5. Discussion
By D.07-09-043, we departed significantly from our previous design of shared-savings mechanisms with regard to MPS true-up provisions. In years past, we rejected the notion that utilities would have to pay back all the interim claims if the final EM&V studies indicated that the MPS had not been met-as long as the portfolio still produced positive net benefits (PEB).6 However, with the goal of designing an incentive mechanism that recognized a "dual objective function" for energy efficiency-to both achieve kWh, kW and therm savings goals and maximize dollar net benefits in the process--we adopted more stringent true-up provisions in D.07-09-043. More specifically, we directed that both the MPS (tied to the savings goals) and the PEB (a calculation of net benefits) be trued up at the final claim, subjecting the utilities to the possibility of paying back all interims claims if the final true-up based on ex post studies indicated that the MPS had not been met. Although we recognized that the possibility of refunding earnings already claimed presented certain problems for the utilities with respect to their financial reporting, we concluded that these problems could be readily addressed by (1) limiting payout of initial claims to 70% (30% hold-back) and (2) deducting any over-collections from future earnings claims.7
However, we are persuaded by the arguments presented by the utilities (and supported by NRDC) that the value of any energy efficiency earnings as a systematic part of the utility's basic business earnings will be seriously degraded, unless we modify the true-up provisions adopted by D.07-09-043. As explained in the Joint Petition, this is because the uncertainty over ex post measurement results (for net-to-gross ratios in particular) coupled with the "all or nothing" nature of the MPS makes it unlikely that the utilities will be able to book any authorized interim earnings during the program cycle under the current true-up provisions:
Utilities must be able to recognize, or book, incentives on a regular basis for accounting purposes in a manner that can be expected and anticipated by the investment community. Otherwise earnings from energy efficiency programs are not truly on par with generation resources in the minds of investors. As the incentive mechanism is currently adopted, if the utilities do not have sufficient certainty, consistent with Financial Accounting Standards Board directives, that the incentive will be earned, they cannot be booked until after the end of the program cycle when the final adjustments have been determined. If the incentives are not booked at regular intervals, they would result in a one time earnings adjustment that would likely be excluded from operating earnings, which are the basis for a company's financial valuation. The uncertainty could result in a higher cost of financing. As a consequence, the utilities would not receive the full benefit of these shareholder incentives from the financial markets.
Because of the uncertainty associated with the net-to-gross factor, for which the final results will not be known until 2009 or 2010 when the Energy Division's measurement consultants issue their reports, it is unlikely that the utilities will be able to timely book any incentives earned without having to simultaneously reserve against that amount, because of the uncertainty over whether the utilities will have to return any interim earnings or meet the final minimum performance standards because of this one factor.8
This was not our intent in establishing the earnings recovery provisions under our adopted risk/reward incentive mechanism. To the contrary, in D.07-09-043, we recognized that an effective incentive mechanism must include provisions for earnings (or penalties) at interim points during the three-year program cycle, as opposed to waiting nearly five years after portfolio implementation for any financial feedback to utility managers and investors.9 In our view, the effectiveness of our adopted incentive mechanism is seriously undermined if the utilities cannot book authorized interim earnings under that incentive mechanism, and we are persuaded by the Joint Petition that this is likely to be the case. None of the arguments that TURN, DRA or CE Council present in opposition to the Joint Petition addresses this fundamental problem to our satisfaction.
As the utilities note in their Joint Petition, their "request to eliminate some of the uncertainty of the interim payments is not intended for...the utilities benefiting from undeserved earnings."10 However, we share the concerns expressed by NRDC that a 30% hold-back provision is likely to be insufficient to ensure that the utilities are not overpaid in interim payments once final performance is trued up with ex post results. The tables that the utilities provide illustrate only up to a 30% drop in PEB from ex ante assumptions to ex post results. (See Tables 2A-2C.) As NRDC points out, this drop may not be inclusive of the possible reductions in PEB that may be seen due to decreased ex post net-to-gross ratios. From the scenario analyses that the utilities filed in their compliance filings to their 2006-2008 portfolio applications, we concur with NRDC's observation that decreases in net-to-gross ratios appear to be amplified when carried through to PEB. (See Table 3.) For instance, a 25% drop in net-to-gross ratios for the utilities' 2006-2008 portfolios results in a 33-48% drop in PEB, depending on the utility.
