4. Developing Cost-Benefit Models According to Perspective
The costs and benefits of any energy project may vary significantly depending on whose perspective a model reflects. For example, a model that reflects ratepayer or utility concerns will focus primarily on the cost of a project relative to other energy resource options available for purchase by the utility. A model that reflects societal concerns will likely incorporate environmental impacts and equity concerns. A model that reflects the concerns of the DG owner will emphasize project profitability and payback period. The Standard Practices Manual presents four perspectives comparable to these and identifies them as follows:
(1) The Participant Test is the measure of the quantifiable costs and benefits to the customer participating in a program. An example of a benefit is the incentive paid by the utility under the program. Since many customers do not base their decision to participate in a program entirely on quantifiable variables, this test cannot be a complete measure of the benefits and costs of a program to a customer.
(2) The Ratepayer Impact Measure (RIM) Test (previously the Non-Participant Test) measures what happens to customer bills or rates due to changes in utility revenues and operating costs caused by the program. This test indicates the direction and magnitude of the expected change in customer bills or rate levels.
(3) The Total Resource Cost (TRC) Test (and its variation, the Societal Test) measures the net costs of the program as a resource option based on the total costs of the program, including both the participants and the utility's costs. The Societal Test differs from the TRC test in that it includes the effects of externalities (e.g., environmental concerns, national security), excludes tax credit benefits, and uses a different (societal) discount rate.
(4) The Program Administrator (PA) Cost Test measures the net costs of a program as a resource option based on the costs incurred by the PA (including incentive costs) and excluding any net costs incurred by the participant. The benefits are similar to the TRC test, but costs are defined more narrowly.
Applying all four models would measure how costs and benefits are distributed among various groups or individuals.
The parties generally do not dispute the purpose of each of these models. They do, however, dispute their relative importance, how they should be applied and what the tests should measure. Each is discussed below.
4.1. Participant Test
The Participant Test measures the economic viability of a DG facility to the developer or customer installing the facility. While those who install DG will naturally have their own calculation of whether an investment is worthwhile, the Commission might want to conduct its own Participant Test to determine the level of incentive needed to promote investment and to help prevent the provision of incentive payments to "free riders." 9 As PG&E observes, it also appears that Section 2827(n) requires the Commission to complete a report on the costs and benefits of net metering from the perspective of "customer-generators." The Itron Framework identifies as benefits the customer's reduction in electricity bills, the value of displaced fuel with the use of waste heat, tax credits and other government incentives. Costs in this test include system costs, interconnection and emission control costs, and operations and maintenance expenses.
No party opposed the use of a Participant Test and we state our intent to adopt a participant cost-benefit model here and to use it to evaluate the efficacy of and need for incentives at various levels. In subsequent sections, we discuss the variables for that test that were a source of controversy in this proceeding. Attachment A lists all of the variables for the test and the source of data for each variable.
4.2. The Ratepayer Impact Measure (RIM) Test
The RIM Test measures the relative costs and benefits of a DG project or program from the standpoint of utility ratepayers. The main difference between this test and the Total Resource Cost test discussed below is that the RIM Test measures potential transfers of wealth between ratepayers and DG facilities. Thus, it measures economic benefits as well as the allocation of costs between DG developers and utility ratepayers.
