According to the statute, the Commission may determine whether net surplus compensation should include either or both the value of electricity itself and the value of the renewable attributes of the electricity, "where appropriate justification exists." (Section 2827(h)(4)(A).) We discuss the value of electricity itself first.
As described in Section 3.1 above, the Commission must consider both state and federal (i.e., PURPA) requirements in adopting an NSC rate. We adopt an avoided cost derived from the DLAP, which represents short-term wholesale energy prices. This avoided cost approach reflects the incremental cost the utility avoids by receiving surplus generation from NEM customers. Such an avoided cost complies both with PURPA and with the mandate of Section 2827(h)(4)(A-B) that the adopted rate be just and reasonable, leave other ratepayers unaffected, and not shift costs between solar customer-generators and other bundled service customers.
We find PG&E's proposal to use DLAP prices is the most reasonable and efficient source for an avoided cost electricity value to include in our adopted NSC rate for several reasons. First, DLAP prices are hourly day-ahead electricity market prices that are transparent and publicly available on the CAISO website.20 DLAP prices represent the price that a utility pays for a quantity of energy sufficient to meet its day-ahead load and are the costs the utility avoids when NEM customers export excess energy between 7 a.m. and 5 p.m. We conclude other ratepayers will be unaffected if the utility compensates net surplus generation at this rolling average of the day ahead market price for power.
Second, PG&E proposes to base its rate on a simple rolling average of these hourly DLAP prices from 7 a.m. to 5 p.m. over the customer's true up period. We agree with PG&E that these hours reasonably correspond to the hours that most NEM customer-generators produce power. We find it appropriate to compensate net surplus generators at an average of the rate the utility potentially avoids by receiving the customer's net surplus generation. We conclude that basing the NSC rate on a simple rolling average of hourly DLAP prices for the 12-month period that corresponds to the NEM customer's true-up period reasonably reflects the costs the utility avoids in procuring power during the time period NEM customers are likely to produce their excess power.
Moreover, this simple rolling average over a 12-month period is a reasonable approach when we consider that the amount of net surplus energy that is likely to be compensated is quite small compared to California's total electricity load. The cost of calculating market prices with more specificity would likely outweigh the value of the program. According to PG&E, data from 2009 indicates that fewer than 10% of its NEM customers, or 2,450 customers, were net exporters of electricity and they generated a total of 5,212,073 kWhs. (PG&E, 3/15/10 at 10 and 12.) Over 40% of the net exporters on PG&E's systems had net exports of 100 kWh or less. (Id. at 10.) Data supplied by CALSEIA/EC supports the conclusion that total dollars that will be paid out for NSC is likely to be minimal. They report that in 2009, SDG&E had fewer than 1,000 customers eligible for NSC, generating on average less than 2,500 kWhs, while SCE reported fewer than 1,500 customers eligible for NSC, also generating on average less than 2,500 kWhs.21 (CALSEIA/EC, 6/21/10 at 6.) This data indicates that SDG&E and SCE net surplus generators would receive individual annual payments of less than $100, if the NSC rate is four cents/kWh.
We find PG&E's NSC proposal is administratively simple because it relies on public electricity prices and uses a simple methodology to convert those prices into an NSC rate that corresponds to the 12-month true-up period for the NEM customer. The three utilities can each use their own published DLAP prices using the same averaging technique proposed in PG&E's application to calculate the NSC rate for their customers. (See PG&E, 3/15/10, Attachment B.) The simplicity of this approach gives us an avoided cost NSC rate at a low administrative cost. We do not want the cost of implementing NSC to dwarf the compensation the customers receive under the program.
Finally, our adopted DLAP pricing approach works well with the annual netting period required by Section 2827(h)(3). DRA and Acton object to the annual average pricing approach and propose an NSC that attempts to compensate based on the time that surplus is placed on the distribution system. (Acton, 6/21/10 at 5; DRA, 6/21/10 at 6.) We conclude that tracking individual customer usage and exports more frequently than on an annual basis would drive up administrative costs and deviates too greatly from the existing NEM program and statute. The statute repeatedly refers to net surplus as calculated over a 12-month period. Within the 12-month period, customers offset their usage with their generation at the full retail electric rate. AB 920 does not change this, but merely adds a compensation requirement over and above the full retail credits. If we base NSC on a rolling average DLAP price, it allows the monthly NEM netting process at full retail rates to continue in accordance with the current NEM program.
