HomeMy WebLinkAboutFuel_Price_Assumptions_Jan2009
Alaska Renewable Energy Fund
Summary of Fuel Price Assumptions Used for Project Economic Analysis
January 23, 2009
!"#$%&’(#)%"*
!"#$%&’()&*%+’$,&#-’$%").%/01$21%34’&56%/7*")&#*6%8/3/9%14+%*"’%:4#;’&$#*6%)<%/01$21%/4,")&15’%
=4$*#*7*’%)<%>),#10%14+%3,)4)?#,%@’$’1&,"%8=>3@9%(&)+7,’+%*"’%<7’0%(&#,’%(&)A’,*#)4$%*"1*%.’&’%7$’+%#4%
’,)4)?#,%14106$’$%)<%(&)A’,*$%&’;#’.’+%74+’&%*"’%@’4’.1-0’%34’&56%B74+%&’C7’$*%<)&%1((0#,1*#)4$D%%
E#<’%,6,0’%’,)4)?#,%14106$’$%.’&’%(&’(1&’+%-6%,)4*&1,*)&$%<)&%?)$*%(&)A’,*$%14+%&’;#’.’+%-6%/3/%14+%
=>3@%$*1<<D%%E#42$%*)%*"’%’,)4)?#,%14106$’$%1&’%()$*’+%)4%*"’%/3/F$%@’4’.1-0’%34’&56%B74+%.’-(15’%
"**(GHH...D12’4’&5617*")&#*6D)&5H@3IB74+D"*?0D%%%!"’%14106$’$%,)?(1&’%<7’0J%)(’&1*#)4J%
?1#4*’414,’J%14+%’C7#(?’4*%,)$*$%)<%’K#$*#45%+#’$’0%)&%41*7&10%51$L<7’0’+%$6$*’?$%.#*"%*")$’%)<%
(&)()$’+%&’4’.1-0’%’4’&56%$6$*’?$%);’&%*"’#&%7$’<70%0#<’D%%M#;’4%*"1*%&’4’.1-0’%’4’&56%(&)A’,*$%7$71006%
"1;’%0#**0’%*)%4)%<7’0%,)$*$J%*"’#&%?)$*%$#54#<#,14*%’,)4)?#,%#?(1,*%#$%)<*’4%*"’%;107’%)<%<7’0%+#$(01,’+%
);’&%*"’#&%0#<’D%%>#4,’%*"’%1$$7?’+%7$’<70%0#<’%)<%&’4’.1-0’%’4’&56%$6$*’?$%,14%&145’%<&)?%NO%6’1&$%<)&%
-#)?1$$%14+%.#4+%(&)A’,*$J%*)%PO%6’1&$%<)&%5’)*"’&?10J%14+%*)%QO%6’1&$%<)&%"6+&)’0’,*&#,%$6$*’?$J%<7’0%
(&#,’%1$$7?(*#)4$%"1;’%1%$7-$*14*#10%#?(1,*%)4%(&)A’,*%’,)4)?#,$D%%
3K(’&#’4,’%"1$%$").4%*"1*%’4’&56%(&#,’$%1&’%;’&6%+#<<#,70*%*)%(&’+#,*D%%=*%#$%0#2’06%*"1*%*"’%(&)A’,*#)4$%
+’$,&#-’+%#4%*"#$%&’()&*%.#00%);’&’$*#?1*’%)&%74+’&’$*#?1*’%*"’%1,*710%(&#,’%)<%+#’$’0%14+%)*"’&%<7’0$D%%
!"’%?1#4%)-A’,*#;’%)<%*"#$%’<<)&*J%").’;’&J%#$%*)%,)4*&#-7*’%*)%1%,)??)4%<&1?’.)&2%<)&%,)?(1-%*"’%
’,)4)?#,%;#1-#0#*6%)<%(&)A’,*$D%%R"#0’%.’%?16%4)*%-’%1-0’%*)%$16%.#*"%,’&*1#4*6%*"1*%14%#4+#;#+710%(&)A’,*%
?12’$%’,)4)?#,%$’4$’J%.’%,14%?)&’%$1<’06%,)4,07+’%*"1*%)4’%(&)A’,*%"1$%*"’%$1?’%)&%-’**’&%’,)4)?#,%
(’&<)&?14,’%*"14%14)*"’&D%
+),-,.*/"&*0#1,$*2)3’)&*4’,.*5$)(,-*
+),-,.*6--’78#)%"-*
=4%’1&06%NOOS%/3/%&’C7’$*’+%*"1*%=>3@%(&’(1&’%+#’$’0%14+%41*7&10%51$%(&#,’%(&)A’,*#)4$%<)&%/3/F$%7$’%#4%
1$$’$$#45%(&)()$10$%$7-?#**’+%74+’&%*"’%T’410#%U)??#$$#)4H/3/%/0*’&41*#;’%34’&56%&’C7’$*%<)&%
(&)()$10$D%%%=4%/(�%NOOSJ%74+’&%,)4*&1,*%*)%/3/J%=>3@%,)?(0’*’+%1%.)&2#45%(1(’&%V/01$21%B7’0%W&#,’%