The utilities advocate for the application of the RIM Test in order to evaluate the financial impact of DG projects on utility customers from incentive payments and the loss of revenue from exemptions to standby charges and nonbypassable charges. SCE observes that the RIM Test is the only test that quantifies the allocation of costs and benefits between customers who install DG and those who do not. SCE observes that this test would measure the cost to ratepayers of such subsidies as exemptions from standby charges and nonbypassable charges, reduced transmission and distribution costs, and DG incentives paid through SGIP or CSI. SCE also states this information is necessary in order for the Commission to comply with Section 353.9, which requires that net costs associated with tariff modifications provided to DG customers be recovered only from the class or classes of customers eligible to receive the tariff modification.10
Some DG proponents oppose the use of such a test, viewing it as too narrow to capture the total benefits of DG projects. CCDC does not believe a RIM Test is necessary to evaluate DG, arguing that the Commission need only apply a modified version of the Societal Test already in use for energy efficiency projects and programs. CCDC cites prior Commission orders that have found the RIM Test inappropriate as a primary test of cost-effectiveness because it only looks at a portion of total costs, its results are affected by rate-design elements, and it does not identify least-cost resource options from an economic efficiency perspective.11
Ratepayer funds support DG programs as part of our policy to promote the development of a more diverse and environmentally sound energy network in California. Among the DG efforts they support through distribution rates are discounted rates, net metering, exemptions from standby charges and the cost responsibility surcharge (CRS), and direct financial incentives offered by the SGIP and CSI. The RIM test is intended to measure whether ratepayers as a group realize a net benefit from incentives paid for DG development. Despite this fact, we note that the Commission does not currently require the RIM Test to be performed and does not rely on it in the context of cost-effectiveness evaluation of utility energy efficiency programs. Rather, the Commission uses the Total Resource Cost Test and the Program Administrator Cost Test, as discussed in D.05-04-051 and described in further detail below, when evaluating energy efficiency programs.12 Therefore, we will not require that the RIM Test be performed as part of our DG cost-effectiveness evaluation efforts. Nevertheless, we will leave discussion of the RIM Test in this order in the event the utilities wish to perform the test for rate design purposes relating to Section 353.9. In subsequent sections of this order, we discuss the variables that the RIM Test should include in the event it is used for that purpose. Where modifications to the Itron Framework approach are not explicitly addressed and adopted, the specifications in the Itron approach are implicitly adopted.
4.3. Total Resource Cost (TRC) Test and Societal Test
The TRC Test measures the relative costs and benefits of a DG project or program to both participants and non-participants, i.e., to society at large. A variant of the TRC test is the Societal Test. The purpose of both the TRC Test and the Societal Test is to determine the net benefits accruing to the subject economy or group.
The Itron Framework proposes using the Societal Test for DG cost-benefit analysis because the test's perspective is comprehensive, considers externalities which affect society as a whole, and uses a lower discount rate than used in the TRC test. The Societal Test also ignores certain tax credits which are benefits to participants, but costs to other taxpayers in the relevant area considered by the test, thereby offsetting each other. When the relevant area for program evaluation is statewide, as is the case for CSI and SGIP, the tax treatment for the TRC Test and Societal Test should be the same. That is, state tax incentives would be considered a transfer and would not be measured by the tests, while Federal incentives would be measured as a benefit.
In commenting on the Itron Framework, most parties supported using the Societal Test, although some parties suggested modifications to Itron's specific interpretation of the test or the inputs Itron used. The DG Proponents suggest the Commission use the Societal Test as the primary test of DG cost-effectiveness. SCE supports use of the Societal Test as proposed in the Itron framework, commenting that the Societal Test is "flexible enough to allow inclusion of any quantifiable cost or benefit the Commission deems necessary to include in a DG cost-benefit analysis." (SCE reply brief, at 7.)
CCDC and ASPv both support use of the Societal Test, but they propose their own modifications to the tests. CCDC proposes modifications relating primarily to air emissions to correct what it believes are underestimates of electric emissions avoided costs and to tailor DG emissions avoided costs by DG technology, time period, and location. ASPv proposes its own cost-benefit methodology, which it calls the PLEASE matrix. According to ASPv, the PLEASE matrix has its roots in the SPM but is expanded to account for the unique benefits represented by DG. ASPv contends the Itron Framework fails to include a number of DG benefits because they are considered too general or too difficult to quantify. In light of the state's support for renewable DG, ASPv explicitly advocates for erring on the side of including too many benefits rather than too few even if some of those benefits are quantified at zero for now. PG&E and SCE both oppose ASPv's PLEASE matrix, alleging it is fundamentally flawed because its elements are highly speculative and unquantifiable.
PG&E suggests the Commission use the TRC test which is used to evaluate energy efficiency programs. City of San Diego also recommends the Commission evaluate the benefits of DG consistently with energy efficiency. DRA points out that the when the Commission evaluates its energy efficiency programs, it uses a hybrid test which more closely resembles the TRC test. (DRA Comments, 2/25/09, at 20.) Moreover, the Commission has declined to use the Societal Test to evaluate energy efficiency due to its lower discount rate and treatment of certain costs as transfers. (D.05-04-051, at 82.) DRA suggests the Commission use either the TRC test or both tests.
