~·~4~3~8~--~0~.~5~5~~2~1~8~9~·~9~--~1~1~2~3~·~8~---- 1986 20498 26824 26784 30.7 0.361 0.43 2547.6 2312.3 1987 21625 27849 27809 28.6 0.390 0.46 2974.7 3573.9 1988 22814 28918 28878 26.6 0.449 0.52 3512.9 4928.2 1989 24069 30118 30078 25.0 0.472 0.53 4134.1 6377.2 1990 25393 31603 31563 24.3 0.496 0.55 4870.8 7929.2 1991 26790 33788 33748 26.0 0.436 Q.48 5899.5 9638.1 1992 28263 35288 35248 24.7 0.483 0.53 6799.6 11428.6 1993 29818 37442 37402 25.4 0.433 0.47 7997.6 13343.2 1994 31458 39850 39810 26.6 0.361 0.39 9383.1 15385.2 1995 33188 41887 41847 26.1 0,404 0.44 10818.0 17525.5 1996 35013 44767 44727 27.7 0.373 0.41 12578.6 19787.9 _1~9=9~7~--3~6=9=3~9~~4~6~8~1~7--~4~6~7~7~7---=2=6~·~6~-=0~·=4~3=2~--=0~·=4~7--~1=4~1=8=9~·=4---=2=2~1=0~7~·=9 ____ __ 1998 38970 49701 49661 27.4 0.418 0.45 i6226.1 24519.8 1999 41114 52575 52535 27.8 0.440 0.48 18681.9 27044.4 2000 43375 55541 55501 28.Q 0.463 0.49 21186.6 29647.0 2001 45761 57916 57876 26.5 0.474 0.51 23717.7 32295.8 2002 48278 61401 61361 27.1 0.399 0.42 26968.0 350~3.7 2003 50933 65261 65221 28.1 0.401 0.43 31167.7 37910.4 2004 53734 68~61 68521 27.5 0.460 0.49 35230.3 40866.4 ~2-0=0=5~--5=6=6=9~0=---7~2-1~5~2&-___ 7~2-l~-1~.2--~2~7~.~2~~0~.~4=0~~.1~--~0~·~4=3~~3~Sl~6.2 43856.2 2006 59807 76116 76076 27.2 0.459 0.48 44550,5 46945.5 2007 63097 80389 80349 27.3 0,488 0.51 50365.6 50120.5 2008 66567 84813 84773 27.3 0.444 0.46 57340.8 53406,6 _.2~0~0x9---L7~o.2.2~8--~8~94~14=-~a=9~3=7~41-~2~7~.3=-~o~.4~3~9~--~o~·~4~6~~6=~10~9~·~s~---~5~6~7~4~6~·w7~----- 2010 74091 93854 93814 26.6 0.459 0.47 72605.2 60185.4 2011 78166 98854 98814 26.4 0.486 0.50 81304.2 63686.1 2012 82465 104819 104779 27.1 0.441 0.45 94524.7 67386.1 2013 87001 110009 109969 26.4 0.484 0.50 104423.0 71101.9 2014 91786 115859 115819 26.2 0.493 0.50 117168.8 74892.2 Figure 13-3. Summary of Load, Capacity, Reserve, LOLP, and Cost 13-5 ... ~. :]··-.. n. ·.~ : 1 I :'] \ 1 I i ~ I 1 I I I 1 l ! l I l I I \ I I r ., . . -. ' • I ~ '"%j 1-'• OQ s:: "l (I) ...... w I -I= • t:::l (I) ...... (1" w I» I 1-'· 0\ 1--' (I) 0. Cll s:: s s Ill '"S '< 0 1-1;) (") 0 [/J (1" [/J f ~ .~ ~~ ~ {B!Jt ~~ !J.liJlilllt =-;.-t..-~ GENERAL ELECTRIC COMPANY. EUSED GENERATJ ON PLANNJ NG PIROGRAM r,GP-6A V6. 10 OGP ELECTRIC SYSTEM USERS MANUAL EXAMPLE ~l ~ ~l ~ ~ .. ,~ PAGE 80 01/13/62 16.886 2526NT POOL TOTAL YEARLY COSTS CMILLION $) YEARLY COSTS C$/MWH>!...--------,-- PEAK ENERGY LOAD ***•****************~**********************~ **************************************** YEAR <MW) (GWH> FACTOR INVEST, FUEL fiHM NUC INV TOTAL INV. FUEL O+M N.l. TOTAL **** ****** ********* ******* ******* ******* ******* ******* ******** ****** ****** ****** ****X* ******* 1985 19429. 102120.0 §o,oo 46.0 1776.7 281.3 85.8 2189.9 o.5 17.4 2.a D.8 21.4 1986 20498. 107735.8 e;o.oo 111.4 2044.8 :299.9 91.4 2547.6 t.o 19.o 2.8 o:.8 23.6 1987 21625. 113662.0 60.00 209.8 2352.6 315.0 97.4 2974.7 1,8 20.7 2.8 0.9 26.2 tS88 22814. 120241.2 6o.oo 313.5 2761.5 ~34.2 103.7 3512.9 2.6 23.0 2.s o.9 29.2 1989 24069, 126508.4 EiO.OO 423.0 3240,7 :360.1 110.4 4134.1 3.3 25.6 2.8 0.9 32,7 1990 25393. 133465.5 60.00 589.8 3768.3 395.3 117.6 4870.8 4.4 28.2 3.0 0.9 36.5 1991 26790. 140807.1 EiO,OO 1111.8 4202.5 443.3 141.9 5899.5> 7.9 29,B 3.1 1.0 41.9 t992 28263. 148957.2 60.00 1294.6 4868.1 485,8 151.1 6799.6 8.7 32.7 3.3 1.0 45.6 1993 29818. 156721.9 60.00 1901.6 5375.6 5.40.7 179.7 7997.6i 12.1 34.3 3.5 1.1 51.0 1994 31458. 165341.6 60.00 2579.6 5991.3 600.7 211.5 9383.1 15.6 36.2 3.6 1.3 56.7 1995 33188. 174433,9 60.00 3249.9 6666.2 665.4 246.6 10818.0 18.6 38,2 3.8 1.4 62.0 1996 35013. 184533.7 60.00 4294.5 7256.8 741.9 285.4 12578.6 23.3 39.3 4.0 1.5 68.2 _llU!Z 36939, 1-94150.6 6Cl.OO 4815.8 8261.4 808.3 303,9 14189.4 24,8 42.S ... 4.2 1.6 73.....J__ 1998 38970. 204828.8 60.00 5742.2 9275.4 884.8 323.7 16226.1 28.0 45.3 4.3 1.6 79.2 1999 41114. 216094.4 60,00 6996.8 10332.7 980.2 372.2 18681.9 32.4 47.8 4.5 1.7 86.5 2000 43375. 228603.6 60.00 8036.8 11675.6 1077.8 396.4 21186.6 35.2 51.1 4.7 1.7 92.7 2001 45761. 240517.8 f'O.oo 8743.2 13380.7 _ 1111_,1 422.2 23717.7 36.4 55.6 4,9 1 .a 98.6 2002 48278. 253746.9 60,00 10017.1 15217.1 1284.3 449.6 26968.0 39.5 60.0 5.1 1.8 106.3 2003 50933. 267703,7 60.00 12566.8 16611.4 1440.0 549.5 31167.7 46.9 62.1 5.4 2.1 116.4 2004 53734. 283201.0 60.00 14416.7 18590.2 1600.6 622.9 35230.3 50.9 65.6 5.7 2.2 124.4 2005 56690. 297961.6 60.00 15932.2 20~3,9 1756,7 663.4 39196.2 53.5 70,0 5.9 2.2 ___j~ 2006 59807. 314347.0 60.00 18754.1 23092.7 1954,5 749.2 44550.5 59.7 73.5 6.2 2,4 141.7 2007 63097. 331773.3 60.02 21643.7 25701.9 2176.6 843.4 50365.6 65.2 77.5 6.6 2.5 151.8 2008 66567. 351011.2 60.03 25413,0 28512.1 2420.6 995,1 57340.8 72.4 81.2 6.9 2.8 163.4 2009 70228. 369230.7 60.02 28221.7 32161.6 2666,7 1059,8 64109.8 76.4 e;7.1 7.2 2,9 1~ 2010 74091. 389548.7 60.02 31799.4 36728.3 2934.5 1143.1 72605.2 81.6 94.3 7.5 2.9 186.4 2011 78166. 410798.6 59.99 35512.1 41256,3 3259.8 1275.9 81304.2 86,4 100.4 7.9 3.1 197.9 2012 82465. 434716.3 60.01 42803.2 46629.6 3636.5 1455,4 94524.7 98.5 107.3 8.4 3.3 217.4 2013 87001' 457363,0 60.01 46351,0 52520,3 4001,7 1550.0 104423,0 101,3 114,Q 8.7 3,4 228,3 2014 91786. 482501.3 60.01 52693,9 58206.5 4476.3 1792.2 117168.8 109.2 120.6 9.3 3.7 242.8 CUMULATIVE PRESENT WORTH < M I LU "N $) *~-*lie ~lkliC li!.lf(JiC lt~_liCll:lk_*lk.lk_:>l<liC liClk lk **-*ll:_jtX'·lk_~.!cliC~lltJiC lk lk llClLl~JIUk~--__ ~-___________________________ _ YEAR INVEST, FUEL O+M NUC INV TOTAL **** ******* ***M*** ******* ***)!(*** *****~JI(JI< 1985 23.6 911.8 144.4 44.1 1123.8 1986 75 6 1865 7 284 3 86 7 231? 3 ------------------------------------------------------- 1987 t64.6 2863.4 417.9 128.0 3573,9 1988 285.4 3928.1 546.7 168,0 4928.2 1989 433.7 5063.9 672.9 206.7 6377.2 1~gn 6?1_6 6264.6 79a_g ?44.1 792~ 2 1991 943.6 7482.0 927.3 285.2 9638.1 1992 1284.5 8763,9 1055,2 325.0 11428.6 1993 1739.7 10050.8 1184.6 368.1 13343.2 1994 2301 1 11354 7 1315 4 414 1 15385 2------------·---------------------- 1995 2944.1 12671.5 1447.0 462.,9 17525.5 1996 3716.5 13976.7 1580.4 514.2 19787.9 1997 4503.9 15327.5 1712.6 563.9 22107.9 1998 ~3~7.4 1~70~ .. 3 1844 1 ~1?_0 ?451~.8 1999 6302.9 18102.5 1976.6 662.3 27044.4 2000 7290.2 19536.U 2109.0 711.0 29647,0 -! T .I I J J I I I I I .I I I I I I L I GENERAL ELECTRIC COMPANY, EUSED GENERATION PLANNING PROGRAM OGP·6A V6.10 ----OGP ELECTRIC SYSTEM USERS MANUAL EXAMPLE TERRITORY PEAK 25393. M~ SPINNING RESERVE 1200, MW OPTIMUM STAG 1990 YEARLY PRODUCTION COST SUMMARY COSTS IN THOUSANDS OF DOLLARS PAGE 01/13/82 16.886 2526NT THERMAL PEAK 19950. MW UNIT STATION NAME CO. UNI1 FUEL RATING MAINTENANCE MIN. ENERGY HRS. CAPACTY FUEL OPER.+ FUEL FORCED PLANNED FUEL I 0 I DENT. TYPE TYPE MW PTRN. MONTHS UP RULE OUTPUT ON MWH l.INE FACTOR COST MAINT. COSTS INVT. OUTAGE OUTAGE PRICE COSTS RATE RATE S/MBTU ~~~6~~N~E~W~T~O~N~~~0~1~·~E~p~J~S~O~N~~2~--~2---7~0~5~.0~~0~J~A~N~.~------~-~1--~5~0~2~6~0~5~9~-~7~3~5~9~·~0~.8~1~4~_1~4~7~5~9~2~.--~1~1~1~3~5~.--~~~0~.--~0~.0~8~2~0~.1UL__.~ 20 SEASHORE 01 EDISON 1 1 960.0 0 OCT. 1 6779612. 7062 0.806 92841. 17347, 11515. 0.119 0.120 1.369 38 EAST PT 02 PUBSER 1 1 960.0 0 MARCH APRIL 1 6170665. 6428. 0,734 84502, 17347, 11515. 0.119 0,120 1,369 4a SEASMORE 02 EDISON 1 1 960,0 0 MAY 1 6779612. 7062. 0,806 92841, 17347. 11515. 0.119 ~.120 1,369 --~3~9--=EA7.=sT=-P~T=-~0~1~-~p~U~a~S~E~R~~~~---~l--~9~2~5~.~0--0~F~E~a~.~M~A~R~C~H~~~~~5~9~9~1~6~0~2~.~6~4~7~7~.--~0~.7~3=9--~8~2~0~5~0~.~~1~6~~l~6~'·~-1wlu0~9~5~,--~0~.~1~1~8~0~,~1~2~0~-71~.~3~6~9- 49 EAST PT 03 PUBSER 1 1 1200,0 5 1 8387943. 6990. 0.798 118312. 20181. 14394, 0.128 0,120 1,369 51 SEASHORE 05. PUBSEP. 1 1 1200.0 0 OCT. 1 8387943. 6990, 0.798 118312, 20181. 14394. 0,128 0.120 1.3~9 54 SEASHORE 06 EDISOI~ 1 1 1200.0 0 APRIL MAY 1 7634535. 6362, 0.726 107685. 20181. 14:'194, 0,128 0.120 1.369 4;! SEASHORE 0;! EDISON 1 1 1200.0 0 MAY 1 8;!85174. 6990. 0.798 1419;!1. 20181. 14;!94. 0.128 0.120 1 ;!69 4<1 SEASHORE 04 EDISON I 1 1200.0 0 APRIL. 1 7843207. 7011. O, 746 133397. 20181. 14394. 0.128 0.120 1.369 10 STATE 02 EDISON 2 2 210.0 0 SEPT. OCT, 2 1406843, 6924. 0.765 4?.003. ~483, O, 0,051 0.100 3,259 6 LINCOLN 02 EDISON 2 2 170.0 0 JAN, FEB. 2 1134987. 