Moreover, the utilities' analysis of the impact of diversification on reducing the risk of total portfolio PEB reductions (Table 1) appears to be highly sensitive to how program elements are defined and the data sources used to identify them. Table 4 presents PG&E's top five elements by end-use based on reported data through 2nd Quarter 2007.11 Using this approach to defining the top five elements for PG&E, we observe a much higher level of variability in PG&E's PEB for the two scenarios prepared by the utilities, i.e., 15% and 20% reduction in savings for the top five elements. It also shows that if ex post load impact results reduces the savings from these top five elements (relative to planning assumptions) on the order of 35%, rather than the 20% level presented in the utilities' worse case scenario, then PG&E's overall portfolio PEB would be reduced by 30% or greater.
Under the utilities' scenario analysis, holding back 50% of the interim claim pending the final true-up of earnings, as NRDC strongly recommends, will eliminate potential overpayments in all of the scenarios they present for our consideration.12 With a 30% hold-back, however, there is a potential for earnings overpayment under the utilities' Scenarios A and C if the ex post level of PEB (net benefits) is less than 80% of the PEB estimated in the interim claims, depending upon the probability one assigns to that outcome. Similarly, under the illustrative scenarios that NRDC presents for our consideration in its comments (where the ex post level of PEB is on the order of 60% of ex ante PEB), a 50% hold-back eliminates the overpayments that are seen to occur if only 30% is held back.13 This is also the case under the utilities' portfolio analysis with alternate assumptions for PEB variability in all but a very "worst case" scenario illustrated in Table 2D, i.e., when one assumes that the probability of ex post portfolio PEB falling to 60% of its ex ante value is 80% or greater.
As discussed below, we will require the utilities to update the ex ante load impact data based on the Database for Energy Efficient Resources. The Joint Utilities contend in their Comments on the Proposed Decision that both increasing the hold-back to 50% and updating the ex ante energy savings estimate each year have the potential to diminish the impact of the shareholder incentive mechanism. The Joint Utilities recommend that the Commission adopt the updating of the ex ante energy savings each year, but continue the 70% interim claims/30% final claims split in the current incentive mechanism.
There is an overlap between the ex ante update and increasing the hold-back percentage. Both mechanisms are intended to balance the ratepayer interest of limiting overpayment with the utility interest in assuring revenues can be booked in a timely manner. If considered independently, these mechanisms can interfere with the balance between these interests by implementing two mechanisms separately intended to mitigate ratepayer risk.
There is no record evidence as to how to quantify the interaction between the ex ante update and the hold-back provision. It is clear from NRDC's evidence that increasing the hold-back to 50% substantially mitigates ratepayer risk. However, the ex ante update only partially mitigates ratepayer risk because there is a significant (though reduced) risk that overpayments may still occur due to incomplete ex ante updates or lack of net benefits. Therefore, we find a combination of updated ex ante values combined with a lesser increase in the hold-back that recommended by NRDC will substantially mitigate ratepayer risk brought upon by the changes we adopt to the true-up mechanism.
Based on the considerations discussed above, we will approve the modifications requested by the utilities to the MPS true-up provisions, but will require a 35% hold-back of interim payouts, rather than the 30% level adopted in D.07-09-043. This 35% hold-back represents a reasonable balancing of the concerns raised by the parties. In this way, our adopted incentive mechanism will provide the utilities with an opportunity to book meaningful earnings during the program cycle, based on verified measure installations and program costs, and at the same time will minimize the potential risk of earnings overpayment once the final ex post load impact studies are completed.