In summary, we will direct PG&E, SCE and SDG&E to use the simple rolling average of their DLAP prices from 7 a.m. to 5 p.m., corresponding to the customer's 12-month true-up period, as the value of electricity incorporated into each utilities' NSC rate. The rolling average should be calculated on a monthly basis and be applied to all customers with a true-up period in the following month. Each utility should file a Tier 3 advice letter containing its revised NEM tariffs to implement the NSC program, the initial calculations for the DLAP-based NSC rate described in this decision, and specifics on a process for monthly updates to the rate. The NSC rate for each utility shall take effect upon Commission approval of that utility's advice letter. According to Section 2827(h)(3), customers were notified in January 2010 that they could opt to receive NSC. Therefore, the NSC rates approved through each utility's advice letter will apply to customers who chose NSC when notified in January 2010, or thereafter.
With regard to Sierra Pacific and PacifiCorp, we realize that these two utilities have very few customers that may qualify for NSC payments, and we agree that we should make administration of NSC as simple as possible given the unique characteristics of these two utilities with small territories in California. According to Sierra Pacific, it operates its own balancing authority and is not part of the CAISO, so it has no DLAP prices. Therefore, for administrative simplicity, we direct Sierra Pacific to base its NSC rate on PG&E's DLAP price. It is unclear if separate DLAP prices exist for PacifiCorp. If PacifiCorp does have DLAP prices, those prices should be used to calculate its NSC rate using the methodology described in this decision. We expect that PacifiCorp can mirror the PG&E DLAP pricing approach without undue burden. In the event DLAP prices do not exist within PacifiCorp's California territory, we direct PacifiCorp to base its NSC rate on PG&E's DLAP price. Sierra Pacific and PacifiCorp shall each file an advice letter in compliance with this order either to provide their calculations for a DLAP-based NSC rate or to notify the Commission they will use PG&E's NSC rate.
5.1. Discussion of Proposals Not Adopted
We decline to adopt other parties' proposals as set forth below.
5.1.1. Utility Proposals
Similar to PG&E's proposed electricity value, SCE's proposal involves an avoided cost. We prefer PG&E's proposal to use DLAP prices because net surplus generation will create exports from NEM customers, which will likely offset other customers' load and result in fewer purchased kWhs of load at the DLAP price. SCE's proposal is less appropriate because the MRTU generation hub price is the price paid for additional kWhs of sales to the CAISO, and NEM customer generators do not sell directly to the CAISO. (PG&E, 3/15/10 at 4, n. 1.)
Moreover, we agree with the Joint Solar Parties and Acton that SCE's method of converting bill credits to NSC using a weighted average ratio is overly complex, results in different prices for different customers, and could result in higher administrative costs. (Joint Solar Parties, 7/23/10 at 17.) As the Joint Solar Parties point out, it is problematic that under SCE's proposed NSC ratio approach, a customer's payment is determined in part by the size of any remaining bill credit. The larger the bill credit, the larger the payout, while a customer with net surplus generation but no bill credit receives no NSC payment. In addition, SCE proposes to weight average hourly MRTU prices based on a customer load profile. SCE provides only one sentence in an attachment explaining this solar profile weighting and we find this explanation insufficient.
SDG&E and PacifiCorp both propose NSC rates based on SRAC prices paid to QFs. We prefer an avoided cost approach to valuing NSC, in line with the authority granted in the recent FERC order that allows us to differentiate avoided costs on the basis of the supply characteristics of the different technologies, i.e. the unique attributes of the excess power received from net surplus generators. (See 133 FERC ¶ 61, 059 (October 21, 2010) at ¶ 23.) Although SRAC QF pricing sounds simple and straightforward, it is not. SRAC QF rates are frequently subject to litigation and adjustment in regulatory proceedings. Plus, there are many different settlements and rates for QFs, depending on whether they are renewable or non-renewable. We prefer a publicly available market-price as the avoided cost for our NSC rate. In addition, SDG&E is not simply using the QF SRAC price, but proposes to adjust that rate based on an annually determined time-of-delivery factor. For non-TOU customers, this adjustment would be based on a representative profile of excess generation derived from SDG&E load research data. (SDG&E, 7/23/10 at 4-5.) We find this adjustment to QF rates complicated and likely to make annual NSC rate updates overly contentious and resource-intensive.
