6. Discussion

As NRDC points out, California's "one high-level, overriding goal guiding its energy efficiency efforts: to pursue all cost-effective energy efficiency opportunities."¹⁸ This overriding goal has been clearly articulated in the Public Utilities Code, in rulings and decisions by this Commission, and in the joint agencies' Energy Action Plan, which calls for conservation and energy efficiency to be first in the "loading order" of resources pursued in procurement.¹⁹ Pub. Util. Code § 701.1(b) provides that utilities should seek to exploit all cost-effective energy efficiency. Commission policies on energy efficiency articulated in D.02-10-062 and D.04-01-050, as well as the Assigned Commissioner's ruling dated July 3, 2003 in this proceeding, echo the Energy Action Plan requirement that energy efficiency be first in the loading of resources in the IOUs' procurement plans.

It is within the context of our objective to capture all cost-effective energy efficiency that we establish numerical targets for electricity and natural gas savings today, and create a process for updating them on a regular basis in the future. In order to meet our objective, the annual and cumulative numerical goals for energy savings must be aggressive, that is, they must "stretch" the capabilities and efforts of all those involved in program planning and implementation. At the same time, these stretch goals need to reflect a pace for increasing program efforts that is achievable, so that the savings goals can also be relied upon for resource planning and procurement purposes.

In our judgment, the Joint Staff final recommendations for electricity and natural gas goals achieve this balance, with certain adjustments. Based on the issues raised in the comments on the draft decision, and after further consultation with Joint Staff, we have made some modifications to the electric savings goals presented in Table 2. Specifically, we adjust the annual 2004 and 2005 GWh savings goals upwards to reflect the values adopted by the Commission for these program years, as several parties propose. ²⁰ In addition, we trend the GWh savings for the remaining years to produce cumulative savings goal for each IOU that take into account the IOU-specific estimates of maximum achievable potential presented in an expanded version of the Hewlett Foundation Report, prepared by the same authors. The expanded study disaggregates the statewide maximum achievable potential estimates presented in that report to each of the three major electric IOU service territories.

We refer to the expanded study as the "disaggregated Secret Energy Surplus Study." It was prepared for the IOUs to assist them in developing their July 2004 procurement plan filings in R.04-04-003. The disaggregated numbers were not available to Joint Staff as it prepared the March 2004 report, discussed its proposed allocation of the statewide numbers to each IOU service territory during the April 2004 workshops, and developed its final recommendations for the draft decision.

In their opening comments on the draft decision, SCE and SDG&E attach a summary of the results of the disaggregated Secret Energy Surplus Study prepared by the study authors, and compare the disaggregated numbers adopted in the draft decision to those results.²¹ As explained in that summary, the study authors went back to interim work products developed as part of the statewide analysis of savings potential. Those interim products provided energy efficiency potential estimates by utility. The authors presented those underlying 10-year estimates of technical, economic and maximum achievable potential by utility and compared them to the Joint Staff final recommendations. That comparison is shown in Attachment 8.

The Joint Staff recommendations and the disaggregated study results both sum to 30,000 GWh of statewide maximum achievable savings by the end of 2013. However, Joint Staff's approach for allocating the statewide numbers between municipal utilities and the IOUs results in IOU cumulative totals that are approximately 1000 GWh higher than the sum of the maximum achievable results for each utility in the disaggregated Secret Energy Surplus Study. Specifically, the sum of the utility-specific numbers underlying the Secret Energy Surplus Study yields a cumulative total of 25,490 GWh in maximum achievable potential for the IOUs combined, out of the statewide total of 30,000 GWh. Joint Staff's approach for allocating the statewide total to the IOUs (based on applying a baseline ratio of savings per dollar of expenditure to each IOU's relative share of program funding), yields an IOU combined cumulative total of 26,511 GWh. In addition, the Joint Staff approach for allocating the statewide savings goals to SCE and SDG&E results in goals for these two service territories that are higher than the maximum achievable potential produced under the disaggregated study. (See Attachment 8.)

We believe that using the utility-specific numbers that underlie the statewide assessment of savings potential is preferable to using the top-down allocation that Joint Staff produced in the absence of having this disaggregated data available when it prepared the March 2004 report. However, in making our adjustments to the savings goals recommended by Joint Staff and in the draft decision, we recognize that several key assumptions underlie the calculation of the maximum achievable savings potential. These include how rapidly funding levels can be ramped up, the role of emerging technologies in contributing to future energy efficiency savings, as well as the overall savings yield of program dollars (kWhs/dollar) over time. There continues to be considerable disagreement over some of these underlying assumptions, as reflected in the comments on the draft decision. For example, PG&E argues that the savings yield of program dollars will decline from current levels as energy efficiency potential becomes "mined out" over the decade. ²² Joint Staff, on the other hand, presents the view that a constant savings yield per dollar appropriately balances the range of factors that could produce either decreasing or increasing saving yields over the next decade. ²³

Today's adoption of specific savings goals is not intended to resolve these issues. Rather, the underlying assumptions that drive the development of both economic and maximum achievable potential should be thoroughly considered and addressed during the next update of potential studies. As described above, our adjustments to the draft decision are designed to reasonably bound the savings goals trajectory at either end of the forecast period, based on the best study information available to date.