More generally, today's adopted modifications effectively make the interim claims a reward or penalty for the success or failure in implementing the energy efficiency programs and the final claim a reward or penalty for the measured load impacts resulting from the programs. These changes will mitigate the largest earnings transition in the adopted earnings mechanism at 85% of the Commission's goals, which could cause a large change in earnings for a very slight change in energy savings. For the reasons discussed above, such changes are reasonable and necessary in order to improve the effectiveness of the risk/reward incentive mechanism adopted in D.07-09-043.
Accordingly, the 1st and 2nd interim earnings claims under the risk/reward incentive mechanism adopted in D.07-09-043 will be based on verified measure installations and verified program costs, but using ex ante assumptions for load impacts, including net-to-gross ratios.14 As TURN points out in its comments, the utilities do not specify in their Joint Petition (or in their joint reply to comments) the specific ex ante assumptions for measure savings to be used in calculating the 1st and 2nd interim earnings claims. It is particularly important to clarify this issue for the 2006-2008 program cycle because the utilities' compliance filing data is too aggregated to be useful for this purpose for a significant number of programs. Moreover, consistent with agreements reached during the development of the Case Management Statement (CMS), some of the utilities have already updated their ex ante savings parameters in consultation with Energy Division since submitting their compliance filings in order to reflect more recent and realistic values for net-to-gross ratios.15 It makes no sense to undo this work by relying on earlier planning assumptions that have since been superceded by the inclusion of more realistic ex ante values in some of the utilities' E3 calculators.16
So that there is no ambiguity as to which ex ante assumptions shall be used for the purpose of calculating the 1st and 2nd interim claims under today's adopted modifications to D.07-09-043, we provide direction today for the 2006-2008 program cycle. Except as otherwise indicated below, we will use the ex ante measure savings parameters that are contained in the utilities' E3 calculators, as of their 4th quarter 2007 report for the 1st Claim and as of their 4th quarter 2008 report for the 2nd Claim. Specific direction for the 2009-2011 program cycle will be provided in our decision approving the utilities' 2009-2011 program plans.
For measures included in the Database for Energy Efficient Resources (DEER), however, we will update the values contained in the E3 calculators with the 2008 and 2009 DEER updates of ex ante measure savings parameters, including net-to-gross ratios and expected useful lives. DEER is a database developed jointly by this Commission and the California Energy Commission and funded by ratepayers. Our adopted evaluation, measurement and verification protocols require that staff undertake the updating of ex ante assumptions for measures included in DEER on a regular basis, based on the most recently completed evaluation studies. 17 Pursuant to D.06-06-063, the utilities are reporting program and portfolio accomplishments for the 2006-2008 program cycle using the 2005 DEER estimates of energy and demand reductions only "until additional [evaluation, measurement and verification] study results are available that would update those DEER values." 18
The 2008 and 2009 DEER updates are expected to be completed in early 2008 and 2009, respectively, under the direction of Energy Division. Accordingly, the 2008 DEER update will apply to the 1st Claim and the 2009 DEER update will apply to the 2nd Claim, which are currently scheduled for submission in September of 2008 and 2009 (respectively) following the release of Energy Division's verification reports in August.19
Updating measure load impacts using the DEER database prior to the payout of interim claims in 2008 and 2009 should help to mitigate the risk of extremely large swings in earnings (positive or negative) at the final earnings true-up, which serves the interests of both utility shareholders and ratepayers. Moreover, incorporating updated DEER ex ante values into these interim claim calculations will improve the consistency between the ex ante load impact assumptions we will be using to calculate earnings during 2008 and 2009 and the ex ante load impact assumptions being used to develop the 2009-2011 portfolio plans. Finally, this direction ensures that all the utilities, without further delay, will adjust their lighting savings estimates to "reflect more realistic and updated assumptions" on net-to-gross ratios, consistent with the agreements reached in the CMS.20
For customized measures or customized projects that represent aggregated measures in the E3 calculators, Energy Division will need to identify the appropriate installed measure(s) based on its measure verification results and develop the associated ex ante load impact values. For this purpose, Energy Division may use the utilities' tracking system information, engineering workpapers, DEER values and methods, or other current measurement and verification results that are available.