6977. 0.762 34013. 4845. 0, 0,050 0,100 3,239 ---Z__WATERSID~ 01 EDISON 2 2 163 Q 0 JAN 2 118Q918 7615. 0 831 35571 4728 0 0 050 0,100 ;! 259 5 LINCOLN 01 EDISON 2 2 1~0.0 0 NOV, 2 1094828. 7638. 0.833 32812, 4503, 0, 0.050 0.100 3,259 9 STATE 01 EDISON 2 2 125,0 0 APRIL 2 903903, 7638. 0.825 27100. 4048, O, 0.050 0.100 3.259 8 WATERSIDE 02 EDISO~ 2 2 117,0 0 JAN, 2 844149, 7615. 0,824 25308, 3894, 0. 0,050 0.100 3.259 45 BLUE LAKE 03 PUBSER 5 2 300,0 0 APR1L 2 2064658 7509. 0.786 62308 7519 0 0 066 0 103 3 259 50 NEWTON 02 EDISON 2 2 750,0 0 SEPT, 2 51828G2. 7365. 0.789 164762. 11545. 0. 0.084 0.114 3.259 35 FRONTIER 02 PUBSER 2 3 621.0 0 SEPT. OCT. 2 380]003. 6742. 0.699 148814, 10338, 0, 0.076 0.111 4,318 33 FRONTIER 01 PUBSER 2 3 320.0 0 JUNE 2 2174064, 7582, 0,776 85638, 7015. 0, 0,057 0.103 4.318 41 BLUE LAKE 04 PUBSER 2 3 210 0 0 MAY 2 1396957 7607 0 759 55513 5483 0 0.051 0 100 4 318 29 BLUE LAKE 03 PUBSER 2 3 146.0 0 JUNE 2 979670. 7638. 0.766 39092. 4433, 0, 0,050 0,100 4.318 32 RIVERSID~ OS PUBSER 2 3 105,0 0 FEB. MARCH 2 628190, 6977. 0.683 25136, 3655, 0. 0,050 0,100 4,318 26 SLUE LAKE 0~ PUBSER 2 3 145,0 0 SEPT, 2 960638. 7638, 0,751 39265, 4433. 0. 0,050 0.100 4.318 __as~S!PE 04 PUBSER 2 3 100 o 0 NOV 2 653784 7638 o 746 26915 3552 0 0 050 0 100 4 31ft_ !';2 NEWTON 01 PUBSER 2 3 750,0 0 JULY 2 4723155, 7343. 0.719 199764, 11545. 0, 0.084 0.114 4.318 46 FRONTIER 03 PUBSER 2 3 225,0 0 JUNE JULY 2 1249224. 6917. 0.634 53003, 5709, 0, 0.052 0.101 4,318 34 LOON MT 03 PUBSER 6 4 550.0 0 MAY 2 3351117. 7599, Q,596 146876, 5720. 0, 0.05~ 0.089 4 798 40 BAY VIEW 04 EDISON 6 4 550? 0 NOV 2 3275351 7622 0 son 143741 5120 0 a 052 a 089 4 796 31 LOON MT 02 PUBSER 6 4 117,0 0 JULY AUG. 2 651148. 7054. 0.635 .2~988, 2313, 0, G.030 0.080 4.798 28 LOON MT 01 PUaSER 6 4 150,0 0 MAY 2 875058, 7776. 0.666 39143. 2675, 0. 0,030 0.080 4.798 14 STATE 03 EDISON 6 4 527,0 0 APRIL 2 2963520. 7638. 0.642 138104. 5579. 0, 0.050 0.088 4.798 15 STATE 04 EDISON 6 4 527 0 0 MARCH 2 2619845 7615 0 611 13174~.--~~~5~7~9~ ____ _uo ___ o~o~5~0~0~0wBwB1-~4~7u9~R~ 13 HARBOR 03 EDISON 6 4 456.0 0 JULY 2 246571:5, 7647. 0,617 115460, 5126. 0, 0.046 0.086 4,798 12 HARBOR 02 EDISON 6 4 209,0 0 MARCH 2 1056709. 7768. 0.583 50162, 3248. o. 0,031 0,080 4.798 11 BAY VIEW 03 EDISON 6 4 320.0 0 JAN, 2 1211801. 7711. 0,432 58739, 4167, 0, 0,038 0.083 4.798 30 NORTKR I ME02 PIIBSER 6 d 163 0 0 APR II 2 59BBt'i:S 7799 0 d 1 S 29200 2806 0 0 030 0 OBO 4 79& 27 MIDLINE 03 PUBSER 6 4 163.0 0 JULY 2 603956. 7776, 0.423 29496, 2808, 0, 0.030 0,080 4,798 4 BAY VIEW 02 EDISON 6 4 170,0 0 NOV. 2 538068, 7799. 0,361 26783. 2878. O. 0,030 0,080 4,798 24 NORTHRIDGE01 PU8SER 6 4 115,0 0 MAY 2 421085. 7776. 0,418 20690, 2290, O. 0.030 0,080 4,798 2 SO SIDE 92 EDTSON 6 4 146 0 n DEC 2 332426 7671 0 415 2fi274 2633 0 0 030 o nan 4 ?QB 3 SO. SIDE 03 EDISON 6 4 146.0 0 2 570496. 8392, 0.446 28201. 2633, 0, 0,030 0.080 4.798 53 NO, SIDE 03 PUBS:::R 4 15 400 0 0 FEB. 3 1755646. 6111. 0.501 85296. 5401. 0. 0,060 0,080 5.876 57 COMMITTED 3 EDISON 3 15 1:50,0 0 MARCH 3 15037. 591, 0.011 2079. 437, 0. 0.060 0,040 5.376 58 COMMITTED 4 EDISON 3 5 150.0 0 3 12742. 506. 0.010 1773. 402. 0. 0.060 0.040 5.876 59 COMMITTED 5 EDISON 3 5 150,0 0 AUG, 3 8649. 3156. 0,007 1232, 341. 0. 0.060 0.040 5.876 60 COMMITTED 6 EDISON 3 5 150.0 0 3 7023. 336. 0.005 1104, 333, O, 0.060 0,040 5.il76 61 COMMITTED 7 EDISON 3 3 150.0 0 OCT. 3 5083, 246, 0,004 004, 296. 0. 0,060 0,040 5.876 19 HAR80R-GT~0~2~~E~D~I~S~O~N~~3~--~5~~1~5~0~-~0--2~------------~3~--~~~6~4~4~.--~1~9~~~·~0~.0~0~3~--~7~524~.----~2~7~4~----~0~.--~0~.~0~6=0~0~.0~4~07-~5~.8~7~6~ 37 G.T. LUMP 3 PUBSER 3 5 100.0 0 3 1843. 141. 0.002 379, 169. 0, 0.067 0,040 5,876 18 UPTOWN-GT 02 EDISON 3 5 100,0 0 SEPT. 3 Hilt>. 122. 0,002 323. 164, O. 0.067 0.040 5,876 62 G.T. LUMP 4 EDISON 3 15 94.0 0 3 1263, 109, 0.002 274. 151. 0. 0,067 0,040 5.876 --~17~~G~.T~-~L~U~M~P~~1~~E~O~I~S~O~N~~3~--~5--~12~8~.0~~0~------------~3~--~1~3~4~0~.--~9~4~·--0~.0~0~1~--~3~1~8~,-----gQ 70~.------7o~,--Q~,0~6~q~o~.0·4~5.876 36 G.T. LUMP 2 PUBSER 3 5 130,0 0 3 866. 63. 0.001 216. 190. 0. 0.063 0.040 5.876 TIE ENERGY 1368. 2:54, TOTAL THERMAL 24979,0 133198873. 3736173. 380798. 117610, 544, CONV. HYDRO PUMPED HYDRO BATTERIES COMPRESP AIR PURCHASE + SALES SYSTEM TOTALS TYPE 1 NUCL, 2 F•COAL 3 G I 4 STAG 5 C·COAL 6 F-OIL TJENG TOTAL 310.0 4224.0 500.0 500.0 1050,0 31!56:L 0 RATING MW 9805. 5013. 29!52 26DO. 300, 4309, 24979. ENERGY OUTPUT MWH 2208000. -2162494. -52438. -3659;!. 310200. 13346:5549 CAPACITY FACTOR. FUEL COST THO!JSANtl S 0, o. 3164. 28931, 10998. 2006, 951, 3768268 39!5298 117610 0 + M THOUSAND $ THERMAL S/MWH 66360293. o. 7726 971872. 169861, 11.21 33347245. 0,7594 1182301. 106341. 38,64 734!59),. ____ ~0~0~2~8~4------~7~9~~•L9~7~------~1~1~4~5~4~~1~2~3~6~8~----------------------------- 8i4556l. 0.3840 426439, 29348. 152.12 2064b5S. o.78~6 62308. 7619, 33.a7 21945160, 0.~814 1013603, 56175. 48. 7~ 1 368 2!'14 1 8!5 13:1198873. 3736173. 380798. 30.91 = • • •MANUAL MAINTENANC~ PATTERNS• • • • PTRN J F M A M J J A S 0 N D 1 110000000000 ----------------------------------------~~--~OL_~OL_J__LOL-~OL-nO~no~~O--~O--Oc_~0~~.----------------------------------------------- 3 0 0 0 0 0 0 0 0 1 1 0 0 4 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 1 NOTE WHEN USED, PATTERNS OVERRIDE THE COMPUTED P.O,R.-A 1 INDICATES SCHfOII! EO MAINTENANCE Figure 13-5. Annual Production Cost Summary 13-7 ~ ·~ ~:1 ~ 1-' w I co ~ r .~ t-zj 1-'• OQ s:: '"S CD 1-' w I 0\ • ::> ::I ::I s:: P> ......, J::l:.:l ::I < 1-'• '"S 0 ~ i3 CD ::s cT P> ......, ::0 CD '0 0 '"S cT 0" '< '1:1 ......, P> ::I cT IW44i ~ tl.Ria ZJ1II!il!!l ~ ;.--,_ ~~ ~ c~ --- GENERAL ELECTRIC COMPANY, EUSED GENERATION PLANNING PROGRAM OGP-6A V6.10 'OGP ELECTR~C SYSTEM USERS MANUAL EXAMPLE PLANT ID PLANT STATE 2 LINCOLN 3 WATERSIDE 5 SOUTH SIDE 6 BAY V:EW 7 HARBOR 8 NEWTOWN 9 UPTOWN 10 SEASHORE 11 EAST POINT 12 MIDLINE 13 NORTH RIDGE 14 FRONTIER 15 BLUE LAKE T AVG. Y OPER. P EFF. E FUEL CONSUMPT-ION 0.356 2 931413. TON 4 6674349. ~ BBL 0.371 2 694599. TON 0.371 2 739065. TON 0.327 4 485543. BBL 0.360 4 5539392. BBL 0.340 4 3721749. BBL 5 99194. BBL 0.358 2 3710853. TON 3 2364464. TON 0.145 5 54414. BBL 0.311 1 405456. LB. U 0.337 0.336 4 0.337 4 175676. LB. U 481013. BBL 1 0-394 13. BBL 0.370 3 3617519. TON 0.367 2 799255. ION 3 1624592. TON ENVIRONMENTAL REPORT ******** PLANT· SUMMARY ******** 1990 YEARLY HEAT REJECTION Cl"lBTU X 1000) S02 ATI•IOS. WATER C TONS) 9177. 2973. 2457. 456. 4949. 3926. 18119. 287. 340653. 142035. 446. 962. 6778. 6430. 32537. 10542. 8711. 1616. 17546. 12069. 64240. o. 0. o. 1582. 3410. 31124. 22797. 180855.9 125243.9 103469, 1 3670.7 41877.8 26386.4 282086.9 137. 1 o . o. 3636.5 7858.0 130627 .. 4 176879.4 NOX CTCINS) 19247.4 5469.0 4516.7 929.6 11663.3 7632.2 47127. 1 69.1 o. o. 944.1 2049.8 42995.9 22923.9 . :~ .. ~ 7 .. PAGE 23 01/13/62 16.686 2526NT CCI CTONS> 689.8 227.9 188.2 31.0 389.4 252.5 1046.5 1.1 o. o. 31.5 68.4 205.9 288 7 PARTICU- LATES CTONS) 17537.1 15720.8 10973.4 341.4 3838.5 3106.4 395107.3 190.4 o. o. 346.7 752.8 59113.5 28848.0 WATER CONSUMP CGX1000) o. o. o. o. o. o. o. o. 53759797 .. 25234914. o. o. o. o. 16 RIVER~I~ 0.362 3 696688. TON 1623. 5754. 27867.5 7671.8 36.7 5407.9 O. 17 LOON MT. 0.368 4 5947708. BBL 5248. 18605. 44964.7 12799.8 18 NORTH SIDE 0.404 25 UNSITED 0.341 100 UNSITED o. 131 1"0TAL SYSTEM 0.337 5 2 3 5 5 ,. 2197286. BBL 3001186. TON 2581988. TON 4148739. BBL 222416. BBL 3628. 4435. 2768.6 7791.2 32811. 72401. 44309.6 43120.9 1 t 90. o. 280.2 256.2 586147. 307368. 1204919.6 237208.0 427.3 5236.4 o. 120.1 21459.8 o. 1504.3 710054.9 o. 3.9 705.6 o. 5513.5 1278740.9 78994"..!1..!1..!.'-- .., ") "" 0' ••• I ~ fi t, - Jt, j I t"'' }; ~ . I r t: y I f""" 'l Jl v I I I ,,.,, h i ~ I f~' I f ., lt I F r -,, i.; I i r l ' j; I fl -· I ~ - r i 1 1 n I £ J, I { .i ' ~ . . '"' I' l . .... LJ I L I. r L~ I L. I L ' u- EXPANSION OUTPUTS SUPPLEMENTARY INFORMATION l. Parametric Sensitivity Method for Establishing Optimum Long-Range Generation Mix, A.M. Adamson, J.F. Kenney and R.W. Moisan, 1973 American Power Conference. 