W&)A’,*#)4$%NOOSLNOPOX%8/((’4+#K%/9D%%!"’%(1(’&%+’$,&#-’$%1%?’*")+)0)56%*"1*%Y9%,)&&’01*’$%,&7+’%)#0%
(&#,’%*)%+’0#;’&’+%+#’$’0%<7’0%(&#,’$%#4%YZO%&7&10%().’&%5&#+$%*"1*%&’,’#;’%W).’&%U)$*%3C710#[1*#)4%8WU39J%
N9%’$*#?1*’$%1%U\N%’?#$$#)4%(&#,’%-1$’+%)4%(7-0#$"’+%#4*’&(&’*1*#)4%)<%5&’’4")7$’%51$%,1(%14+%*&1+’%
0’5#$01*#)4J%14+%P9’$*#?1*’$%14%1++’&%<)&%4’.%&’C7#&’?’4*$%<)&%70*&1L0).%$70<7&%+#’$’0%D%
B)&%*"’%&’C7#&’+%,&7+’%)#0%(&#,’%#4(7*%1$$7?(*#)4$%=>3@%7$’+%*"’%:>T\3%34’&56%=4<)&?1*#)4%
/+?#4#$*&1*#)4%83=/9%/44710%34’&56%\7*0))2%NOOS%8/3\OS9%"#5"J%0).J%14+%?#+%8&’<’&’4,’9%,1$’$D%%=4%*"’%
/3\OS%&’<’&’4,’%,1$’%*"’%NOO]^%(&#,’%)<%,&7+’%.1$%(&)A’,*’+%*)%+’,&’1$’%<&)?%^ZPH--0%#4%NOOS%*)%
^_‘H--0%#4%NOY]D%%=4%a16%NOOSJ%5#;’4%,&7+’%(&#,’$%*"1*%&145’+%<&)?%^YYO%*)%^YPOH--0%14+%bca3d%
<7*7&’$%?1&2’*$%1-);’%^YOOH--0J%/3/%&’C7’$*’+%*"1*%=>3@%$#?(06%7$’%^YYOH--0%);’&%100%6’1&$%1$%*"’%
%
2
?)+’0%#4(7*D%%!"’&’<)&’%+#’$’0%<7’0%(&#,’$%7$’+%#4%’,)4)?#,%1$$’$$?’4*$%#4%*"’%/0*’&41*#;’%34’&56%14+%
@’4’.1-0’%34’&56%B74+%$)0#,#*1*#)4$%1&’%-1$’+%)4%,)4$*14*%&’10%^YYOH--0%,&7+’%)#0%(&#,’$D%
!"’%3=/F$%/44710%34’&56%\7*0))2%NOO‘%31&06%@’0’1$’Y%(7-0#$"’+%#4%T’,’?-’&%NOOS%&’;#$’$%/3\OS%
(&)A’,*#)4$%$7-$*14*#1006D%%=*%(&’+#,*$%*"1*%V.)&0+%,&7+’%)#0%(&#,’$%&’,);’&%<&)?%1%4’1&L*’&?%+’,0#4’%14+%
&’1,"%^YPO%(’&%-1&&’0%8#4%NOOZ%+)001&$9%-6%NOPOXD%
B)00).#45%(7-0#,1*#)4%)<%*"’%1**1,"’+%=>3@%&’()&*J%/3/%14+%=>3@%?1+’%?#4)&%?)+#<#,1*#)4$%*)%#?(&);’%
*"’%(&)A’,*#)4$D%%!"’$’%#4,07+’%
YD /++#45%14)*"’&%6’1&%)<%WU3%<7’0%(&#,’%+1*1%*)%*"’%&’5&’$$#)4$%)<%<7’0%(&#,’%)4%,&7+’%)#0%(&#,’D%
ND =+’4*#<6#45%())&%&’5&’$$#)4$%*"1*%.’&’%+7’%*)%?#$$#45%)&%74&’(&’$’4*1*#;’%+1*1%14+%$7-$*#*7*#45%
<7’0%(&#,’%’$*#?1*’$%<&)?%$#?#01&%,)??74#*#’$D%
=>3@%(&)A’,*#)4$%’4+%#4%NOPOD%%>#4,’%*"’%’,)4)?#,%14106$’$%)<%"6+&)’0’,*&#,%14+%)*"’&%(&)A’,*$%’K*’4+%
(1$*%*"#$%6’1&J%/3/%<)00).’+%’1&0#’&%(&1,*#,’%)<%<&’’[#45%<7*7&’%<7’0%(&#,’$%1*%NOPO%0’;’0$D%
9,/#)":*4’,.*
e’1*#45%<7’0%(&#,’$%<)&%/4,")&15’J%B1#&-142$J%f74’17J%g’41#J%g’*,"#214J%W10?’&J%14+%R1$#001%.’&’%
(&)A’,*’+%7$#45%1%$#?#01&%?’*")+%1$%&7&10%+#’$’0%().’&%<7’0%(&#,’$%8$’’%/((’4+#K%/9D%
T’0#;’&’+%"’1*#45%<7’0%(&#,’$%#4%&7&10%,)??74#*#’$%.’&’%(&)A’,*’+%-6%1++#45%7(%*)%^YH510%*)%*"’%
(&)A’,*’+%WU3%+#’$’0%().’&%<7’0%(&#,’%+’(’4+#45%)4%").%?7,"%<7’0%1%<1,#0#*6%7$’$%’1,"%6’1&G%
