A. Parties' Positions
Although the Commission provided a broad standard in D.02-03-055 for bundled ratepayer indifference relating to DA suspension, the specific methodologies to implement that standard were left for this proceeding. Parties disagree in a number of respects concerning the manner in which indifference costs should be computed and assigned among DA customers. Parties' disagree on various issues, including whether indifference should be determined only with reference to DWR costs, or whether it should incorporate the entire procurement portfolio, including both DWR and utility-related costs. Among those who agree that utility-related costs should also be considered, there is disagreement as to whether the calculation should incorporate all URG costs, including below-market resources, or be limited to those specific categories of so-called above-market transition costs authorized for recovery under Public Utilities Code Section 367. There is also disagreement as to how the above-market component should be calculated, and what form of market proxy should be used.
There is also disagreement in how the cost-shifting effects of customer migration from bundled to DA load should be captured and measured. Most parties rely on a computer-simulation modeling approach to compute the net difference in DWR procurement-related costs on a before-and-after-DA-suspension basis. The simulations thus compare the cost difference between: assumed bundled load that (1) includes the incremental load that migrated to DA between July 1 and September 20, 2001, and (2) excludes that incremental DA from bundled load. The difference represents the cost-shifting effects of the DA migration.
DWR's approach calculates the change in the unit cost of the total net short (i.e., the DA load served via DWR's long-term contracts and DWR's spot purchases) between alternate suspension dates of July 1 DA load of 2% and September 20, 2001 DA load of 13.62%.15 DWR thus defines "indifference" to mean that the rates paid by bundled customers should not increase as a result of suspending DA as of September 20, 2001 rather than July 1, 2001. The difference in costs between these two DA load levels represents the increase in the average cost of net short power to bundled customers due to the migration of customers from bundled to DA load between July 1 and September 20, 2002. The cost differential represents the portion of the DWR revenue requirement incremental DA customers would need to pay to avoid cost shifting to bundled customers. In modeling indifference costs, DWR focused on only its own costs and ignored utility-related costs.
CLECA, CMTA, SCE, and TURN (among others) agree with DWR's general approach of comparing costs based on the change in incremental DA load between these two dates, but disagree with focusing only on DWR power. CLECA's approach defines indifference in reference to the change in unit cost of the total bundled service portfolio (i.e., DWR's long-term contracts, DWR's spot purchases, and the IOUs' URG) between the two suspension dates.16 CLECA et al. point out that the DWR power represents only a fraction of the power sources serving bundled customers. DWR's method thus assigns zero uneconomic DWR costs to the portion of bundled customers' load served with URG resources.
The majority of power used to serve bundled customers comes from URG sources. The DWR power share of total resources varies by utility and changes over time. In all cases, the DWR share of total power requirements in any given year will reflect the amount of utility nuclear generation (which varies when there is refueling), weather, and hydro availability, for PG&E and to a lesser extent SCE. Furthermore, the share will be influenced by load growth and the percentage of DA load. Under the DWR/Navigant approach, the cost of the bundled portfolio actually declines under a September 20 suspension date, once the DA cost responsibilities are included.17 CLECA proposes as an alternative that the Commission does not focus on DWR costs alone but rather on the entire bundled energy portfolio costs.
The total cost of generation used to serve bundled customers is the combined weighted average cost of both URG and the DWR power. DWR power has been, on average, more expensive than the weighted average cost of URG power, to date. DWR's own analysis shows its average power prices to finally drop to $69 to $70/MWh after several years. PG&E's URG cost under the recent URG decision is about $52/MWh. SCE's is about $52/MWh, and SDG&E's is about $57/MWh.
If DA customers leave bundled service, their share of URG power is thus made available to serve remaining bundled load. DA customers will not receive DWR power either, and any excess DWR power from non-dispatchable sources can be sold in the market. Fixed costs, however, will still have to be covered. The departure of the DA load will leave more of the lower cost URG power available to serve bundled customers and help offset the impact of DWR power costs.
The CLECA approach mixes DWR and URG unit costs into a single blended rate, and does not segregate a rate just representing URG-related costs. Both SCE and SDG&E argue that the CLECA methodology needs to be refined to provide for a separate URG rate because under their proposals, all DA customers will pay a CTC rate, but not necessarily a DWR rate. SCE and SDG&E have differing proposals as to how the indifference calculation should be made, but both agree on the overall approach which incorporates a separate calculation of above-market URG costs based on a market proxy. SDG&E defines indifference as (1) payment by migrated DA customers of their share of post-migration above-market DWR long-term contracts costs, and (2) payment by DA customers of their share of AB1890 above-market URG costs.18
CMTA proposes an alternative approach to that of DWR and CLECA. CMTA defines indifference as there being no change in the amount of above-market DWR costs paid by bundled service customers between the two suspension dates and allocating above-market URG costs to both DA and bundled service customers.19 Under CMTA's alternative approach, there is no specific comparison of the cost difference between DA loads at discrete points in time. DA and bundled service customers would each be allocated an equal cents per kWh charge for the recovery of uneconomic DWR costs.20 By contrast, CMTA claims that Navigant's method results in incremental DA loads bearing 39% of uneconomic DWR costs in 2002, even though they represent only 11.6% of total loads.21 CMTA argues that Navigant's approach thus does not result in ratepayer indifference, but actually leaves bundled ratepayers with lower rates.
Under CMTA's approach the uneconomic costs of the DWR portfolio would be determined by comparing per-unit costs of the DWR contracts against a market benchmark price based on the all-in costs of a new gas-fired combined-cycle power plant. CMTA notes that this is the same benchmark that the Commission used in its FERC complaint concerning DWR contracts. This market proxy includes both variable and fixed capital costs.
