The Commission established CSI funding levels for 2007 through 2016 in D.06-01-024. (D.06-01-024, p. 6.) Table 1 in that order sets the annual revenue requirements by investor-owned utility, and Table 2 indicates the portion of the total CSI budget allocated to each utility. The order provides for funding flexibility between program years because the Commission recognized that actual demand for solar incentives may vary from year to year. Further, the Commission specified that 10% of the total CSI budget should be reserved for administrative costs, including program evaluation, marketing, and outreach, 10% for assistance to low income residential customers and affordable housing projects, and up to 5% for research, development and demonstration.
The Staff Proposal recommends refinements to the funding approach adopted in D.06-01-024 in conjunction with the overall proposal to bring a performance dimension to the incentive payments and adjust incentives to account for federal tax credits. Specifically, Staff proposes the following:
· Adhere to the budget schedule established in D.06-01-024, with each utility's budget based on its prorated share of CSI collections.
· Consider dividing budgets based on customer classes or system sizes.
· Allow fund shifting in the first half of each calendar year to residential and small system applications only.
· Allow fund shifting in any direction in the second half of each calendar year.
In describing its proposal, Staff highlights the importance of preserving equity across service areas by limiting CSI funds to each utility's pro rata share of funds. The Staff Proposal specifically requested parties comment on whether and how to divide the CSI budget based on customer class or system size.
A. Parties' Positions
Several parties, including ASPv, CFC, DRA, EPUC, the Joint Solar parties, PG&E, SDREO, and TURN, support the concept of reserving portions of the total CSI budget for discrete customer classes. ASPv recommends the Commission reserve 50% of funds for residential solar incentives, and 50% for commercial incentives, while the Joint Solar parties suggest funds be reserved based on the collections from residential and non-residential customers. DRA suggests a set-aside of 30% of the annual CSI budget for residential solar rebates, corresponding to the approximate percentage of residential sales to total system sales among the electric utilities. CFC, TURN, and EPUC contend funds should be reserved based on how funds are collected from each class of customers. They are concerned with equity and want to avoid cross-subsidization, where the majority of funds are collected from residential customers but the majority of incentives are paid to non-residential customers. TURN recommends the Commission establish volume triggers for several customer segments to account for the various external factors on each customer group. EPUC suggests that customer classes should contribute to CSI based on their benefits received.
Unlike the other proposals, PG&E proposes a reservation of CSI funds based on system size rather than customer class, with 50% of funds reserved for projects under 100 kW, and 50% for projects over 100 kW. DRA opposes reserving funds based on system sizes because it fears gaming might occur. For example, applicants might size their systems solely to fall in one category under the assumption funding will be easier to obtain, and disregarding other key sizing considerations.
In contrast to the other parties, SCE and SDG&E/SoCalGas see no need to set aside funds based on customer class, although SCE would not oppose funding allocations based on system size. SCE argues the benefits of solar power accrue to all ratepayers regardless which customer installs a system, and these benefits will be the same whether the program results in fewer large installations, or many small ones. SCE also cites the administrative burden of managing separate incentive budgets.
B. Discussion
First of all, this order does not modify the adopted yearly revenue requirements by utility that were set forth in D.06-01-024, nor does the order modify the reservation of 10% of the total CSI budget for administration, 10% for low income and affordable housing solar programs, and 5% for RD&D, as set forth in that order.40
The key funding issue that needs resolution is whether we should reserve CSI funds for specific customer classes or project sizes, i.e. residential versus commercial, or projects under 100 kW versus those over 100 kW. The Staff proposed the concept of reserving funds, but did not provide specifics other than suggesting a limit on fund shifting within each calendar year, to allow small customers better access to program funds. In this order, we have determined that we will not use a calendar year basis for incentive changes, but will reduce incentives as volume triggers of program participation are reached. Thus, the Staff approach focused on calendar years of funding no longer applies. In response to parties' comments, however, we must decide whether CSI funds should be reserved based on customer class or system size.
After considering the parties' comments, we are persuaded to reserve a portion of CSI funds for residential and non-residential customers based on equity concerns and the desire to ensure all customer classes have access to CSI incentive funds. This is responsive to parties' concerns that we avoid residential ratepayers cross-subsidizing large commercial solar projects. We conclude it is better to reserve funds based on customer class distinctions rather than system sizes because this will be administratively simpler and less prone to gaming. By reserving a portion of CSI incentive funds for residential customers, customers who install small solar facilities will not have to compete for funds with large commercial customers, who have the added bonus of larger tax incentives and typically build larger solar projects. Without differentiation between residential and non-residential sectors, the CSI program could be heavily dominated by commercial rather than residential systems.
An additional reason to reserve funds for residential applicants is linked to the change in administration from the CEC to the Commission. Formerly, the CEC administered residential rebates from a single budget source, while the SGIP administrators handled medium and large solar projects through the SGIP budget. Now, residential retrofit and small commercial incentives will be administered together with non-residential incentives under Commission oversight. We do not know the future level of demand for residential retrofit solar rebates, and for this reason, we find it prudent to reserve a portion of CSI funds specifically for the residential market.