B1,#0#*6%:$15’%8510H6&9%/++’&%*)%T#’$’0%W&#,’%8^H5109%
O%h%YJOOO%^YDOO%
YJOOY%h%NQJOOO%^ODQO%
NQJOOO%h%YOOJOOO%^ODNQ%
YOOJOOO%i%O%
;/8#1/*
j1$’+%)4%+#$,7$$#)4$%.#*"%M)0+’4%k100’6%30’,*&#,%/$$),#1*#)4J%/3/%1$$7?’+%1%(&#,’%)<%^NDZ‘H510%#4%NOOS^%
<)&%41(*"1%5#;’4%*"’%^YYOH--0%,&7+’%(&#,’%1$$7?(*#)4D%
;/#’$/.*</-*5$)(,-*
=>3@F$%?’*")+$%<)&%41*7&10%51$%(&#,’%(&)A’,*#)4%1$$7?’+%*"1*%(&#,#45%("1$’$%#4%<&)?%,7&&’4*%,)4*&1,*$%*)%
e’4&6%e7-%-6%NOYQ%8$’’%/((’4+#K%/9D%%>#?#01&%*)%,&7+’%)#0%(&)A’,*#)4$%=>3@%&’<’&&’+%*)%/3\OS%"#5"J%0).%
14+%&’<’&’4,’%,1$’$%<)&%e’4&6%e7-%(&#,’%(&)A’,*#)4$D%%!"’%/3\OS%&’<’&’4,’%,1$’%(&)A’,*’+%*"’%e’4&6%
e7-%(&#,’%#4%NOO]^%*)%5&1+71006%#4,&’1$’%<&)?%^QDSZH??j*7%#4%NOYQ%*)%^ZDNNH??j*7%#4%NOPOD%%M#;’4%
bca3d%(&#,’$%1-);’%^YYH??j*7%#4%a16J%<)&%*"’%(7&()$’$%)<%*"’%/0*’&41*#;’%34’&56%14+%@’4’.1-0’%
34’&56%B74+%$)0#,#*1*#)4$J%/3/%&’C7’$*’+%*"1*%=>3@%7$’%1%0)45L*’&?%(&#,’%)<%^SDZ_H??j*7%8^‘DOOH?,<9D%
3=/F$%/44710%34’&56%\7*0))2%NOO‘%(&)A’,*$%41*7&10%51$%1*%^‘DNQH??j*7%-6%NOPO%8NOOZ^9D%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1 http://www.eia.doe.gov/oiaf/aeo/index.html
%
3
%
688,"&)=*6>**6./-?/*4’,.*5$)(,*5$%@,(#)%"-*ABBCDABEB*
1 17 April 2008
Alaska Fuel Price Projections 2008-2030
prepared for:
Alaska Energy Authority
prepared by:
Steve Colt
Ben Saylor
Nick Szymoniak
Institute of Social and Economic Research
University of Alaska Anchorage
16 April 2008
ISER Working Paper 2008.2
2 17 April 2008
Contents
Introduction.....................................................................................................................................2
General methods and assumptions..................................................................................................2
Natural Gas .....................................................................................................................................3
Fuel Oil ...........................................................................................................................................8
References.....................................................................................................................................10
Introduction
We generated Low, Medium, and High case fuel price projections for the years 2008-2030 for
the following fuels:
• Incremental natural gas in Southcentral Alaska delivered to a utility-scale customer
• Incremental diesel delivered to a PCE community utility tank
• Incremental diesel delivered to a home in a PCE community
• Incremental home heating oil purchased in Anchorage, Fairbanks, Juneau, Kenai,
Ketchikan, Palmer, and Wasilla
This memorandum provides documentation of the assumptions and methods that we used. Two
companion Excel workbooks contain the detailed projections.
General methods and assumptions
Base year and time horizon
Our projections run from 2008 to 2030. They are computed and reported in inflation-adjusted
year 2006 dollars, consistent with the EIA Annual Energy Outlook 2008. We recognize that a
“projection” for 2008 is unlikely to match actual 2008 data. However, much of the data that we
rely on has only been published through 2006 or 2007.
Ultra low sulfur diesel premium
We include a 5-cent additional cost starting in year 2008 for rural areas only, to account for the
additional refining costs of ultra low sulfur diesel. This value can be quickly changed within the
workbook
Carbon pricing
For the Low case, we assume that the cost of greenhouse gas allowances is zero.
For the Medium case we assume that the price follows a trajectory broadly consistent with
relatively lenient national policy such as that embodied in the Bingaman-Specter “Low Carbon
Economy Act.” This legislation requires cap-and-trade but contains a safety valve provision that
caps the allowance price. While various versions of the bill have been generated, our scenario is
3 17 April 2008
based on a maximum allowance price of about $10 per metric ton CO2-equivalent (in year
2006$) beginning in 2012 and rising 5% in real terms each year thereafter.
For the High case we have assumed an allowance price trajectory based on the MIT Future of
Coal study.1 The MIT group described their “High CO2 Cost” case as a $25 per metric ton CO2
allowance cost measured in 1997$, imposed in 2015 and increasing at 4% per year above
inflation thereafter. We have adjusted the 1997$ for inflation through 2006 and also assumed
that the price trajectory begins in 2010 at a lower level that passes through the MIT benchmark
in year 2015.
Figure 1 summarizes the assumed carbon price trajectories.