2. Impact of Uncertainty on Long-Range Generation Planning, L.L. Garver, H.G. Stoll and R.S. Szczepanski, 1976 American Power Conference. 3. Solving Today's Capital and Fuel Supply Problems in the Selection of New Generation, W.D. Marsh, R.W. Moisan and H.G. Stoll, 1975 American Power Conference. 4. Analysis Approach to Evaluate the Impact of Electric Heating Loads on Utility Operations, J.L. Oplinger, 1975 Workshop on Solar Energy Heat Pump Systems for the Heating and Cooling of Buildings. 5. Power Plant Productivity--Techniques for Assessing Benefits and Cost Effectiveness, R.M. Nelson, Jr., M.A. Korn, R. Habermann, Jr., J.B. Tice, R.W. Keller and M.J9 Smith, 1978 American Power Conference. 6. Market Potential for New Coal Technologies, O.D$ Gildersleeve and D. Spencer, EPRI Journal, May 1978. 7. Reducing Oil Consumption Through Economic Generation Reserve Margins, D.L, Dees, G.E. Haringa and H.G. Stoll, 1980 American Power Conference. 8. Utility Generation Planning Within an Interconnected Power System, D.L. Dees, B.W. Erickson, G.E .. Haringa, H.G. Stoll and J.B. Tice, 1981 American Power Conference. 13-9 • -Jr • g ft ~ v, i: I I -r.·-, ·' ", L ,, u ( I I f'' f! ll 1 I r·" " ~ L I f ~ I' ~ " I " f I f ~ "'! I !>" I r. i I F " " £ I ~ ~ ~~ t ··' f ., I ; /i t j, I ' & _, I ~ ' ~ Lr ,. I ' tj I f i ) I i -~} I ~ .~ FINANCIAL DATA The Financial Simulation Program (FSP) is designed to serve as a tool for evaluating the financial impact of alternative generation expansion plans. FSP is a simplified corporate model which focuses mainly on the generation plant. There are two basic categories of input data for FSP: system data and financial data. The follow·ing types of system data must be input into FSP: • Generation additions • Peak loads • Annual energies • Fuel and O&M costs The following types of financial data must also be input into FSP: • Initial balance sheet • Financing ratios and limits • Regulatory and tax rules • Future projections (e.g., interest, inflation, etc.) The OGP program has the capability of storing the system data on a separate file for input into the FSP program. If the user runs FSP independently of OGP, the information transferred from OGP to FSP can be input to FSP through a separate Data Preparation program. The Data Preparation program reads in the data normally transferred from OGP and writes it onto a file in the proper format for FSP. The following system data is transferred . from OGP or the Data Preparation program to FSP via this file. Generation Data The following generation data for each unit is transfer~ed to FSP: • Name • Rating • Years of installation and retirement e Unit Type No • • Plant cost relative to the year costs were quoted in OGP or the Data Preparation program • Per-unit p.lant cost modifier by company for future units 14-1 L ·------.---·-·--·---~------·---·-···-·-.. ··--·"··--· .. ·-·"·--- ·I':J t ;; i\·.: .. ,<''''"''''"'" I I I I I I I I .I I I I I I I .... I I I Annual Data The following data is transferred from OGP or the Data Preparation program to FSP for each year cf the study: G Peak load • Generated energy, including contracts • Generation production costs, including emergency energy and contracts • Fuel cost by type of unit • Demand and energy charges for purchases and sales • Per-unit plant cost inflation modifiers The generation and annual data from a generation expansion study, such as one done with OGP, defines a fixec expansion which will be input into FSP. One of these fixed expansions may become the basis for a number of parametric FSP studies. As is true with much of the OGP input data, the user also has the option of manually modifying the FSP input data which may change on an annual basis .. In addition to the generation and annual data just discussed, the user must provide additional basic data input, including the financial data, directly to FSP. This basic data input ia divided into the following thirteen areas: • Run identification a Initial balan~e sheet • Income statement data • Common and preferred dividend data • General financial information • System data • Generation plant data • Other electric plant data • Tax data • e Regulatory data 14-2 ·r· ·. ····· ·_. .--.. -~------'-·--·--------··------~---~---·--····"-.. ---......... ~--~---"-------... -...... ____________ ~·----------~---~--~---------------. ;.: ~ \ '"' ' .· ' I I 11i""'l ~~ tl t• I «""'"" ; !I I ~ J I I ~ t I' ~ ' I ~ .. / • Second business data • Nuclear fuel data • Optional output specifications A description of each of these data items is included later in this section. FSP Jt"equires sufficient data input . in order to calculate i terns for each year: balance sheet, income statement, cash miscellaneous data such as tax information and financial earnings per-share), etc. In all instances, dollar values thousand8 of dollar·s. the following report. other ratios (e.g., are input in The FSP logic is not designed to operate in an unstructured env.ironment. ThereforE~, the initial data input must be realistic and must reflect the assumpti1:~ns that were incorporated into the accompanying OGP or other system expansion study. For example, the beginning balance sheet must be balanced; Construct;ion Work in Progress ( CWIP) accounts must reasonably reflect the ongoing projects represented in the expansion study; yearly issue size limits imposed on long-term financing must be consistent with the load growth, capacity growth and real dollar inflation specified. In most situations, the input data required may be taken directly from documents such as a utility's annual report or EEI Uniform Statistical Report. However, because of the extensive variation among methods of reporting by different electric utili ties, the user may have some choices to conside:r• when determining where to place certain accounts. This is particularly true with the specification of the initial balance sheet. For example,. the user may maintain separate accounts for nuclear fuel which has been purchased and that which has been leased. Before this data can be input into FSP, these accounts must be added together. The discussion of how the financial simulation actually is done, which is presented in Section 15, may help tbe user make such decisions concerning account placement. Thet remainder of this section of the handbook provides a detailed description of the input data required for FSP, except for the data that previously has been noted as originating from OGP or the Data Preparation program. All of the individual pieces of input data will not be listed separa1~ely. Instead, comments will be made in selected areas to facilitate the user's understanding of FSP's processes. 1. Run Identification a. b .. If the user desires a printout of a unique description, a case identification is required because this type of information will not be carried through from the OGP program. The FSP study must begin with the first year of the OGP or Data Preparation case, and it cannot extend beyond the thirty-year time limit of a single OGP or Data Preparation run. 14-3 I t • I I I I I il I I I I I I I I - I I I I - ··:~ ';'-" ," ,'i· '~,c:.::;:~i·~c:;;.;• i ;,,::; ~/'";,,,~·~·""'"'"''"~~ 2. Initial Balance Sheet a. All monetary values, as of December 31st of the year before the start of the study, must be input in thousands of dollars. It is not necessary to input all of the items on the initial balance sheet. FSP will use the information supplied to calculate the remaining accounts not input. The following items are actually input to FSP: . Assets • Total plant in service, including the Allowance for Funds Used During Construction (AFDC), but not including CWIP • Total generation plant in service, including AFDC, but exclu~ing CWIP • Total second business (e.g~, gas or steam) plant in service, in.cluding AFDC, but excluding CWIP • Generation plant CWIP, including AFDC • Other electric plant CWIP, including AFDC • Second business plant CWIP, including AFDC • Electric system depreciation reserve • Second business depreciation reserve • Net nuclear fuel • Cash balance, end of year • Accounts receivable and deferred debits • Fossil fuel inventory • Total inventory, excluding nuclear fuel Liabilities • Short-term debt • Long-term debt, including current maturities • Common stock, including premiums • Retained earnings • Preferred stock 14-4 ( I ~ (r ii !i j I p _, ••• ,~ I r ·- I :r- A, ;I f ~ " I ' l l I ! t I t I I! ! t- I· ~ 1 . ~" 'ti I ~ t~ I ~ t: ~- I !I ~ " • "'''! I t i ~;;., . :~ c. • Accounts payable and deferred credits • Accumulated deferred federal income tax • Accumulated deferred investment tax credits Total assets must equal total liabilities. By combining all of the items on the company's actual balance sheet with the quantities listed in Item 2. b., the user's ability to attain this goal without undue iteration is enhanced. d. CWIP for the generation plant, including AFDC, may be omitted if the user chooses to supply this information for each individual unite 3. Income Statement Data 4. a. The net operating income for non-electric sources will be held constant by the program. b. During the study, other plant operating expenses will increase proportionately according to the value of the other plant in service· account. Common and Preferred Dividends Data Ratios specified in the input will determine dividend payouts during the study. 