Finally, we put the utilities on notice that there is increasing feedback from recent ex post evaluation studies to support lowering of 2005 DEER and non-DEER ex ante assumptions for net-to-gross ratios during the forthcoming DEER updating and 2009-2011 planning process, particularly for certain lighting measures. Energy Division staff will be working closely with the assigned Commissioner and assigned ALJs to ensure that the ex ante planning assumptions used for the 2006-2008 interim claims and 2009-2011 portfolio plans are realistic and reflect the results of the most up-to-date studies available.
6 See D.94-10-059, 57 CPUC 2d, 1, pp. 43-46.
7 D.07-09-043, mimeo., pp. 121-124.
8 Joint Petition (Amended), p. 13.
9 See, for example, D.07-09-043, Conclusion of Law 7.
10 Joint Petition (Amended), p. 17.
11 This table was prepared by Energy Division at the request of the assigned Administrative Law Judge.
12 At the request of the assigned ALJ, the utilities provided the excel spreadsheets underlying Tables 2A-2C, allowing us to modify the hold-back provisions and review the results. Table 2D presents Scenario 2C (the utility scenario with the largest potential for overpayment) with a 50% hold-back, including more conservative assumptions about the 3rd Claim PEB (i.e., that it could be as low as 60% of the 2nd Claim PEB estimate).
13 See Response of NRDC to Petition for Modification of D.07-09-043, November 30, 2007, p. 8 and Tables 2 and 3, for Scenarios 1-4. We do not include the results for NRDC's Scenarios 5 and 6 because they do not reflect possible outcomes under the utilities' proposed modification to D.07-09-043. More specifically, these scenarios assume that the utilities can keep the interim payments even if the ex post results indicate that they are in the penalty range. At the request of the assigned ALJ, the utilities confirmed that this retention of earnings would not be the case under their proposal-they would be required to return all interim payments and pay applicable penalties if the final true-up (based on ex post results) indicated that portfolio performance fell into the penalty range.
14 As we directed in D.07-09-043, we will be revisiting the adopted risk/reward incentive mechanism--including the modifications we authorize today--and considering Energy Division's recommended modifications in time for the 2012-2014 program cycle. See D.07-09-043, Section 13.
15 The CMS was developed during the planning phase for 2006-2008 portfolio programs. Its purpose was to reflect discussions among the utilities, Peer Review Group members (including Energy Division staff) and interested parties that filed opening comments in Application 05-06-004 et al. Specifically the CMS was intended to (1) summarize the areas/issues in dispute based on the June 2005 filings on the utilities' proposed portfolio plans, (2) describe issues/areas where resolution had been reached based on further discussions among the CMS participants, (3) describe the extent to which cost-effectiveness issues raised by the report developed by Energy Division's consultant (TecMarket Works) had been addressed during the process, and (4) identify the remaining areas of disagreement that required Commission resolution.
16 The "E3 calculators" contain the inputs and formulas to perform cost-effectiveness calculations and report portfolio performance for the utilities' authorized energy efficiency programs. This model is named after the consultants ("Energy and Environmental Economics, or E3") that developed them for use by the utilities for this purpose.
17 ALJ's Ruling dated September 2, 2005 in R.01-08-028, p. 18.
18 D.06-06-063, p. 21; see also D.06-06-063, Finding of Fact 6; see also our direction to the utilities to present updates of ex ante load impacts for customized rebate programs based on DEER savings values for the installed measures to staff every six months. (D.06-06-063, pp. 27-28.).
19 See Attachment 6 of D.07-09-043.
20 Among other things, the CMS noted that Peer Review Group members were frustrated that the utilities used net-to-gross ratios for a variety of strategies that were outdated and inaccurate and probably to high. (CMS, p. 6.) The Peer Review Group requested that PG&E in particular reduce its reliance on lighting measures, especially residential lighting, to which PG&E responded that it would "adjust its 2006 portfolio lighting savings to reflect more realistic and updated assumptions on NTG ratios." (CMS, pp. 17-18.)