Figure 1. Carbon price trajectories (year 2006$ per metric ton CO2)
Natural Gas
Background
Chugach Electric Association (CEA) has four locations where it generates electricity using
natural gas: Beluga, Bernice Lake, International Airport Road, and Nikiski. Marathon provides
all of the gas to Bernice Lake, International and Nikiski. Marathon also provides 40% of the gas
to Beluga. ML&P, ConocoPhillips and Chevron provide 20% each of CEA’s gas needs at
Beluga.
Existing gas supply contracts specify a formula for determining the contract price. The price for
each CEA contract is based on the same general formula with adjustments made on an individual
1 Massachusetts Institute of Technology. 2007. The Future of Coal: Options for a Carbon-Constrained World.
(March). Available at: http://web.mit.edu/coal/
4 17 April 2008
contract basis. In developing these projections we used the basic CEA gas contract formula with
no adjustments or added transportation costs.
Under the basic formula, the contract price of gas is determined by the inflation-adjusted, or
“indexed”, prices of fuel oil, natural gas and crude oil.
The contract price in $/mcf is:
$1.35
x {(reference natural gas – 84.54 ¢/mcf)/84.54
+ (reference fuel oil – 74.93 ¢/gallon)/74.93
+ (reference crude oil - $17.08 /barrel)/17.08}/3
The reference price for natural gas is the U.S. Bureau of Labor Statistics (BLS) Producer Price
Index natural gas component index. The fuel oil reference price is the fuel oil component index
of the BLS Consumer Price Index for all urban consumers (CPI-U). The crude oil reference price
is the West Texas Intermediate (WTI) spot price traded on the New York Mercantile Exchange
(NYMEX). The contract price is adjusted quarterly and the reference price of the 12 month
average price lagged one quarter (the average of the months from 15 to 3 months previous.
Figure 2 shows our estimates the historical price of CEA purchased gas using this formula.
Figure 2. CEA estimated historical gas purchase price (2006$ per mcf)
source: ISER estimates based on contract formula and formula inputs.
5 17 April 2008
Key Assumptions
Assumption 1. Any project that offsets electric use on the Railbelt would offset Cook Inlet gas
purchases. Chugach Electric is the marginal user of Cook Inlet natural gas. For this reason we
assume that the cost of marginal or incremental gas is the price that Chugach pays and will pay
for gas.
Assumption 2. All future contracts will be based on Henry Hub gas prices. As each existing
contract expires it will be replaced by a new contract or an amended old contract based on Henry
Hub prices. The Henry Hub price includes the transportation costs and taxes associated with
moving the gas from the wellhead to Henry Hub so it is not necessary to add any further taxes or
transportation costs to the Henry Hub price.
Assumption 3. In 2009 new contracts begin to come into use based on the Henry Hub price. By
2014 all gas contracts are based on the Henry Hub price. This is consistent with the method used
in the CEA’s Integrated Resource Plan2.
Assumption 4. The fraction of total gas priced based on Henry Hub steadily increases from zero
to 100 percent during the transition period from 1 January 2009 through 31 December 2014.
Figure 3 illustrates this pattern for the Medium case.
Assumption 5. Absent compelling evidence for some other approach in a particular case, none
of the pipeline transportation costs or local distribution costs are avoidable when considering
Cook Inlet gas. Therefore no additional components are added to the price projection here.
2 http://www.chugachelectric.com/news/projects/CEA_Integrated_Resource_Plan.pdf
6 17 April 2008
Figure 3. Assumed shares of incremental gas supply from existing contract terms
and from Henry Hub during contract transition phase, Medium case,
(2006$ per mcf)
Projection method
1. Obtain the EIA Short-Term Energy Outlook (SEO) and the Annual Energy Outlook (AEO).
These contain Low, Medium, and High projections of Henry Hub natural gas spot prices.
2. Convert EIA projections to appropriate base year real dollars, if necessary, using the CPI-U.
For these projections we did not make any adjustment because the AEO 2008 numbers and our
projections are both in year 2006 real dollars.
3-5. Project the base contract formula reference prices for natural gas, fuel oil, and crude oil to
use during the 2009-2014 transition period:
3. The reference price for natural gas in the contract formula is the BLS PPI commodity price for
natural gas. Since no forecast exists for this price, use the existing regression in the workbook
that relates the PPI index value to the Henry Hub spot price, which is forecasted by the EIA. The
short term price is forecasted using monthly forecasts from the SEO through the end of 2009.