We also reject the interim rate proposed by PG&E and the rate proposed by Sierra Pacific. PG&E suggests an interim NSC rate based on its system average generation rate, while Sierra Pacific proposes an NSC equal to the generation component for baseline quantities from its retail rates. PG&E acknowledges that its proposed interim rate is the generation component of retail rates and represents embedded or average, costs and not avoided or marginal costs. (PG&E, 3/15/10 at 5, n. 3.) Sierra Pacific's proposed rate is similar. Both of these rates are set in regulatory proceedings based on utility costs. We consider our approach of identifying an avoided cost, such as publicly available DLAP prices, to be superior. In our view, a utility's DLAP price more reasonably reflects the costs a utility avoids when NEM customers deliver excess generation to the grid and it is more appropriate to compensate net surplus generators using an avoided cost rather than an embedded cost.
We reject the Joint Solar Parties' suggestion to base the NSC rate on the Commission-adopted MPR, plus additional adjustments, for several reasons. First, we reject the proposal to rely on an MPR-based NSC rate because the MPR represents the cost to construct, operate and maintain a 500 MW combined cycle gas turbine, and we do not believe that surplus generation from NEM customers will result in avoided procurement from such a facility. Rather, we find that surplus generation from NEM customers is more likely to avoid short term wholesale purchases by the utilities. Thus, an MPR-based NSC rate would be inappropriate. As SCE and TURN both note, the MPR is a legislatively mandated metric intended as a cost benchmark for RPS projects with a known energy output, while NSC involves payment to NEM customers for an inherently unknown amount of net surplus generation. We agree with TURN that net surplus generation bears greater similarity to short term energy purchases by the utilities than the output of a long-term renewable generator under a power purchase agreement. As TURN notes, surplus generation cannot be forecast and only reduces real time market purchases. It does not serve as a hedge against gas price volatility so it should not be compensated as such. Since an individual NEM customer has no obligation to provide any energy to the utility, the only generation cost that the utility avoids when an NEM customer provides surplus is the reduced procurement of electricity from the CAISO wholesale market. While Joint Solar Parties contend the MPR should determine the NSC rate because it is used to pay generators under AB 1969 tariffs, we find the AB 1969 program is distinguishable from NSC because AB 1969 involves contracted power, while NSC involves payment for incidental, non-contracted power production.
Second, we reject the proposal to use the MPR to set the NSC rate because we agree with the utilities and TURN that it is not appropriate to pay net exports which can be occasional, intermittent, and unpredictable, using a cost methodology that assumes a long-term projection of costs and includes a value for capacity. As SDG&E notes, NEM customers are not under a long-term contract to provide surplus generation. Thus, the NSC program cannot be counted on for resource adequacy and the utilities do not avoid capacity costs when a customer installs net metered solar or wind generation. We do not agree with the arguments by Joint Solar Parties that prior decisions to pay QFs a capacity payment require payment for capacity to net surplus generation by NEM customers. NEM customers are required to size their systems to be no larger than onsite load and for most NEM customers, there is little or no net surplus generation over a 12-month period. In addition, as PG&E notes, AB 920 describes payment for the "value of electricity itself," implying an energy only payment. We agree with SCE that NEM customers already receive a form of compensation for capacity by having their total generation netted with their total consumption at the bundled retail electric rate. An additional payment for capacity from net exports would over-compensate them and violate the customer indifference requirement in AB 920.
Third, we reject the proposal by the Joint Solar Parties to include avoided T&D costs in an MPR-based NSC rate. We agree with the utilities that we should not include avoided T&D in the NSC rate because it has not yet been demonstrated that net surplus generation avoids these costs. We have no way of knowing if each of the net metering customers would be in congested areas or would always be in a situation where they do not need to purchase electricity from the utility. As PG&E, Sierra Pacific and SDG&E explain, NEM customers make use of the T&D system to export their generation and it is impossible to forecast when and if exports will occur. As a result, net surplus generation may not defer capital investments. Moreover, NEM customers with exports during peak months already receive compensation for exports that offset their usage at the bundled retail electric rate.
Fourth, we reject the concept of fixing the NSC rate for each customer for the life of that customer's system based on the MPR adopted in the year the system becomes operational. We agree with DRA, PacifiCorp and TURN that it is unreasonable to fix an NSC rate for the system life based on a 20-year forecasted rate such as the MPR. Further, we agree with SDG&E that basing the NSC rate on a fixed MPR value would create a mismatch between NEM credits, which are based on current retail electric rates that vary annually, and a fixed MPR-based payment for net exports. As we determine here, the NSC rate should reflect a short-term market price for electricity that represents the costs avoided by the utility over the same 12-month period in which the exports were produced.
