6. What Methodology Should the Commission Use to Allocate and Collect the Revenue Requirement for Bond-Related Costs?
Almost all parties propose that the revenue requirement for bond-related costs be allocated based on some measure of kWh. However, parties differ on which kWh should be included in the allocation and whether the allocation should be adjusted to reflect certain specific factors.
A. PG&E, CLECA, SDG&E, ORA: Allocate and Collect Bond Costs Based on kWhs
The simplest position of parties is to allocate and collect, on a uniform statewide basis, per kWh charges for all bond-related costs. Of the parties in this proceeding, PG&E, CLECA and SDG&E support a per kWh charge with no exemptions.31 Although ORA proposes certain adjustments to the calculation of bond charges (as do CLECA and PG&E), ORA characterizes its position as allocating "the bond charge on a simple equal cent per kWh basis across the vast majority of kWh forecast to be sold in the service territory of investor-owned utilities (IOUs) in 2003, and 2004."32 SCE similarly characterizes its position as a uniform allocation, yet it supports ORA's proposed adjustments to the assessment of bond charges.33
Moreover, ORA provides several rationales for assessing the bond charge on a simple equal cents per kWh basis. In particular, ORA notes the expected duration of the bond charges of 20 years. ORA concludes that with the "inevitable changes in customers and circumstances"34 that will occur over this time period, these charges will be paid by customers who did not even live in California during the crisis and that some who did will not pay these charges if they move away. In light of the inability to link bond costs to a customer's consumption, ORA concludes that a simple per kWh charge is fairest.
In a similar vein, PG&E states:
"The California Legislature and the Commission have determined that DWR's actions during the energy crisis were undertaken `for the health, welfare, and safety of the people of this state' and on behalf of all ratepayers `to ensure reliable electricity service and, therefore, all ratepayers should contribute to the effort to pay down the unprecedented debt incurred by the state to help weather the energy crisis.' (See AB X1, Section 7; D.01-09-060, mimeo at pp. 3, 6). PG&E's approach is consistent with this policy because it shares the burden of the cost of the `energy crisis' on each customer based on usage. It is easy to explain to customers. It treats all California ratepayers equally."35
CLECA, in addition to the arguments listed above, notes that the bond charge will be small in relation to the overall cost of DWR power, and that this reduces "the need for utility specific allocation."36
B. ORA: Adjust PG&E's Rates in Light of WAPA Contracts
ORA proposes one specific departure from a per kWh allocation of bond charges in this proceeding. ORA recommends "that the dollar impact of the forecast net sales by PG&E to the WAPA be assigned to PG&E's ratepayers."37 ORA states that it "assumes that PG&E's ratepayers benefited from the contract between PG&E and WAPA"38 and concludes that they should bear WAPA-related costs. Noting that PG&E must provide WAPA with substantial amounts of power in 2003 and 2004, ORA recommends inclusion of these amounts in the allocation of revenue requirement between utilities, and the subsequent assignment of this revenue requirement to PG&E's other retail customers. Thus, under ORA's proposal, PG&E's customers would pay a higher bond charge than customers in the service areas of SCE or SDG&E. SCE supports this adjustment. PG&E opposes this position, arguing that it is simply an allocation based on a modified "net short" position, and, after citing ORA's own arguments for an equal allocation, argues that fairness requires the rejection of this position.
C. TURN: Allocate Revenue Requirement Per D.02-02-052
TURN takes a very different approach. TURN states that "[w]hile the equal cents methodology has the advantage of simplicity, it ignores any and all differences among the three companies that resulted in their making very different contributions to the accrual of the DWR `undercollection' during the first nine months of 2001."39 TURN argues on behalf of the allocation factors previously used in D.02-02-052, stating that the allocation factors are "generally consistent with cost causation."40 Finally, concerning ORA's proposed WAPA adjustment, TURN states "no WAPA adjustment is needed if that methodology [i.e. that of D.02-02-052] is followed here."
D. PG&E: Adjust Bond Charge to Reflect Line Losses
PG&E also proposes an adjustment in the allocation of revenue requirement to different customer categories. PG&E states "DWR bond charges should be differentiated by voltage to reflect differential line losses for different service level voltages, but otherwise set equally on all included load."41 PG&E argues that differentiating bond charges by service voltage appropriately reflects that less energy is needed "to serve a given quantity of electric consumption at transmission service voltage levels, relative to service at primary and secondary distribution service voltages."42 CLECA supports PG&E's proposed voltage-based adjustments for line losses, and SCE states that it does not object to such an adjustment.