Since the SEO and the AEO forecasts do not line up perfectly at the beginning of 2010, we use a
pro-rata formula that rolls in the AEO forecast over a 5 year period. There is only one set of
projections provided in the AEO 2008 early release for these
4. The reference price for fuel oil in the contract formula is the CPI commodity price for fuel oil.
Since no forecast exists for this price, use the existing regression coefficients in the workbook
7 17 April 2008
that relate the CPI index value to the SEO forecast for residential heating oil (including taxes)
and the AEO forecast for commercial distillate fuel. The SEO forecast is used to 2009 then is
transitioned to the AEO forecast. There is only one set of projections provided in the AEO 2008
early release for this component, so the same numbers are used for this component in the Low,
Medium, and High cases.
5. The reference price for crude oil is the WTI spot price. This price is forecasted in both the
SEO and the AEO. The SEO is used to forecast through 2009 then the forecast is transitioned to
the AEO forecast. There is only one set of projections provided in the AEO 2008 early release
for this component, so the same numbers are used for this component in the Low, Medium, and
High cases.
6. Apply a 6-year linear interpolation to slowly increase the weight on Henry Hub projected spot
price and to decrease the weight on the projected “CEA contract” formula price during the years
2009 through 2014.
7. Beginning in 2015, use Henry Hub projected spot prices.
8. Add appropriate carbon prices to determine final projected price
Figure 4 shows the projected Medium case price for Cook Inlet natural gas using the above
procedure. This graph does not include a carbon price in order to more clearly show how the
Henry Hub and existing contract formula prices blend together during the transition period.
Figure 4. Cook Inlet gas price vs. Henry Hub gas price, Medium case, excluding
carbon allowance cost ($2006 per mcf)
8 17 April 2008
Fuel Oil
Background
Fuel oil prices are simpler (although not easier) to project because there are no existing complex
contracts with formulas to be followed. Our projections are based on EIA AEO projections of
crude oil. We use the Composite Refiner Acquisition Cost of crude oil (CORAC) as the basis for
the fuel oil projections.
Key Assumptions
Assumption 1. The price of diesel to a particular PCE utility bears a stable linear relationship to
the RAC crude price, but the coefficient is allowed to be different than 1.0 and is allowed to vary
by community. A coefficient above 1.0 indicates “percentage markup pricing” as opposed to a
straight pass-through of a crude price increase dollar for dollar.
Assumption 2. Due to uncertainties and ambiguities about the delivery points underlying the
actual price data for “delivered” fuel oil, it is currently not possible to rigorously determine a
home delivery surcharge by statistical methods. Instead, we suggest that $1.00 per gallon be used
as the avoidable cost of home delivery when a substantial amount of delivered fuel is avoided
(eg, a community district heating system or mass retrofit for biomass heating). However, when
only small amounts of home-delivered fuel are being avoided, then we suggest that the
appropriate credit for avoided delivery charges is zero.
Projection method
The fuel oil price projection is based on the imported crude oil price projection from EIA’s
Annual Energy Outlook (AEO). The mid price case is based on the AEO 2008 reference case,
but because low and high crude oil price forecasts from AEO 2008 are not available at the time
of this writing, AEO 2007 numbers were used to estimate low and high price cases for AEO
2008. In future updates, it is likely that AEO Low and High case prices will be available.
1. Obtain EIA’s Annual Energy Outlook 2007 from the following URL:
http://www.eia.doe.gov/oiaf/archive/aeo07/
Obtain the forecast Imported Crude Oil Price from Table 1 for the reference, low price, and high
price cases.
For each year, determine the ratio of the low price to the mid price and the ratio of the high price
to the mid price.
2. Obtain EIA’s Annual Energy Outlook 2008 from the following URL:
http://www.eia.doe.gov/oiaf/aeo/
Obtain the forecast Imported Crude Oil Price from Table 1 for the reference case.