B. Discussion
We conclude that the comparison of the difference in costs between incremental DA load in and out between July 1 and September 20, 2001 more closely conforms to the intent of D.02-03-055 than does the CMTA method. Specifically, the key elements of our adopted methodology shall be based on the alternate DA suspension dates, consistent with the objective of D.02-03-055 that we adopt surcharges in lieu of an earlier suspension date. Thus, the adopted surcharges computed on this basis shall ensure bundled service customers are indifferent to costs under the two suspension dates of July 1 or September 20, 2001.
We conclude that the CMTA approach does not satisfy the Commission requirement that bundled service customers be indifferent between two discrete suspension dates. CMTA's method provides no connection between the alternate suspension dates that can be tied to bundled service customer indifference with respect to costs. CMTA's proposal also incorporates the use of a market proxy to measure uneconomic costs. We address the issue of market proxies in Section XIV.
We also find that the proper approach to computing ratepayer indifference must take into account the total portfolio of energy sources, not just those provided by DWR. ORA objects to CLECA's indifference approach, arguing that the cost of URG resources are "off limits" to DA customers, but are dedicated to service of bundled customers. ORA argues that it blurs the distinction between DA and bundled service to assign an offsetting savings to DA customers.
The intent underlying the indifference calculation, however, is to determine the cost shifting that resulted from the migration of certain bundled customers to DA. An accurate measure of cost shifting cannot be determined if we selectively focus only on certain components of cost shifting while ignoring others. The directive in D.02-03-055 was to consider all cost shifting, not just those effects attributed to the DWR portion of the total portfolio. The netting of URG savings does not imply that those URG resources are somehow dedicated to serving DA customers. The attribution of savings to DA customers merely reflect the change in costs experienced by bundled customers associated with their use of those dedicated resources.
The total portfolio approach to computing bundled ratepayer indifference, as adopted herein, will require the computation of two rate components, one relating to remittances to DWR and the other relating to payment to the utility for utility-related uneconomic costs.
The calculation of indifference costs on a total-portfolio basis still incorporates the use of the DWR modeling of costs on a DA in/out basis. The DWR model already incorporates variables for both DWR and URG resources to determine resources to be dispatched. Although DWR's model scenarios only focused on the costs associated with its long-term contracts and spot-market purchases, both the DWR and the URG costs for the pre- and post-DA migration scenarios are available from the DWR-supplied spreadsheets. This DWR modeling information can thus be used to compute an indifference cost on a total portfolio basis. Once the total indifference cost level is determined, the DWR portion of that indifference cost can be identified by calculating the above-market cost and related kWh of the IOUs'own resources and subtracting that from the total portfolio indifference cost. The CLECA total portfolio methodology mixes URG and DWR revenue requirements. Therefore, a separate benchmark must be determined to identify the stand-alone, uneconomic portion of URG. This stand-alone component is needed because those continuous DA customers who will not pay DWR-related CRS, will still be responsible for utility-related CRS.
PG&E points out that the split between URG and DWR components of CRS does not affect the aggregate division of costs between bundled and DA; it will affect the allocation among classes of bundled customers. The effect is due to the fact that the ongoing URG and DWR charge are established using different cost allocations. Thus, the larger the ongoing URG component, the more the allocation of the DA total portfolio indifference amount is weighted toward top 100 hours, and the less toward equal cents per kWh allocation.
Accordingly, we shall adopt a DA CRS component representing the above-market portion of the URG portfolio for each utility. To the extent the utility operates its URG portfolio to meet bundled service load, its variable costs of operation will be at or below the alternative costs of procuring energy in the market. Nevertheless, the economics of fixed and variable costs within the portfolio will vary yearly depending on market conditions. For example, baseload generation may be more costly than market purchases during off-peak hours, but less costly than market purchases during on-peak hours.
The above-market portion should consist of the difference between the cost (revenue requirement) of the URG portfolio and an estimate of its value in the market.22 This CRS component shall be calculated using the same "stranded cost" approach the Commission previously adopted for the calculation of the CTC. This will ensure that DA customers will be responsible for the same proportional share of "stranded costs" as bundled service customers will bear. This charge shall then be deducted from the indifference cost calculation to determine the amount that should be remitted to DWR. We consider the issue of a market benchmark at Section XIV.
15 DWR/McMahon, Ex. 4; DWR/McDonald, Ex. 8. 16 CLECA/Barkovich/Yap, Ex. 28, p. 36. Strategic Energy proposed a method similar to CMTA, except that it involves liquidating a portion of DWR's contracts by the amount of increased DA load and assigning the contract cost above the revenue derived from the liquidation to DA customers. (Strategic Energy/Lacey, Ex. 37, p. 5) We find no evidence, however, that DWR is willing to liquidate a portion of its contracts. (Strategic Energy/Lacey, Tr. 6/767.) Even if DWR was willing to liquidate a portion of the contracts, there is no evidence that the exact portion associated with the increase DA level could be liquidated and liquidated in a manner equitable for each of the IOU service territories. 17 Id., p. 25. 18 SDG&E/Trace, Ex. 54, pp. 5 - 7; SDG&E/Nelson, Ex. 57, pp. 1 - 4. 19 CMTA/Beach, Ex. 39, p. 10 - 19. 20 CMTA, p. 4. 21 CMTA's supporting calculation of the relative allocation of uneconomic costs using the Navigant method is set forth in Table 5 of CMTA Exhibit 39. 22 SCE also proposes to include the Independent System Operator (ISO) costs associated with the operation of this portfolio in this cost responsibility.