5. General Financial Information a. b., Inflation rates for items related to other plant are used to establish the value of these items through time. Many of the values entered in this group of general financial data are ultimately related to the present worth and fixed charge rates which were used for the OGP input datao The input data provided for both programs (OGP and FSP) should be consistent. 6. System Data Historical data will be used to determine the portion plant-in-service accounts that was placed into service in each years prior to the start of the study. This will directly retirement quantities during the FSP analysis. of the of the affect 14-5 I I I I I I ·I _, I .I I I I I I t· - I I I J I .. 7. Generation Plant Data a. If individual values of $/kW and/or inflation factors for the existing generation plant are supplied, the resultant aggregate value should be consistent with that listed on the initial balance sheet. b. FSP will simulate the addition of generating units beyond the end of its designated study period by assuming a continuation of the same generation mix and load growth trend exhibited during the last five years of the study. Although loss-of-load probability (LOLP) is not explicitly considered in the FSP logic, the percent reserve margin is maintained at the same level as in the last year of the study. These additional generating units are necessary to enable FSP to take into consideration a continuing stream of plant construction expenditures. The user, however, has the option of manually defining the additional units. 8. Other Electric Plant Data The other electric plant includes all of the non-generation plant (e.g., transmission, distribution, and general plant) on the system. 9. Tax Data a. Deferred investment tax credits will be carried forward until they are used up. b. The tax rates input into FSP should be consistent with those which were reflected in the fixed charge rates used in the expansion study. 10. Regulatory Data a. If CWIP is included in the rate base, it should be reflected in the plant cost ($/kW) input into the OGP or Data Preparation program. b. A lag in regulation means that the rate change indicated for the current year will not be implemented until a future year. 11.. Second Business Data The second business is completely defined through the input of growth and inflation rates on plant, sales, and expenses~ 12. Nuclear Fuel Data The net amount of nuclear fuel on the balance sheet and the portion treated as a direct expense are controlled through input. 14-6 I ~ /1 ;I ~ ; I ~c<t! ' " I• .. I I 4 ~ .I ! r I I f • I t J I ; I ~- 1 I · I L 13. Optional Output Specifications Use of optional output specifications may provide more "bottom lines," other than conventional financial quantities, for use in evaluating alternate expansion plans. 14-7 "",;>).~ I i ' -' I i{ "' I i ,..,~ -. I c ., ' j r I I ~ I I .I • ~ . I I i: ...... I ,, ~ I y ! ,., I t Cot I I I 1,. t E ~- 0 FINANCIAL SIMULATION A corporate model is a logical structure by which inputs from the planning, operating and financial components of a company are combined to produce financial statements. The Financial Simulation Program (FSP) performs a financial simulation of the capital cost portion of a system expansion plan developed by the Optimized Generation Planning (OGP) Program. Thus, FSP is an extension of OGP; it is not part of the optimization process. FSP is a strategic corporate model, designed to provide the user with information needed to do long-range expansion planning. As a long-range planning tool, FSP needs less detail than is found in corporate models used to study the near-term expansion requirements. As a result, the input data requirements for FSP are simplified; the simulation is done on an annual, rather than monthly or ~eekly, basis; and FSP assumes there is just· one average customer class, rather than segregating the customers into residential, commercial, and industrial classes. As a generation planning tool, FSP focuses on the generation plant. Transmission and distribution are treated as an aggregate. Provisions to model a second business, such as gas or steam, allow the program to calculate consolidated financial results for the company i.n the form of balance sheets, income statements, and cash reports. Thus, FSP allows utility planners to quickly and inexpensively evaluate future expansion plans by providing more "bottom lines" for comparison than does the revenue requirements approach. Some typical functions of FSP include evaluating the financial effects of non-financial decisions, performing general studies of long-term financing, investigating the effects of different cash management and dividend policies, and evaluating the consequences of different tax rates and costs of capital. In this section, the models and techniques used in FSP to perform these functions will be described. The logical structure of FSP is shown in Figure 15-1. The analysis to be performed by FSP has been categorized into ten major areas, each of which is described in detail in this section. The main loop on Figure 15-1 advances the FSP simulation one year at a time. If calculations indicate an unacceptable rate of return, FSP, through an inner loop, will adjust the rates so satisfactory revenues will be obtained, and the desired level of return will be achieved. When the expansion period has been completed, overall financial summaries are provided which enable the user to evaluate the total expansion. 1. Pre-study Initialization Before doing the annual financial simulation, quantities relating to the plant in service at These quantities are necessary to calculate book other plant expenditures during the study period. 15-1 FSP must calculate some the start of the study. and tax depreciation and .. I ··' I I I I I I I .I I I I ,I I ·I I I I I ' ~ . . ~<;;.-# RATE CHANGE lF REQUIRED PRE-STUDY CAPITAL EXPENDITURE·S PLANT RETIREMENT DEPRECIATION REVENUE EXPENSES FINANCING ACCOUNTING TAXES REGULATION STORE YEARLY FINANCIAL SUMMARIES PRINT FINANCIAL SUMMARIES NEXT YEAR Figure 15-1. Logical Structure of the Financial Simulation Program 15-2 ' I P"' I ' ·I ·~ ~ -. '!'