9 17 April 2008
For each year, multiply the AEO 2008 reference case price by the low-to-mid ratio for that year
to produce a low price forecast, and by the high-to-mid ratio to produce a high price forecast.
3. Obtain the monthly “U.S. Crude Oil Imported Acquisition Cost by Refiners (Dollars per
Barrel)” (CORAC) from the following URL:
http://tonto.eia.doe.gov/dnav/pet/pet_pri_rac2_dcu_nus_m.htm
Calculate average CORAC for Alaska fiscal years 2002 through 2007 by averaging the prices for
months July (FY-1) through June (FY). Divide by 42 to obtain price per gallon.
3. Obtain the PCE Statistical Reports for fiscal years 2002 through 2007 from the following
URL:
http://www.aidea.org/aea/programspce.html
4. Extract the field “Average Price of Fuel” for each community and year. Some communities
without data require data from other communities as a proxy. The proxy communities suggested
by AEA, listed with the original community first, then the proxy, are as follows:
• For Dot Lake: Substitute: Tok
• Hollis: Craig
• Klawock: Craig
• Thorne Bay/Kasaan: Craig
• Kasigluk: Nunapitchuk
• Pitkas Point: St. Mary’s
Make the following additional substitutions:
• Chignik Lake: Chignik Lagoon
• Klukwan: Kake
• Kobuk: Shungnak
• Napakiak: Napaskiak
Perform these substitutions not by copying data points from the proxy community into the
missing slots, but by copying the regression coefficients (calculated later) from the proxy
community.
5. Adjust the CORAC FY averages and PCE fuel prices for inflation to 2006 dollars using the
U.S. Consumer Price Index for All Urban Consumers (http://www.bls.gov/CPI/).
6. For each PCE community, do a linear regression with PCE fuel price per gallon as the
dependent variable and CORAC per gallon as the independent variable, each year being one
observation for that community consisting of a crude oil price and a heating oil price. (There are
only six years of data for each community.)
7. Use the regression coefficients to predict fuel oil price per gallon for each PCE community as
a function of Imported Crude Oil Price per gallon (Low, Mid, and High cases) for each year from
2008 to 2030.
8. The above prices are for utilities. For avoided use of home-delivered fuel, add $1/gal if a
significant amount of fuel is avoided (no clear relationship was found between surveyed home
10 17 April 2008
heating oil prices and PCE utility fuel prices, but the average difference was about $1.00). Do
not add anything if only a small amount is avoided, as most of the costs of storage and delivery
can only be avoided in “lumps.”
9. For urban places (Anchorage, Fairbanks, Juneau, Kenai, Ketchikan, Palmer, Wasilla), obtain
prices for heating oil from Alaska Housing Finance Corporation’s annual fuel price surveys
conducted in years 1999 through 2007 (contact ISER or AHFC to obtain this data). Use the
average of #1 and #2 heating oil. Where prices are missing, use the price included in the Alaska
Food Cost Survey conducted for December (http://www.uaf.edu/ces/fcs/) (there will still be
some missing data points).
10. To obtain crude oil prices corresponding to the time frame of the heating oil prices, calculate
the average CORAC per gallon for October through December of each year from 1999 to 2007
in nominal dollars.
11. For each place and year, subtract the average CORAC just calculated for that year from the
fuel price for that place and year. Put this difference into real 2006 dollars using the same CPI as
above. Put the average CORAC numbers in real 2006 dollars as well.
12. For each place, do a linear regression with the price difference as the dependent variable and
CORAC as the independent variable, each year being one observation for that place consisting of
a fuel-crude price difference and a crude oil price.
Use the regression coefficients to predict the difference between fuel price and CORAC for each
place and year as a function of Imported Crude Oil Price per gallon (Low, Mid, and High cases)
for each year from 2008 to 2030. Add to these the projected CORAC to obtain a projected
heating oil price.
References
Massachusetts Institute of Technology. 2007. The Future of Coal: Options for a Carbon-
Constrained World. (March). Available at: http://web.mit.edu/coal/