E. EPUC, CLECA and Modesto: Adjust Bond Charge on Departing Load Customers to Exclude Revenue Requirements
Although issues associated with DL were assigned to R.02-01-011, certain parties, including EPUC, Modesto and CLECA proposed adjustments to bond charges based on the structure of the bond financing. We have not addressed the elaborate discussion from these parties regarding DL customers, as well as the responses by TURN, PG&E, and SCE, because the issues raised are outside the scope of this proceeding. We note that pursuant to the August 13, 2002 ALJ Ruling, the evidentiary record of this proceeding was incorporated into R.02-01-011. That is the proper forum for these arguments, and we make no determinations of these issues in today's decision.
F. Discussion: Allocate and Collect Bond Charges Based on
All Non-Excluded kWh Consumption
We will allocate and collect the bond-related costs on a simple per-kWh basis, spread over all customer usage, with the exceptions of residential sales below 130% of baseline, medical baseline, and CARE-eligible customer usage. This policy makes sense for several reasons. First, as ORA points out, the long period over which the bond charges will be collected breaks the linkage between those for whom the power was purchased and those responsible for repayment. In addition, these bond-related costs were incurred to stabilize the grid, which benefited everyone. Thus, the assessment of a bond charge is simply a mechanism for raising the revenues needed to repay these bond-related costs. In light of these considerations, absent a rational reason to exclude particular usage or customers, it is reasonable and equitable to allocate these bond-related costs over the largest base of customers on a simple per kWh usage basis.
Second, because the purpose of the bond charge is simply to raise revenues to pay for bond-related costs, the simplicity of the per-kWh fee recommends it. It is transparently fair to all who must pay it.
Third, the one thing that the Commission knows from the period of the energy crisis is that the prices paid for power had little relationship to the cost of producing that power. Thus, the use of the principle of "cost causation" to allocate bond-related costs, as recommended by TURN, is unwarranted, for this principle assumes a relationship between cost and price that may not have existed at that time
In particular, TURN suggests that we allocate these bond-related costs consistent with a modified "net short" position, as adopted in D.02-02-052. TURN fails to note, however, that D.02-02-052 did not allocate past responsibility for energy purchases, but instead allocated responsibility for current and ongoing purchases by DWR on behalf of the customers in the service territories of investor-owned utilities. Moreover, unlike the crisis period in which these bond-related costs were incurred, the current relationship between power prices and power costs better meets the principles of "cost causation" ratemaking.
Moreover, as we observed in D.02-02-051, we are not dealing with routine costs arising from utility operations:
"The establishment of a separate Bond Charge also recognizes the nature of the costs that DWR will finance with its bond transaction. These are costs that DWR incurred at the height of the crisis. . . Because the costs that DWR incurred to save the grid have future benefits, they should be amortized over time."43
In addition, as we noted in D.02-02-051, we have broad discretion in assessing a Bond Charge:
"The Commission's authority under Pub. Util. Code § 451 and § 701 to impose rate mechanisms such as Bond Charges extends to situations where the charge is not in proportion to the direct benefit received by each customer paying the charge. (Footnote omitted) This would be the case, for example, for future ratepayers who will pay Bond Charges despite the fact that they only received the benefits of DWR's grid-stabilizing activities, and did not receive any of the electric power that was procured by DWR during the height of the electricity crisis."44
From this discussion, it is clear that the Commission did not contemplate a strict adherence to the economic principle of "cost causation" in allocating responsibility for bond-related costs. We believe that it would not be equitable to do so.
PG&E's argument (which SCE and CLECA support) to make a voltage-related adjustment to the bond charge because it took less power to serve high voltage customers is not persuasive. In our view, this argument does not warrant a departure from the equitable principle of assigning bond responsibility to each kWh of bundled consumption not otherwise excluded. As noted in D.02-02-051, some people who have never lived in California will pay these costs even though they consumed no power during the crisis period. In light of this harsh fact, making an adjustment based on a customer's service voltage lacks a reasonable basis.
Even ORA, whose testimony and brief state that the purpose of this charge is to fund bond-related costs, succumbs to the temptation to allocate WAPA-related costs based on the notion that PG&E's customers obtained some benefit from these WAPA contracts, and therefore should pay a higher share of bond-related costs. As PG&E and TURN point out, ORA's proposal to allocate a revenue requirement based on power provided to WAPA and then collect it from other PG&E customers introduces a revenue requirement allocation based on one factor contributing to the "net short" position. We reject a strict adherence to the principle of "cost causation," for the reasons stated above. Thus, we find that this argument does not warrant a departure from our equitable decision to allocate the Bond Charge on all non-excluded consumption by bundled customers.