~"'<;:'. I ' I f' " I I I{'" I I {t ':;. I . L. I .. ; r ·I· ~''-""' I. ' "l:$1 I ~~ For each generating unit in service at the start of the study, FSP calculates the unit's cost, including and excluding Allowance for Funds Used During Construction (AFDC)o To perform this calculation, each generating unit is first described by an installation date, construction period, total installed cost, and construction expenditure pattern. The construction expenditure pattern is a series of per-unit numbers that shows the rate at which the actual construction expenditures were made during the construction period. The total installed cost (including AFDC) of each unit is then adjusted so the sum of the installed cost for all units on the system equals the generation plant account on the initial balance sheet. The expenditures on the other plant (e.go, tr~ansmission, distribution, and miscellaneous plant) for the years before the start of the study are calculated from the book life for the other plant and the other plant account on the initial balance sheet. FSF assumes that these expenditures increased at a constant rate equal to the average historical load growth times t:1e average historical inflation rate. Similar calculations are made for the second business. The user has the option of inputting historical expenditures year by year for the other plant and/or the second business. Also, as part of the pre-study initialization, FSP, based on the installation dates and costs of the generating units on the system at the start of the study, determines the historical issue dates for the long-term debt shown on the initial balance sheet. The user has the option of inputting this historical bond schedule. 2. Capital Expenditures The area of capital expenditures computes additions to the plant accounts during the FSP study. Plant additions are accomplished in two \vays: (1) by using an explicit ~reject-by-project basis to model generation additions on a unit-by-unit basis; and ( 2) by calculating expenditures on a continuous basis as a function of growth in electric load demand (this method is used for modeling other plant) or as a function of the plant growth rate input for the second business. Each generation unit is described by an installation date, a construction period, a pattern describing per-unit construction expenditures, total installed cost, and, for projects underway at the start of a study, the Construction Work in Progress (CWIP) opening balances. This data is used to calculate the year-by-year expenditures during the study period. AFDC is calculated based on the mid-year CWIP balance for each project. At the close of a project, the accumulated charges are transferred from CWIP balances to the appropriate plant-in-service account. In order to accurately represent the CWIP account for the entire period included in the FSP study, the program defines the units to be added after the last year studied. Alternatively, the user may specify a stream of unit additions for the years following the end of the FSP study. 15-3 • \ ~-;-:-------------------- I I I I I .I I I . t ll- ,, FSP assumes th~t units installed during the study are pl~ced in service on July 1st of the installation year? regardless of the actu~l installation month in the OGP study. The user has the option of specifying the month of installation by unit type or unit by unit. If the unit is inst~lled in mid-year (i.e. , other than January lst) , the user can specify whether construction expenditures are to be made in the months before the unit goes into service or if they ~re to end in the previous year. The month of installation is also considered when calculating book and tax depreci~tion. In a given study year, other plant expenditures are first made to replace, at current costs, any equipment being retired. Then, additional capital expenditures are made so that the current replacement cost of the other plant changes in p_roportion to system dem~nd and the inflation rate for other plant expenditures. Or, at the user's option, other plant expenditures m~y be specified on a year-by-year basis. FSP assumes that other plant expenditures are made on January 1st and that a user-specified portion goes into service on July 1st, e~rning one-half year's AFDC. The remainder is added to the CWIP account and goes into service on January 1st of the following year, earning a full year's AFDC for the year in which the expenditure was made. The user can also define the portion of other plant expenditures eligible for AFDC. The second business capital expenditures are treated in a similar manner except th~t they are grown proportional to the input second business plant growth rate rather than the lo~d growth. The actual expenditures may also be inP.ut year by year. Capital expenditures for nuclear fuel are also calculated by FSP. Three pieces of data are input for each year of the FSP study: (1) the fraction of nuclear fuel to be capitalized, (2) the number of years of nuclear fuel costs to be carried on the company books, and ( 3) the ratio of net to gross nuclear fuel. The fraction of nuclear fuel to be capit~lized tells FSP how much of each year's nuclear fuel will be capitalized as an asset and how much will be treated as a direct expense. The number of years worth of nuclear fuel to be carried on the company books is combined with the yearly nuclear fuel costs from OGP or the Data Preparation program to determine the net nuclear fuel that will appear on the balance sheet. This reflects the lead time involved in purchasing and processing nuclear fuel several years before it is used. The ratio of net to gross nuclear fuel determines the rate at which nuclear fuel is disposed of and removed from the pompany books. The capital expenditures for nuclear fuel are equal to tbe change in gross nuclear fuel from the previous year to the current year plus the cost of the nuclear fuel being removed from the company books in the current year. The cost of nuclear fuel removed from the company books includes the capitalized portion of the burn-up cost for the current year and the change in amortization • 15-4 • 3. Plant Retirement A generation plant is retired and removed from the balance sheet unit by unit, based an the retirement date assigned to each unit by OGP or the Data Preparation program. Other plant and the second business plant are · retired at the end of their respective book lives. I t ~;~i 4. Depreciation Book and tax lives are input separately for each of the ten types of generation, the other plant, and the second business. Straight line depreciation is used for book purposes. The depreciation method used for tax purposes depends on the year in which the asset was installed. Sum-of-the-years' digits (SYD) depreciation, based on the input tax lives, is used for equipment installed after 1954. Shorter tax lives, reflecting the Asset Depreciation Range (ADR) guidelines, are used for assets placed in service during the years 1971 through 1980. Assets installed after 1980 are depreciated according to the Accelerated Cost Recovery System (ACRS) as outlined in the Economic Recover-y Tax Act of 1981 and modified by the Tax Equity and Fiscal Responsibility Act of 1982. If the ACRS tax lives are not input, FSP will calculate them based on the ADR tax lives that were supplied. The tax savings due to liberalized depreciation can be normalized over either the book life (full normalization) or the tax life (partial normalization). 5. Revenue The revenue section of FSP computes the annual revenues obtained from the saleS of electricity and from the second business. The MWh sales are calculated from the MWh generation and the energy loss factor. The MWh sales times the average electric rate yields the total electric revenues. These revenues can be adjusted by the fuel cost adjustment which allows a user-specified portion of the change in average fuel cost since the last rate adjustment to be passed thPaugh to the customers. The revenues from the second business are calculated from the second business sales (an input item) and the price per unit of second business sales. The electric and/or se~ond business revenues may also be input on a yearly basis. 6. Expenses Expenses calculated include fuel and O&M costs and taxes other than federal income tax. The fuel expense and the O&M costs associated with generation are obtained from OGP or the Data Preparation program. The O&M expenses for the other plant are calculated as a per unit of the initial other plant in service. Then they are increased through time in proportion to the load growth and a user-input inflation rate. Production expenses for the second business are calculated from the second business plant at the start of the study, then increased in proportion to inflation and the rate of growth in the second business plant. Sales, customer, and 15-5 •• , •. _..,,. .• ~---.!-'~ *"-'' ~~~ -·~···~'""' -~--~~·~~ --·-"'-·-~~----~-~--.,....