95725.docx

DECISION MEMORANDUM
TO:COMMISSIONER NELSON
COMMISSIONER SMITH
COMMISSIONER HANSEN
MYRNA WALTERS
TONYA CLARK
STEPHANIE MILLER
DAVID SCHUNKE
DON HOWELL
RICK STERLING
GARY RICHARDSON
WORKING FILE
FROM:BRAD PURDY
DATE:JULY 25, 1995
RE:ADJUSTMENTS TO AVOIDED COST MODEL
STAFF'S PROPOSAL TO MODIFY THE AVOIDED COST MODEL
Staff proposes to make several changes to the avoided cost model to more accurately reflect costs of a gas-fired combined cycle combustion turbine (CCCT), the surrogate upon which the avoided cost rates are based for projects smaller than 1 MW.  Staff is not proposing to change any of the generic or utility-specific variables adopted by the Commission in the recent avoided cost case, but rather is simply proposing changes in the method in which these variables are used to calculate avoided cost rates.  Some of the proposed changes correct inconsistencies in the calculation methods employed for fueled and non-fueled rates, others more accurately levelize costs, while still others are cosmetic only and are intended to better accommodate annual updates to generic variables.
In addition to model changes, Staff wishes to suggest alternative methods of publishng avoided cost rates.  Current methods Staff submits, are deceiving, difficult to explain and difficult for many people to interpret.
Finally, Staff wishes to make a proposal concerning annual updates to avoided costs.  Clearly, the adjustable component of avoided costs is intended to capture changes in the cost of natural gas and should be applied to fueled rates.  However, it has never been addressed whether the starting fuel price assumed for non-fueled rates should be re-set annually or left unchanged from the price set in the most recent avoided cost proceedings.
Inconsistency in Calculation of Levelized Rates for Non-Fueled Projects
Currently, in computing levelized rates for non-fueled projects, surplus energy prices prior to the first deficit year are not being included.  Instead, the levelized rates are based on fuel costs, fixed O & M, and variable O & M during the deficit period.  This is inconsistent with the computation methods used for non-levelized rates for non-fueled projects, and for both levelized and non-levelized rates for fueled projects.  This inconsistency was pointed out in Staff's post-hearing brief, but the Commission chose to forego making a model change because of its minimal effect on rates and because it could potentially affect only very short term levelized contracts.  However, the effect of the inconsistency becomes more significant as utilities' first deficit years are extended further into the future as in the case of WWP.
Staff's proposed change is to base all rates prior to the first deficit year on surplus energy prices, whether for fueled, non-fueled, levelized, or non-levelized contracts. The effect of this change alone is very minor, and the result is slightly lower levelized, non-fueled rates.  
Levelization of Surplus Energy  Rates
Staff also proposes that surplus energy prices be included in computing levelized rates for fueled and non-fueled projects.  Exhibit 1, which represents Idaho Power's current levelized and non-levelized avoided cost rates for fueled projects, illustrates the problem that exists when the current calculation methodology is employed.  According to the tables, a three-year non-levelized contract would receive payments of 19.86, 20.75, and 21.68 mills/kwh in 1995, 1996 and 1997 respectively.  However, if a levelized contract was signed instead, three equal payments of 21.68 mills/kwh would be received in each of the three years.  Clearly, the value of a levelized contract is greater than the value of a non-levelized contract.  Exhibit 2 further illustrates the problem by comparing the net present value of levelized and non-levelized contracts for various on-line dates and contract lengths.  Except for contracts which do not begin until after the utility's first deficit year, the value of levelized contracts always exceeds the value of non-levelized contracts.  Again, this is occurring because the surplus energy prices prior to the first deficit year are not being included in the levelization.
Staff has prepared several exhibits using Washington Water Power as an example to illustrate the proposed adjustment to the existing methodology.  Exhibit 3 represents non-levelized rates for fueled projects.  As indicated in the exhibit, there are three distinct parts used to calculate the rates:
A.  Fuel costs,
B.  Capital plus O & M costs, and
C.  Estimated market energy prices during the surplus period.
The capital and O & M component, and the surplus energy component are non-adjustable.  They are escalated at pre-established rates, and thus are considered fixed costs.  The fuel cost component is shown in Exhibit 3 as being level throughout the term of the contract.  However, for fueled contracts, fuel prices will be adjusted annually based on actual fuel prices at Sumas Washington.  Thus, actual future fuel prices are unknown, but are expected to increase.  Since the fuel component is adjustable, each year's fuel cost is simply added to the non-adjustable components, thus making up the total avoided cost payments for that year.  
Currently, the surplus component (C) is not being levelized.  Staff contends the capital plus O & M component (B) and the surplus component (C) should be levelized in computing levelized rates since both represent fixed costs.  When these two components are levelized, the levelized rates for just these two components appear as shown in Exhibit 4.  
Finally, if fuel costs are added on top of these two levelized components, the total levelized rates appear as shown in Exhibit 5.  Note that for all exhibits, non-levelized rates are depicted as bars and levelized rates are depicted as solid shading.  All exhibits assume an on-line date of 1995.  Consequently, where non-levelized rates are shown, the bars represent the actual avoided cost payments in each of the corresponding years.  For levelized rates however, since the same rate is paid in each year of the contract (except in the case of fueled projects where an adjustable component is added every year after the surplus period), the shading represents the levelized rate that would be paid for various contract lengths.