We must now decide what portion of CSI funds to reserve for residential customers. Parties suggested numerous methods, but we find the simplest and most reasonable method is to reserve one-third of total CSI funds for residential customers, and two-thirds of funds for non-residential customers, i.e. commercial and tax-exempt segments combined. We accomplish this by reserving one third of the total MWs for residential solar applicants. DRA had suggested a 30% reservation for residential customers because they represent approximately 30% of total system sales based on data from recent general rate cases. (DRA, 5/16/06, p. 5 and n. 1.) The data cited by DRA actually suggests that residential customers approximate one-third of system sales, so we will use one-third rather than 30%. This method is consistent with our pro rata allocation of CSI funding and MW among the four utilities based on their percentage of total sales. After more experience with the CSI program, we can determine whether a reservation of one-third of MWs for residential customers is reasonable. If we find that one class is achieving its MW targets and facing precipitous incentive reductions, we will reassess whether to reconsider the allocation of MW goals between the residential and non-residential sectors. If necessary, we can consider adjusting the total amount of MWs available for residential vs. non-residential customers when we review the CSI program in two years. We describe the future review of CSI more fully in Section VII.B.3 below.
Now that we have decided to allocate CSI funds between residential and non-residential customer groups, we must make another key refinement to our "trigger" process for incentive adjustments. If we reserve one-third of CSI funds for residential customers, we should also allow residential incentives to adjust based on demand in the residential solar market. This means we need to establish MW triggers not only for each investor-owned utility, but also for the residential and non-residential customer segments41 within each utility. We recognize this adds more complexity to the CSI program, but we find this complexity is necessary to ensure residential customers have access to solar incentives.
In D.06-01-024, the Commission had established a ten-year schedule for incentive reductions based on either calendar year or MW levels. We will use the same MW levels of participation for each step-down in our volume-based trigger mechanism. The tables below indicate the MW triggers for each utility, separated into residential and non-residential portions and the total allocation of MWs between residential and non-residential sectors. In the second table, we show Steps 2 through 10 only because the first 50 MW were allocated in Step 1 of the 2006 SGIP Program.
Table 11
CSI MW Targets by Utility and Customer Class42
|
|
PG&E (MW) |
SCE (MW) |
SDG&E (MW) |
So Cal Gas (MW) | ||||
Step |
MW in Step |
Res |
Non-Res |
Res |
Non-Res |
Res |
Non-Res |
Res |
Non-Res |
1 |
5043 |
-- |
-- |
-- |
-- |
-- |
-- |
-- |
-- |
2 |
70 |
10 |
21 |
8 |
16 |
3 |
6 |
2 |
4 |
3 |
100 |
15 |
29 |
11 |
23 |
4 |
9 |
3 |
6 |
4 |
130 |
19 |
38 |
15 |
30 |
6 |
11 |
4 |
8 |
5 |
170 |
25 |
50 |
19 |
39 |
7 |
15 |
5 |
10 |
6 |
230 |
33 |
68 |
26 |
52 |
10 |
20 |
7 |
14 |
7 |
300 |
44 |
88 |
34 |
68 |
13 |
26 |
9 |
18 |
8 |
400 |
58 |
118 |
45 |
91 |
17 |
35 |
12 |
24 |
9 |
500 |
73 |
147 |
56 |
114 |
21 |
44 |
15 |
30 |
10 |
650 |
94 |
192 |
73 |
148 |
28 |
57 |
19 |
39 |
Totals |
1122 |
867 |
332 |
230 | |||||
Percent |
44% |
34% |
13% |
9% |
Table 12
CSI MW Allocations by Customer Sector
Customer Sector |
MW |
Percent44 |
Residential MW |
842 |
33% |
Non-Residential MW |
1708 |
67% |
2006 SGIP Program |
50 |
|
Total MW |
2600 |
100% |
Essentially, we have taken the CSI budget allocations for each utility initially established in D.06-01-024 and used those percentages to assign each utility a pro rata portion of the total goal of 2,600 MW. Then, we have further subdivided each utility's MW goal into a residential and non-residential segment on a one-third, two-thirds basis. As an example, when PG&E conditionally reserves 10 MW of solar incentives for residential customers, its incentive level will automatically lower from Step 2 to Step 3. When that occurs, if PG&E has not yet received 21 MW of conditional reservations from the non-residential sector, then the incentive level for non-residential customers, both commercial and tax-exempt, will stay at Step 2 even if residential incentives have dropped to Step 3.
Additionally, since we changed the starting incentive level from the one originally adopted in D.06-01-024, and set a higher rate for tax-exempt customers, we must create a new schedule for how these incentives decline over the course of the CSI. The table below indicates how the incentive levels will alter as they decrease from Steps 2 through 10.