---~ .............. -....,__ ··:;rc_·, .. ----~---.. --·--·-·· \''u __ •.tl.~-.:_, -_~'& •.•• . ' :lll!l\1r"· ·r . ,, •••••• co I I I I I I I I ' 7. ·o other general administrative expenses are usually included in the O&M expense category. Property and revenue taxes are also computed. The property tax is calculated from a user-input tax rate and the total plant in service, excluding CWIP and nuclear fuel inventory. The tax rate may also be specified for individual generating units. The revenue tax is computed as a per unit of the total revenues. Financing External financing requirements for the year are determined by estimating the long-term financing requirements, including 'the retirement of existing bonds, and subtracting from this estimate the internal sources of funds such as depreciation~ deferred taxes, and earnings. Debt finanning, preferred stock, and common stock will be issued, in that order, subject to the user-input constraints on minimum and maximum issue sizes, debt ratio, and preferred stock ratio. If the amount required from a certain type of financing (e.g~~ debt, preferred, or common) is less than the minimum issue size for that type, no financing with that type will occur, and FSP will consider the next type. FSP issues short-term financing as needed to maintain the company's cash position above a mlnlmum level specified through input. Additional short-term financing can also be issued to maintain short-term debt as a user-specif~ed per unit of total capitalizat~on. The market price of common stock is determined from the current year's earnings per share and the price-to-earnings ratio input by the user. The user can inhibit the issuance of common stock if the ratio of market price to book price falls below a user-specified limit. When this occurs, any remaining financing will be obtained from long-term debt or short-term debt, as specified by the user. 8. Accounting At this point in the FSP logicy the income statement and cash report ~re computed. This is done iteratively by computing operating income after income taxes and combining this value with other sources of income and the estimated interest and dividends for the existing and new financing to determine the year-end cash position of the company. The dividend rate for preferred stock is a per-unit multiplier of the long-term interest rate. The user has two options in specifying the common stock dividends. One option is to input a payout ratio which is multiplied by the earnings per share to calculate the dividends per share. The second option is to input a growth multiplier which is applied to the previous year's dividends. 15-6 II I I J I , I The minimum allowable cash position is checked. When cash is outside the limit, the short-term financing pattern will be adjusted. If this occurs, the income taxes and year-end cash posi.tions are recalculated using the new short-term financing pattern and the new amounts of interest, dividends, and taxes paid. This procedure is performed until convergence is achieved. The balance sheet is then updated and reviewed to make sure total assets equal total li.abilities.. The company's cash flow is audited by comparing the differenee between sources and applications of funds for the year with the balance sheet change in cash. These comparisons ensure that the accounts have not become unbalanced. 9. Taxes Since FSP does not explicitly calculate state income taxss, they are usually included with the federal income taxes. To determine taxable income, the book income is reduced by the AFDC, the net non-operating income, and the difference between tax and book depreciation. Then the federal income tax is obtained by multiplying the taxable ; n~Jome by the tax rate input by the user. This quantity is then rec:uced by the investment tax credits allowed in that year to determine the federal income tax liabilitye A user-specified portion of the tax savings due to liberaliz,ed depreciation can be normalized, and the remainder will be flowed threugh t.o current income. For equipment governed by ACRS (i.e., Elquipment installed after 1980), all of the tax savings will be normalized. Normalization can be done over the book life (full normalization) or tax life {partial normalization). The tax effects of tlhe borrowed-funds portion of AFDC can also be normalized. The total investment tax credit allowed for an asset may be taken when the asset is placed in service or when the annual construction progress payments occur. Unused investment tax credits will be carried forward until they can be used. The user can specify the portion of investment tax credits to be normalized, and the period over which normalization is to occur. The port.ion not normalized will be flovred through to current incwme. FSP can also simulate the tax effects of stopping construction on a generating unit before it is placed in service. A unit cancellation is mod·eled by assigning a negative tax life to the type of generation associated with the unit to be cancelled. In the year designated by the unit's installation year, FSP will refund to the Internal Revenue Service any accumulated investment tax credits taken on progress payments, and it will depreciate the unit fully for tax purposes.. For tax book purposes, depreciation will be calculated on a straight line basis using the absolute value of the tax life input. 15-7 I • I I 10. Regulation Eate regulation in FSP is simulated by maintaining within specified limits the rate of return on rate base 1 the percentage earned on common equity, or the pre-tax interest coverage. The user also specifies the nregulator•y lag" from the time a rate change is requested until it is actually implemented. The regulatory process begins with the calculation of the rate of return or interest coverage based on revenues determined from the electric and second business rates currently in effect and any applicable fuel rider revenues. If the return produced by these revenues is within the aoceptable range, the program proceeds to the next year of the study. If the return is outside the acceptable range, a rate change will be initiated. Based on the input value for the desil"ed rate of return or coverage, the program will estimate the new rates needed for an acceptable return. As shown in Figure 15-1, the program will feed these new rates back to the revenue calculation and repeat part of the FSP simulation. It will continue iterating in this manner until the return falls within the range of acceptable values. If the regulation is being done currently, the new rates will be implemented immediately.. However, if the regulation is being lagged by one or two years, the new rates will ·become effective one or two years later. Whenever a rate change becomes effective, a new fuel adjustment basis is calculated for use in future years • The rate regulation in FSP will automatically adjust both the electric rate and the second business rate. The second business rates can be held constant between runs by inputting the second business revenues or rates. The electric rate would then fully reflect any changes due to sensitivities being studied. It is also possible to input the electric revenues or rates for some or all of the study years. The dollar returns and bases for the computation of rate of return on rate base, percentage e~rned on common equity, and interest coverage are as follows: • Rate base = Gross plant in service + CWIP (optional) -Book depreciation reserve + Net nuclear fuel + Materials and supplies -Deferred federal income tax (optional) -Deferred investment tax credits (optional) ( CWIP may be included in the rate base to the extent desired by the user, and AFDC will automatically be adjusted acco~dingly.) • Return on rate base = • Common equity base Net operating income/Rate base Common stock outstanding + Retained ea.rnings· (0.5)(Common stock issued during current year) 15-8 . ....,...,,"""'lr-·--· -·~·------------·--·------. <t~ i L ·" • ..., I \.i I I I I 1 I I .. I '"'"' • Return on common equity = (Net income -Preferred base dividends)/Common equity • Pre-tax interest coverage = (Income before interest funds + federal taxes, adjustments)/(long-term