In carefully examining the rates shown in Exhibit 5, several things are apparent.  First, there is a big jump in rates after the first few years.  This occurs at the utility's first deficit year when rates begin to be based on CCCT surrogate costs and when fuel costs are included.  Second, rates after the first deficit year appear quite flat, and in fact, slightly decrease in the middle years.  The flatness is because the same fuel cost is assumed for all years of the contract.  The fuel cost component will be adjusted annually however, and will almost certainly increase.  This will result in an increase in the total avoided cost in future years, rather than the relatively flat rates shown.  The dashed line indicates how rates might increase if gas prices escalate as expected.  The slight "dip" in the rates in the middle years is caused by the levelization of the surplus and capital plus O & M components, and the fact that the surplus rates exceed the capital plus O & M components in the years immediately following the first deficit year. The manner in which levelized rates for fueled projects are presented is critical to insure that calculated rates are accurately represented.  Presentation of rates is discussed in detail in a later section of this application.  
Exhibit 6 shows the three avoided cost components for non-fueled projects.  The components are the same as before, except that fuel costs are assumed to escalate at the six percent rate adopted by the Commission.  Since none of the cost components are adjusted annually and since all increase, Staff contends all three components represent fixed costs and should be levelized.  Currently, just the fuel, and capital and O & M components are being levelized.  When all three components are levelized, the resulting levelized rates appear as shown in Exhibit 7.
In addition to these changes in the levelization methodology, Staff proposes that fixed O & M and variable O & M be combined in the model's computations, since both are escalated at established rates and neither are subject to annual adjustment.  In other words, variable O & M is variable but not adjustable; consequently, it should be treated the same as fixed O & M.  The effect of all of the computational changes is to increase the levelized rates for both fueled and non-fueled projects.  The increase ranges from less than one mill for non-fueled levelized rates to nearly 10 mills for fueled levelized rates, depending on the utility.  The slight increase  in the non-levelized rates (1.45 mills for a 20 year contract) is due to the variable O & M component being escalated and included in the computations, instead of being added later and adjusted annually along with fuel costs.  Exhibit 8 shows rates computed for each utility using the current methodology.  Exhibit 9 shows rates computed using Staff's proposed new methodology.  Exhibit 10 shows the effect of the model changes on rates for 20-year contracts for each utility.  Staff contends that rates calculated using the proposed methodology more accurately reflect the levelized and non-levelized costs of the CCCT surrogate.
To verify the comparability of levelized and non-levelized rates using the proposed methodology, Staff compared the net present values of contracts of various lengths and with different on-line dates.  Exhibit 11 shows that the net present values are identical for levelized and non-levelized contracts for each utility.
A cell-by-cell summary of Staff's proposed changes to the avoided cost model is included as Exhibit 12.  A diskette containing the model is available from Staff upon request.
Presentation of Rates
Staff also wishes to suggest alternative ways of presenting avoided cost rates and solicit comments from the utilities on which alternative is preferred.  Whichever method is chosen, Staff recommends that one standard be adopted for all utilities to maintain consistency.
Exhibit 13 illustrates two alternatives for displaying rates.  Method "A" is the method currently used.  The primary criticism of this method is that it is very deceiving because it includes in all rates only the current year's adjustable portion, even in the rates shown for future years.  Since the adjustable portion will likely increase in the future, the rates shown are not truly representative for future years.
Methods "B" and "C" are two alternatives for presenting rates.  If method "A" is selected, then the entire table of rates must be prepared annually.  If method "B" or "C" is chosen, then the tables of rates will not change annually but the adjustable portion listed below the tables will change annually.  Staff recommends Method "B" or "C" be used and feels they would be more easily understood by more people.
Annual Updates of Avoided Costs
Staff believes the intent of the Commission is that fuel costs, as used by the avoided cost model, be adjusted annually.  It appears clear that these annually updated fuel costs should be applied to fueled rates.  However, it is not clear whether these annually updated fuel costs should be applied to non-fueled rates.
Staff believes a non-fueled contract, once established, should not be subject to changes in rates for the duration of the contract.  The starting gas price and established escalation rate in effect at the time of contract signature should serve as the basis for rates throughout the life of the contract.  However, in the case of new contracts, the question is should new rates be published every year for new contracts based on the annually updated fuel prices?  Put another way, should a project that signs a contract in 1999 have 1999 gas prices as the starting basis for rates, or should 1995 gas prices which have been escalated at six percent for four years be used as the starting basis for the contract?
Staff proposes that the starting gas price for non-fueled rates remain fixed at the level established in the recent avoided cost case (1995 gas prices), and that new published rates not be recalculated every year using updated gas prices.  The starting fuel price and the assumed escalation rate provide estimates of future prices.  If, over the course of time, these estimates deviate too far from actual fuel prices, then a new proceeding can be initiated to re-set fuel prices and escalation rates.  Re-establishing a new starting fuel price each year based on actual prices would not necessarily be any more accurate, since in subsequent years the actual price may be higher or lower than the assumed price.  Several years of record would probably be needed to assess whether the assumed prices and escalation rates are accurately tracking actual prices over the long term.
If the starting gas price for non-fueled rates is not updated annually, then it will be unnecessary to publish new non-fueled rates each year since they will not change from year to year.  The only new tariff needed each year would be for fueled rates.  Furthermore, depending on how rates are ultimately presented, it may only be necessary to annually publish the adjustable portion of fueled rates, which is just a single number.
CONCLUSION
Staff believes that the proposed modifications to the avoided cost model are relatively non-controversial and suggests that if the Commission is inclined to initiate a formal case to consider them , that it be handled under Modified Procedure.
Commission Decision
Does the Commission wish to initiate a formal case to consider Staff’s proposed modifications to the avoided cost model?  If so, should the matter be handled under Modified Procedure?
Brad Purdy
vld/M-ACM.bp