In order to develop the table below, several important assumptions were necessary. First, for the governmental/non-profit sector, we have kept the same $0.75 per watt differential relative to the other non-residential rebates that staff proposed. This difference strikes a reasonable balance between the additional benefit available to the non-residential taxable entities through the federal investment tax credit and the longer payback period that comments suggest government/non-profit customers can accept (see Section III.A of this decision for more discussion). Second, we have assumed that the composition of the non-residential installation market will be 30% governmental or non-profit, with the remainder taxable entities. Thus, overall, government/non-profit is assumed to make up 20% of the market, residential 33%, and other non-residential 47%. These assumptions may need to be revisited as we gain more experience with the market during the CSI review process described in the next section.
We also relied upon the MW amounts adopted in D.06-01-024 to determine the MW size of each step. Working from these assumptions, while staying within the overall incentive budget constraint, Staff optimized to determine the maximum incentive levels that could be paid at each step and still reach the goal of 2,600 MW. We placed several constraints on this optimization process. First, we wanted incentive drops no bigger than $0.45 and no smaller than $0.05; any larger drops would be disruptive to the market, and any smaller would not be meaningful. Second, we wanted incentive drops of no more than $0.30 in the first two steps, in order to minimize the potential disruptive impact on the market during the early phases of the program. Third, we determined $0.20 per watt to be the minimum meaningful incentive to offer during the last step to close out the program (if the incentive were any lower, the incentive payment would not make a significant contribution to customers' system costs under any scenario). Finally, since the government/non-profit sector starts with a higher incentive, we allowed a larger drop in the incentive rate for this sector in Steps 9 and 10.
Utilizing all of these assumptions, the final resulting per-watt equivalent incentive levels are shown in the table below. If assumptions prove to be invalid, review of the incentive levels may occur sooner than described below.
CSI Incentive Levels by Incentive Step
and Customer Class
Step |
MW in Step |
Gov't/ Non-Profit |
Res |
Commercial |
Total $ Disbursed in Step ($ in millions) |
1 |
5045 |
n/a |
n/a |
n/a |
n/a |
2 |
70 |
$3.25 |
$2.50 |
$2.50 |
$186 |
3 |
100 |
$2.95 |
$2.20 |
$2.20 |
$235 |
4 |
130 |
$2.65 |
$1.90 |
$1.90 |
$267 |
5 |
170 |
$2.30 |
$1.55 |
$1.55 |
$289 |
6 |
230 |
$1.85 |
$1.10 |
$1.10 |
$287 |
7 |
300 |
$1.40 |
$0.65 |
$0.65 |
$240 |
8 |
400 |
$1.10 |
$0.35 |
$0.35 |
$200 |
9 |
500 |
$0.90 |
$0.25 |
$0.25 |
$190 |
10 |
650 |
$0.70 |
$0.20 |
$0.20 |
$195 |
Total |
$2,088 |
The table indicates total CSI expenditures of approximately $2.1 billion, equivalent to the CSI Budget less administrative, marketing and outreach, evaluation and RD&D costs.
Throughout the order, we have described issues we will review after we have two years of experience with CSI. The Commission should institute periodic reviews, every two years, through the duration of the program. Thus, in 2009 or earlier, we anticipate opening a new rulemaking to review, among other issues, the following:
· Whether to continue to offer government and non-profit customers a higher incentive rate.
· Assess the need for incentive changes depending on federal tax credit status or other factors.
· Review the capacity factor used in the PBI payment calculation, based on M&E findings.
· Consider the impact of applying a PBI mechanism to all systems over 30 kW.
· Review whether it is reasonable to reserve one-third of CSI funds for residential customers.
· Evaluate and investigate a "feed-in tariff" approach.
· Consider adding trackers to the EPBB design factor.
The commissioner assigned to this future review proceeding may determine whether additional CSI program elements should be included in the scope of the review, or whether the above issues should be modified.
40 We address plans both for marketing and outreach, and for measurement and evaluation in Phase 2 of this proceeding. Until that time we direct administrators to spend no more than half of the funding reserved for administration (thus, up to 5% of total spending from the 10% reserved for all administrative components). We understand that administrative activities will be detailed in the CSI Handbook, expected to be resolved by the end of 2006. Thus we direct each of the administrators to submit estimated CSI administrative costs for 2007 and 2008 to Energy Division staff by March 31, 2007.
41 The non-residential sector includes commercial and tax-exempt customers.
42 During Phase 2, we will adopt a decision regarding the program rules for affordable housing participation in CSI. The above table now treats residential targets alike, and may be amended in Phase 2 to separate out affordable housing solar goals.
43 The first 50 MW are allocated under the 2006 SGIP program and are not pro-rated by customer class or service territory.
44 The percentages are based on one-third of 2,550 MWs because we do not include the approximately 50 MWs of solar applications received in 2006.
45 The first 50 MW are disbursed under the 2006 SGIP program at a uniform rate of $2.80 per watt.