interest) + AFDC from borrowed. including deferred and interest + short-term Regulation based on a combination of return on rate base and return on common equity is also available. The common equity associated with equipment in the rate base will earn the rate of return specified by the user. The common equity associated with CWIP will earn a return consistent with the AFDC rate.. Based on the mix of capitalization in the company and the range of values input for return on common equity in the rate base, FSP will calculate the values for return on rate base to be used in the regulation. This is done as follows: Return on Rate Base where L.T. Int. + S.T. Int~ + Preferred Div. + Common Div. = --~--------------~~----~----------~------------~--~--L.T. Debt + S$T. Debt + Preferred Stock + Common Equity Base L.T. Debt includes current maturities, S.T. Debt is the average outstanding for the year, and Common Dive. equals the common equity base times the input value for return on common equity in the rate base. At this point in the FSi1 logic, all quantities necessary for the yearly output. financial statements have been calculated. Three basic reports are developed: annual balance sheet, income statement, and cash report. In addition, tax statements and numerous other financial ratios and indicators are calculated and displayed for reference. Present worth treatments of key quantities such as earnings per share, revenues, capital expenditures, etc., are also available to the user. 15-9 • ~-~---~--···-· --·------------------~-~-----·----------------·"''"''-··· I I, .. I ." Ji, ,, I' I' ~, I ", ... I I. i l ·::; FINANCIAL RESULTS The output from FSP begins with a display of the input data, which includes the beginning consolidated balance sheet. Then, for each year of the FSP study 7 the user can obtain remote summaries or batch output consisting of annual balance sheets, income statements, cash reports and miscellaneous tax data and financial information such as earnings per share, etc. A yearly breakdown of the construction expenditures for each new generating unit is also provided. Sample output pages are shown in Figures 16-l, 16-2, 16-3, and 16-4. More detailed information regarding the output is included in the FSP User's Manual. 16-1 t! }~ l'1 'l \lilW '.) 01/29/82 17.250 JOB NUMBER 1862YT OGP ELECTRIC SYSTEM FILE UMFSP6L USERS MANUAL EXAMPLE EDISON AND PUBLIC SERVICE CONSOLIDATED B A L A N C E S H E E T AS OF DECEMBER 31 <THOUSANDS OF DOLLARS) 1990 1991 1992 1993 1994 ASSETS UTILITY PLANT ELECTRIC PRODUCTION OTHER PLANT GAS PLANT CONSTR. WORK IN PROGRESS TOTAL NUCLEAR FUEL TOTAL UTILITY PLANT LESS DEPRECIATION RESERVE LESS AMORT OF NUCLEAR FUEL NET UTILITY PLANT CURRENT + ACCRUED ASSETS CASH + EQUIVALENT ACCTS. REC. + DEF. DEBITS MATERIALS ANQ SUPPLIES TOTAL CURRENT ASSETS Te!TAL ASSETS LIABILITIES CAPITALIZATION COI'•li"lt;lN STFICJ<. t:lUTSTAND T NG RETAINED EARNINGS COMMON STOCK EQUITY PRE:.FERRED STOCK I. IT I DEBT+CLIRR I t1ATUR IT I ES TOTAL CAPITALlZATJON 13371559 918.4020 0 13090024 1o0918o 37254783 7531141 402295 29321347 9859 821984 885951 1717795 31039142 7029800 3117951 10147751 3205931 13356947 26710629 CURRENT + A (";CRllED L I AB I I I I I ES SHORT TERM DEBT 245525 ACCOUNTS PAYABLE + MISC. 1754564 TOTAl CIJR. +ACGRIIED I 1 AB 2000Q8q DEF. TAXES+OTHER CREDITS 2328423 16424413 1 0124191 0 13601454 1682848 41832905 8406564 420712 33005629 10034 933482 963993 1907510 3.4913139 7961800 3468065 11429865 3610931 15044536 300135332 222576 1992547 2215124 2612683 17500606 11166822 0 1660.4326 2023485 47295240 9370376 505871 37418992 10662 1023137 1137342 2171140 39590132 9173800 3829622 13003422 4107931 17118171 34229524 386127 2177970 2564097 2796512 21073872 12323095 0 17564014 2374839 53335819 10444046 593710 42298063 10884 1184605 1238901 2434390 44732453 10457800 4279307 14737108 4654931 19387327 38789366 260687 2528575 2789262 3153825 25018993 13605413 0 19205592 2750062 60580059 11696965 687515 48195579 8969 1347858 1365561 2722388 50~17967 11970800 4790334 16761134 5294931 22062327 44118392 333673 2877046 3210718 3588856 Tl"lTAI I I AB + CAP I TAl 31039142 ~4913139 3~5901~3 44732453 50917966 NEW FINANCING CHAI'JGE IN BONDS BON[J RETIREMENTS TOTAL BOND FINANCING PREFERRED STOCK COMNe!N STOCK "fOTAL 1811021 ?9~79 1841000 435000 1072000 3348000 1687588 18412 1706000 405000 932000 3043000 2073635 4f\365 2119000 497000 1212000 3828000 Figure 16-1. Annual Balance Sheets 16-2 2279156 71844 2351000 547000 1234000 4182000 2665000 a 2665000 640000 1513000 4818000 \, ' ..... ,. ~"-""'7 ,.,..... .... ----------·-"-"-·-""•'"''''"'"'"······· ···--·-----.. ·• ............... ____ , ___ ,_,_._~--,_,,,,,, ____ , __ ,_ > ·-···-·> ~-----·-·-.... ~··-·· -,----·-·---··--·-··-·-·•-"""""•'"-"""'"'-'''""""•"-"" -··-~;-----·-~-~-~~ ., ~:- I I I I I I I I I. I, I 01/29/A2 17 250 JOB ;-JUI'1BER EDISGN AND PUBLIC SERVICE CONSOLIDATED I N C 0 M E S t A T E M E ELECTRIC REVENUE GAS REVENUE ELEC. FUEL ADJ. REVENUE TOTAL ELECTRIC OPER. EXPENSE FUEL AND O+M NET PURCHASED POWER OTHER PROD. EXPENSES TOTAL ELECTRIC PROD. N T AS 1990 9768835 0 0 9768835 4252245 61155 575233 4888633 OF DECE:i"!BER 31 J <THOUSANDS OF 1991 1992 1993 11093922 11736056 14078380 0 0 0 0 423364 0 11093922 12159420 14078380 4756478 5471139 6059626 65575 70448 75408 637215 705875 781933 5459268 6247462 6916966 0 0 0 GAS PROD.~E~X~P~E~--~~lS=E~S~------------~--------~--------~--0 DEPRECIATION EXPENSE F.I.T. LIABILITY OTHER TAXES DEFERRED + APJUSTMENTS TOTAL OPER. EXPENSES OPERATING INCOME AFPC-EQUIJY FUNDS NET OTHER NON-OPER. INCOME INCOME BEFORE INTEREST LONG TERM INTEREST SHORT TERM+OTHER INTEREST AFDC-BORROWED FUNOS(CREDl NET INCOME PREFERRED DIVIDENDS AVAILABLE TO COMMON COMMON DIVIDENDS NET INCOME AFTER DIV COM110N SHARES YE.AR AVG. EARNINGS PER SHARE DIVIDENDf. PER SHARE PAYOUT RATIO STOCK BOOK VALUE -$/SHARE STOCK MKT PRIC~ -$/SHARE MARKET/BOOK RATIO PRICE/EARNINGS RATIO RET! IRN f'lN RATE BASE RETURN ON COMMON EQUITY RETURN ON CAPITALIZATION 805824 731338 882559 145092 7454445 2314390 378198 16601 2709189 1103280 24702 242248 H323454 289507 1533947 1227158 306789 117731740 13.0292 10 4233 0.8000 86.1938 1 1 7 2626 1. 3605 9.0000 a 1352 o. 1596 0.1105 907055 741103 1000613 284269 8392299 2701623 394883 16601 3113107 1262016 23171 251849 2079769 329197 1750572 1400457 35(11 14 126030359 13.8901 1 J 1121 0,8000 90.6914 125 QJQ:Z 1. 3784 9.0000 0 1326 0. 1597 O.i118 100i902 1113756 758566 936654 1124810 1264055 183829 357313 9316569 10588744 2842851 3489636 476142 502123 16601 16601 3335594 4008360 1432043 1629759 30131 32017 306182 322990 2179602 2669574 371817 421146 1807785 2248.1128 1446228 1798743 361557 44<:::lf>86 134778024 14437854.11 13.4131 15,5731 lQ 730~ 12 ~585 0.8000 0.8000 96.4803 102.0727 120 :Zl:Z5 l~Q 158~ 1 . 251 2 1 . 3731 9,0000 9.0000 0 1295 0 1344 0. 1458 0. 1595 o. 1064 0.1117 i862YT DOLLARS) 1994 16018560 0 0 16018560 6764138 80869 866186 7711193 0 1323470 1073246. 1465155 435031 12008095 4010465 542187 16601 4569253 1855060 29421 347594 3032366 477232 2555135 2044108 511027 154026030 16.5890 ]~ 2:Zl2 0.8000 108.8201 ]~q :1008 1. 3720 9,0000 0 1321 0. 1596 0.1114 3 3944 3 4161 3 1352 3 3851 3 4095 PRET~X INTEREST CO\/ERA~~~E-------.~~~-----~~~---~~~~--~~~~--~~~~----­ AFTER TAX INTEREST COVRAGE DEBT RATIO PREEERRED RA rr n AFDC AS PRCNT OF EARNINGS GROSS P~ANT/REVENUE OPERA1ING RATIO ELECTRIC RATE CENTS/KWH GAS RATE CENTS/CF 2.6166 0.5001 0 1200 40.4477 3. 6489 0.6733 8.0433 o. 2.6183 2.4907 0.5001 0,5001 0 1200 a 1..2Q.O 36,9441 43.2753 3,6191 3.7232 0,6641 0.6887 8.6580 8.9704 0. o. Figure 16-2. Annual Income Statements 16-3 2.6065 2.6091 0.5001 0,5001 a j?oa a l2aa 36.6973 34.8232 3.6193 3.6102 0,6602 0,6555 9,8715 10.6463 0, o. -- 01 /29/82 1 7. 250 JOB NUr1BER 1862YT EDISON AND PUBLIC SERVICE CBNSOLIDATED C A S H R E P 0 R T AS OF DEC~MBER 31 (THOUSANDS OF $) 1990 1991 1992 1993 1994 BEGINNING CASH BALANCE 11374 9859 10034 10662 10884 FUNDS FROi"l OPERP.T IONS NET INCOME AFTER DIVIDEND 306789 350114 361557 449686 511027 NON CASH EXPENSES DEPRECIATION 805824 907055 1001902 1113756 1323470 AMORT. OF NUCLEAR FUEL 1089482 1337839 1399084 1682283 1974391 P ROV l S I eiNS F:;;.!Oo!.!R:lo-T.wA:liX~E~S2.