The purpose of our inquiry here is to develop a model for DG programs and facilities that best reflects the value of DG to society and ratepayers. To achieve this goal, we will use both the TRC and the Societal variant to asses costs and benefits of DG to both participants and non-participants, i.e., to Californians at large. While the Itron Framework suggested use of only the Societal Test, we see value in performing both tests, as suggested by DRA, because each test provides a unique perspective given different accounting for federal and state tax incentives and varying discount rates. In addition, we agree with the parties that have suggested our analysis of DG should mirror the cost-effectiveness analysis we currently perform for energy efficiency programs, which uses the TRC Test.
We decline to adopt the DG Proponents recommendation to make the TRC and/or the Societal Test the primary test because we prefer to assess DG from various perspectives, and not purely a societal one. However, we will include both the TRC and the Societal Test in our cost-benefit methodology.
We find the Societal Test, as presented in the Itron Framework, and the TRC Test as described in the SPM will each provide a useful perspective in assessing the costs and benefits of DG projects and programs. We consider both tests flexible enough to incorporate the inputs that we discuss in the remainder of this order. We prefer these tests to the PLEASE matrix proposed by ASPv because we favor using the tests already described in the SPM to adopting new and unique tests for DG.
Subsequent sections of this order address each variable that presented controversy between the parties. For example, the air emissions modifications suggested by CCDC are discussed in the section on environmental values. Attachment A lists all of the variables we adopt for each model and the data source for each. While the variables may not measure costs and benefits perfectly, they are reasonable for our purposes and may be modified as better information becomes available.
4.4. Program Administrator (PA) Cost Test
The PA Cost Test measures the net costs incurred by the PA for programs such as SGIP or CSI, including incentive costs, but excluding any net costs incurred by the program participants. In this test, revenue shifts are viewed as transfer payments between participants and all ratepayers.
The benefits measured by the PA Cost Test include avoided energy supply costs, and the reduction in transmission, distribution, generation, and capacity valued at marginal costs for the periods when there is a load reduction (e.g., when a DG facility is producing energy). The costs measured by the PA Cost Test include the program costs incurred by the PA, the incentives paid to the customers, and increased supply costs for any periods in which load is increased. Administrator program costs include initial and annual costs, such as the cost of utility equipment, operation and maintenance, installation, program administration, and customer dropout costs.
PG&E objects to use of the PA Cost Test as duplicative of the RIM Test. FCE maintains the PA Cost Test is confusing and could be replaced with a simple metric such as program administration costs per kilowatt (kW) of DG energy produced. SCE and DRA support use of the PA Cost Test, favoring a multi-perspective approach to our DG cost-benefit analysis. We conclude this test may be useful as a tool to evaluate program budgets and expenditures. We have already discussed in Section 4.2 above how this test is an important element in the Commission's evaluation of energy efficiency programs and how we prefer for our DG cost-benefit analysis to largely mirror energy efficiency evaluation. We will include the PA test in our cost-benefit methodology to be consistent with our use of the test for evaluation of energy efficiency programs.
9 "Free riders" are beneficiaries of a subsidy designed to motivate certain actions who would have taken that action without the subsidy.
10 The language of Section 353.9 is as follows: "In establishing the rates required under this article, the commission shall create a firewall that segregates distribution cost recovery so that any net costs, taking into account the actual costs and benefits of distributed energy resources, proportional to each customer class, as determined by the commission, resulting from the tariff modifications granted to members of each customer class may be recovered only from that class." The original proposed decision (mailed for comment in September 2005) referred to Section 353.9 and stated the Commission could employ a cost-benefit methodology to ensure compliance with this statute. The Commission addressed the DG rates referred to in this code section in D.01-07-027 and D.03-04-060, and in Resolutions E-3777, E-3778, and E-3779. We are satisfied that any obligation under Section 353.9 has been handled by those decisions and resolutions. We do not anticipate changes to these DG rates based on this cost-benefit methodology. Any review of DG rates or tariffs, or changes thereto, will not be considered in this rulemaking but are more appropriately considered in the proceeding wherein those rates or tariffs were adopted.
11 See CCDC Comments, 2/25/09, at 6, citing D.92-02-075, at 22-24.
12 See D.05-04-051, Ordering Paragraph 5.