Finally, we disagree with the Joint Solar Parties' use of an estimated solar generation profile to adjust the MPR. While we understand the Joint Solar Parties' argument that use of a generation profile is a simplifying assumption because a profile of net exports (i.e., generation minus usage) is customer specific, we find this is too great a simplifying assumption. We agree with PacifiCorp, PG&E and TURN that a profile of net exports would likely yield a different shape than a profile of generation alone.
5.1.3. Other Proposals
We decline to adopt proposals by CALSEIA/EC and Solutions for Utilities to base NSC on retail electric rates because, in our view, this would over-compensate NEM customers by paying them a rate above and beyond the value of electricity. As the utilities point out, costs associated with T&D infrastructure, billing, and other utility services are not avoided when a customer installs a generation system. If we set the NSC at the full retail electric rate which includes T&D, and other utility administrative and overhead costs, this would shift the collection of these costs to other ratepayers in violation of Section 2827(h)(4)(A) which requires that non-participating customers be indifferent to the NSC rate.
Further, we will not adopt the proposal by CALSEIA/EC to set the NSC rate equal to the feed-in tariff rate that will be adopted in R.08-08-009 pursuant to SB 32. AB 920 requires the Commission to set the NSC rate by January 1, 2011. Proceedings to set the SB 32 feed-in tariff rate are at a preliminary stage and it is unclear when the feed-in tariff would be available and whether it would even apply to PacifiCorp and Sierra Pacific. CALSEIA/EC fail to suggest what rate should apply in the interim, while there is no feed-in tariff rate in effect. Moreover, it is unclear that a rate under SB 32 to compensate for power provided pursuant to a tariff should apply for incidental and occasional net surplus power provided by NEM customers.
Moreover, we reject the adders to the feed-in tariff rate proposed by CALSEIA/EC to account for avoided emissions costs and health benefits to society. As PG&E notes, emission control costs are not avoided by the utility when NEM customers provide net surplus power. This is because any power purchased by the utility must comply with emission requirements and those costs are built into the price the utility pays, in this case, the DLAP price. If the NSC rate included emissions adders on top of avoided energy values, PG&E would be paying twice for the same costs. Moreover, we agree with Acton, TURN and PG&E that non-monetized benefits such as "health benefits for avoided state emissions" have not been recognized by the Commission to date and are too vague.
We will not adopt the proposal by Acton to set NSC based on RPS contract prices. We agree with SDG&E, PG&E and SCE that it is not reasonable to pay RPS contract prices when there is no long-term commitment from NEM customers to generate surplus power. In our view, ratepayers would not be indifferent if they were paying a premium contract price for non-contracted power. As SCE notes, NEM customer generators' systems are intended primarily to offset part or all of each NEM customer's own electrical requirements and are not dedicated generating systems built solely to serve a utility's load. We agree with SCE that the only generation cost avoided by the utility when an NEM customer delivers surplus generation is reduced procurement of electricity in the short-term wholesale market.
The proposal by the City of San Diego to equate the NSC with the price paid for utility-owned solar generating plants is also rejected. 22 SCE and TURN contend that the Commission-adopted price of utility-owned solar plants is an "all-in" price for a long-term central station generating facility with a long-term contract and it is inappropriate to compensate intermittent surplus generation at this rate. TURN adds that it is unreasonable to pay the full, levelized cost of utility-owned solar for excess NEM generation, especially when ratepayers already subsidize NEM customer-generators through California Solar Initiative (CSI) and the Self Generation Incentive Program (SGIP) incentives and by paying NEM bill credits at the full retail electric rate. We agree.
Finally, we reject the proposal by DRA to base the NSC on DG avoided cost calculations. While the Commission adopted a methodology to analyze DG costs and benefits in D.09-08-026, and this methodology includes DG avoided cost estimates, certain components of the methodology were not finalized in D.09-08-026, and work continues to address these components. This proceeding is not an appropriate venue to examine these debated components and settle long-standing disputes over DG avoided cost calculations. To undertake this effort would inappropriately delay a final NSC rate beyond the statutory deadline in Section 2827 and require an interim rate.
20 PG&E's DLAP prices can be found at http://www.caiso.com by clicking on the link to "OASIS," then clicking on "Prices" and choosing the "Locational Marginal Prices" report from the Report dropdown menu.
21 Assuming an NSC rate of four cents/kWh (based on the 2009 average) and annual net surplus generation for PG&E, SCE, and SDG&E of 12 million kWh, total annual net surplus compensation would equal $480,000.
22 See D.09-06-049 (SCE), D.10-04-052 (PG&E), and D.10-09-016 (SDG&E) where the Commission adopted prices for utility-owned solar plants.