-_____ 1w7r...:.5:L:92..:9;:!.l;82->90!...-___..!:2=.:,0~2::..:.5.?.i:9~7~6~--!E2=.:.0~6u.7:..!2~0:!.:.5f----..62~5:!.:.5~8~0!..::!:2..s2,______s2..::9!..£7..!:3~4t..:::3~1--- ·TOTAL FUNDS FROM OPER. 3962084 4620984 4829748 5803747 6782320 FUNDS FRQM OUTSIDE SOURCES EQUITY SFCURIIIES LONG TERM DEBT SHORT TERM INSTRUMENTS TOTAL OUTSIDE FUNDS PROVISION FOR INT. + DIV. INTEREST DIVIDENDS TOTAL SOURCES OF FUNDS APPLICATION OF FUNDS CAPITAl EXP. -GEN. CAPITAL EXP. -OTHER CAPITAL EXP. -GAS CAPITAL EXP. -NUC. FUEL INTEREST PAID TAXES PAID DIVIDENDS PAID CHANGE IN WORKING CAP. TOTAL APPLIED FUNDS ENDING CASH BALANCE FEDERAL INCOME TAX BOOK INCOME W/0 TN~ TAX LESS OTHER TAX ADJUST. TAXABLE INCOME I='I='DI='RAI I NCf'!MF TAX LESS INVESTMENT TAX CR. F. I .T. LIABILITY F I I DFFFRRAI , DEPRFC F.I.T, DEFERRALJ AFDC TAX DEPRECIATION M I SCEI I A NFf'!IIS RATE BASE COMMON EQUriY BASE 1507000 1 811 021 -7988 3~10032 1127982 1516665 9916763 344~282 885824 0 1313528 1 127882 1613897 1516665 11099 9918277 9859 2700884 839687 1861198 856151 124813 731338 49582 0 1008464 1337000 1687588 -22948 3001639 1285187 1729654 10637464 3553t:\96 982391 0 1393089 1285187 1741716 172965-'l -48443 10637289 10034 3105131 947534 2157598 992495 251392 741103 66748 0 1191255 1709000 2073635 163551 3946186 1462174 1818044 12056153 4070?90 1089497 0 1654563 1462174 1883376 1818044 77581 12055526 10661 3121996 1155868 1966128 904419 145853 758566 74656 0 1358846 1831000 2279156 -125439 3984716 1661776 2219888 13670127 45?1014 1208297 0 1945798 1661776 2200709 2219888 -87577 13669905 10884 3963541 1288011 267553G 123074.4 294090 936654 106338 0 1560053 2665000 72985 4890985 1884481 2521339 16079125 ssqg,.agg 13L!0069 0 2255808 1884481 2538401 2521339 -58557 16081039 8969 4540642 1489185 3051458 1403671 330425 1073246 154904 0 1906274 17117275 20368169 219520Q8 25972951 30355548 9611751 10963865 12397422 1409~~ 16004635 Figure 16-3. Annual Cash Reports 16-4 • I I I I I I I I I I I. ~ at li; 'iii ' ·;l!' GENERAL ELECTRIC COMPANY, EUSED FINANCIAL SIMULATION PROGRAM FSP-6 V6. 10 01/29/82 17.250 J08 NUMBER 1862YT BASED ON 0GP6 JOB 2526NT DATED 01/13/82 OGP ELECTRIC SYSTEM, FILE UMFSP6L USERS r·1ANUAL EXAMPLE PLANT ACCOUI'JTS ( THOUSAI'ID DOLLARS) YEAR 1995 ********* UN!T INSTL YR ******* 1995 liO{OI: !l: :1: )j( )j( ***** CUMULATIVE ***** ID UNIT NAME TOT.COST *EXPEN. AFDC TOTAL* )j( CWIP AFDC TOTAL* 43 NUCLEAR 3602552 112273 173293 285566 2806819 795733 3602552 44 FOSSIL-COAL 2442130 171161 124006 295168 2053936 ·388194 2442130 45 GAS TURBINE 113455 53365 5043 58408 106731 6724 113455 46 GAS TURBINE 113455 53365 5043 58408 106731 6724 113455 47 FOSSIL-COAL 2588658 362862 114302 477164 1995742 280039 2275780 48 PUMPED.HYDRO 275254 9512 11 9f.\5 21497 194993 63071 258064 49 PUMPED HYDRO 275254 9512 11985 21497 194993 63071 256064 50 GAS TURBINE 1191 28 56034 17€.5 57799 56034 1765 57799 51 FOSSIL-CelAL 2743978 576951 90870 667820 1730852 175682 1906534 52 FOSSJL-CelAL 2743978 576951 90870 667820 1730852 175682 1906534 55 NUCLEAR 4351701 542479 145249 687728 2576777 380210 .2956987 56 FOSSIL-COAL 2908616 611568 57793 669361 1223136 89900 1313036 59 FOSSJL-CelAI,. 3083133 432175 27227 459402 648262 34034 682296 60 FOSSIL-COAL 3083133 432175 27227 459402 648262 3··1034 682296 63 PUMPED HYDRO 340991 35351 8908 44259 159077 25797 184875 64 PUMPED HYDRO 340991 35351 8908 44259 159077 25797 184875 65 PUI"1PED HYDRO 340991 35351 8908 44259 159077 25797 184875 66 FOSSIL-COAL 3268121 229053 7215 236268 223053 7215 236268 70 PUMPED HYDRC'l 359746 37295 7049 44344 130532 17818 148350 71 PUMPED HYDRO 359746 37295 7049 44344 130532 17818 148350 74 NUCLEAR 5598317 348941 21983 370924 523-111 27479 550890 75 NUCLEAR 5598317 348941 21983 370924 523411 27479 550890 79 NUCLEAR 5962208 185811 5853 191664 1 6581 1 5853 191664 80 PUMPED HYDRO 400406 27673 3487 31160 69183 6538 75721 8] PU(1Pt;Q HYQRQ 4004J6 27673 3487 31160 69183 6538 757~1 82 PUMPED HYDRO 400406 27673 3487 31160 69183 6538 75721 8.5 PUMPED HYDRO 422429 21897 2069 23966 43793 3219 47012 86 PUMPED HYDRO 422429 21897 2069 23966 43793 3~19 47012 ~2 E!.!~'lEEQ l:!'iQ!3el 4~;.!22~ ] !24QJ ~7Q ] 2~7] 2~1Ql 121~ ~~~]~ 93 PUI"lPED HYDRO 445662 15401 970 16371 23101 1213 24314 94 PlWlPED HYDRO 445662 15401 970 16371 23101 1213 24:3:14 95 PUMPED HYDRO 445662 15401 970 16371 23101 1213 24314 ~~ EUr1EEQ I::IYQB~ ~ZQlZ~ f212~ 2~2 f2;lf2Q e12~ 2~2 a3ao 100 PUMPED HYDRO 470174 8124 256 8~80 8124 256 8380 TOTAL 5498432 1003506 6501938 18677890 2707329 21385217 ~II::IEB Et ~m· l~Z1Z~~ -l~~ez L4e22;JQ 36Z932 QZ~~ 3Z3Z~l GAS 0 0 0 0 0 0 GRAND TOTAL 6970174 1017994 7988168 19045825 2713124 21758948 YEAB l 996 ****l!:**** UNIT INSTL YR ******* 1996 ******* ***~* CUMULATIVE 1>"1C't<:J."]( ID UNIT NAt'lE TOT.COST *EXF'EN. AFDC TOTAL* * CWJP 1\FDC TOTt\L-..: 47 FOSSIL-COAL 2588658 181431 131447 312878 2177173 411.:18G ;~5S8t'58 48 PUMPED HYDRO 275254 4756 12434 17190 199749 75~(1~ 275254 49 PUMPED HYDRO 275254 4756 12434 17190 1 !.hl7·19 75~05 ~75254 50 G~~s Il!al?ci ~IE 119128 560~~ 5295 6l32q I 12..0G7 7060 --L1..9.12A._ 51 FOSSIL-COAL 2743978 384634 121160 5057S"3 21 15·1SG ;'lSH.1841 2-11 2. ... ~ .. ?7 52 FOSSIL-COAL 27-43978 384634 121160 505793 2115·1N'i ?~1(i841 2·11 :;~~~:7 53 GAS TURBINE 125084 58835 1853 60689 58$~15 1853 Gt.'f'~"9 Figure 16-4. Annual Plant Expenditures 16-5 I . ·' ..... ·M.·,'f!C. [r ~ ·:. ~-,:-------,....--·---·~·~-~~·-~----~·-···--.. -···· .. ·····-----·~·-·-·-,···---·-·--···--·--· ........... ·-··---···-----·---·····-·------····-·------···--·-·~· ···---· ,. ~-~ ···,~-··-------~-------~ ...... _.~,....,,.~, ,.··-~~--.-.·-------~--,.....-- '~ 'jlli\ . ·",tJ' .·.·-·-······"··•" ' . ~> ' ' FINANCIAl RESUlTS SUPPlEMENTARY INFORMATION 1. Adding Financial Simulation to Long-Range Generation Planning, R.P. Felak, W.D. Marsh, R.W. Moisan and RoM. Sigleyo, 2. The Effect of Load Factor on Generation Mix and Financial Planning, R.P. Felak, B.M. Kaupang, W.D. Marsh and R.W. Moisan, 1976 Frontiers of Power Technology Conference. 3. Planning to Improve Utility Profitability, V.A. Rydbeck and R.M. Sigley, 1975 American Power Conference. 4. Economic Implications of Growth, G.N. Creighton and R.M. Sigley, 1976 Joint Power Generation Conference. 5. The Effect of Load Growth Uncertainty on Generation System Expansion Planning and Financial Analysis, D.L. Dees, R .. P. Felak and G .. A. Jordan, 1978 American Power Conference. 6.. The Necessity of Including Financial Simulation in Long-Range Generation P~anning, R.P. Felak, C.Do Galloway, G.E. Haringa, R.M. Sigley and H.G. Stoll, 1978 American Power Conference. 7. Integrating Financial Analysis with Generation Planning, R. P. Felal<: and R.M. Sigley, 1978 Pennsylvania Electric Association System Planning Conference. 16-6 I I I I I I I I I I 7 ADDITION OPTIMIZATION ~ 5 14 2 PREDETERMINED GENERATION STUDY r--DATA 12 13 •• DATA PREPARATION .. J -- RELIABILITY EVALUATION 111 PRODUCTION COSTS , ENVIRONMENTAL IMPACTS II INVESTMENT COSTS ~ n OPTIMIZATION RESULTS EX~ANSION OUTPUTS ..__ ___ -r--__ ,._ FINANCIAL .. F"INANCIAL DATA -· SIMULATION ..___ _____ __, 16 FINANCIAL I ._, ___ RESULTS 1 6 8 r . ..1 9 10 II 15 LOAD MODEL -LOAD MODIFICATIONS 4 * Numbers indicate handbook section in which each topic is discussed. Figure 1-1. OGP/FSP Schematic Flow Ghart I I I I I I I I I I I I I I I I I I £~ .! ,., ~ '-\ LOAD 3 MODEL -LOAD 4 MODIFICATIONS - 1 DATA 6 )ARATION --1 lABILITY 8 _UATION ~ r DUCTION 9 ~OSTS ~ON MENTAL 10 \1PACTS • ESTMENT II :>STS - ·~ !MIZATION ~ESULTS ~ PANS ION UTPUTS I ~ANCIAL 15 ULATION 1 'JANCIAL ESULTS .ch each topic is discussed. hematic Flow Chart 2 PREDETERMINED GENERATION 5 STUDY ~ DATA ADDITION 7 OPTIMIZATION 0 - 12 13 14 FINANCIAL -DATA .. 16 • DATA PREPARATION -I -. RELIABILITY EVALUATION ' PRODUCTION COSTS •• ENVIRONMENTAL IMPACTS • INVESTMENT COSTS ' OPTIMIZATION RESULTS EXPANSION OUTPUTS r . FINANCIAL SIMULATION , FINANCIAL RESULTS 6 s 9 10 II 15 LOAD MODEL - 3 LOAD MODIFICATIONS 4 ' '• * Numbers indicate handbook section in which each topic is discussed. Figure 1-1. OGP/FSP Schematic Flow Chart I. ---~---·---·-·--; ·--:--·--;;-··--=t .. i :.J