In doing so, we note that attempts to use the maximum achievable savings potential presented in the disaggregated study to limit the cumulative GWh savings goals for SDG&E creates very anomalous results. This is because the disaggregated study starts with a baseline of maximum achievable savings for SDG&E that is much lower than either the Commission-adopted energy savings goals or the savings estimates presented in SDG&E's long-term resource plan (LTRP) over the first two years (2004 and 2005) of the 10-year period. ²⁴ As a result, creating a mathematical trajectory of energy savings for SDG&E that sums to the cumulative maximum achievable potential at the end of 2013 requires one to assume an unrealistic trajectory for spending and/or savings yield per dollar of expenditure over the next decade. In particular, one would need to assume dramatically declining funding levels at current savings yield ratios, or else constant funding levels at dramatically declining savings yield ratios.

Rather than force a result using unrealistic assumptions for future funding or savings yield ratios, we adopt a cumulative GWh savings goal for SDG&E that is somewhat higher than the maximum achievable potential presented in the disaggregated study for SDG&E's service territory, but that does not increase the numbers above the maximum achievable potential for all three electric IOUs combined. As a result, our adjustments result in an adopted trajectory of GWh savings goals for SDG&E that is 118% of the cumulative maximum achievable potential presented in the disaggregated Secret Energy Surplus Study, whereas the adopted GWh savings goals for PG&E and SCE are more on the order of 88% the cumulative maximum achievable potential presented in that study.

For the reasons discussed above, there is no way to equalize these percentages of maximum achievable potential across all three service territories without (1) ignoring current performance and adopted savings goals for SDG&E and substituting the much lower values of maximum achievable potential presented in the expanded Secret Energy Savings Study for 2004 and 2005, (2) requiring greatly disproportionate increases in program effort and funding for SCE and PG&E, relative to SDG&E, to bring all three IOUs to the 100% level of maximum achievable potential, and/or (3) dramatically decreasing both funding levels and savings yield ratios for SDG&E in 2006 and beyond to result in the same percentage (88%) of maximum achievable potential as PG&E and SCE. Clearly, we will need to take a fresh look at the underlying assumptions that create the disparity in the 2004/2005 savings baseline and estimated savings potential across the three service territories when we update our savings potential estimates in the future. In the meantime, we have adjusted the Joint Staff recommended trajectories of GWh energy savings over the 2004-2013 period for PG&E, SCE and SDG&E in a manner that we believe is reasonable and appropriate based on the record in this proceeding.

The peak savings (MW) goals presented in the Joint Staff recommendations and draft decision were derived as a fixed percentage (26%) of the GWh savings goals. This percentage (or "conversion factor") was selected as a reasonable mid-point estimate of the historical relationship observed between GWh and MW savings over the past 5 years. ²⁵ We are persuaded by the comments that a lower conversion factor (on the order of 20%) that reflects the average relationship between and MW savings for the 2004 and 2005 program years alone is a reasonable alternative assumption. Using this assumption in the calculations will yield cumulative peak savings values that do not exceed the cumulative total maximum achievable peak savings potential (4959MW) projected by Xenergy over the 10-year period.²⁶ We adjust the peak MW goals presented by Joint Staff, accordingly. However, as we look to develop energy efficiency programs for 2006 and beyond that more aggressively reduce peak loads, we may need to adjust the conversion factor upwards.

As PG&E and others point out, there are several different ways to define and calculate peak savings. For the purpose of establishing energy savings goals and measuring performance against those goals, we adopt the Joint Staff definition of the term, which is the average number of GWhs saved by energy efficiency measures during the summer peak period. That period is defined as the 560 hours from 12 noon to 7 pm, each weekday, from July through September. We recognize that the IOUs will need to develop a consistent method for translating average peak savings to coincident peak for resource planning purposes. For this purpose, we anticipate the need for workshops or another forum, in which the IOUs, Joint Staff and interested parties can evaluate alternative approaches for developing a consistent method. We leave it to the assigned ALJs and the assigned Commissioners in our energy resource-related proceedings to determine how and when this issue can best be addressed.

In response to comments, we also make adjustments to the Joint Staff recommendations for natural gas savings goals to reflect the 2004/2005 annual savings targets adopted by the Commission in recent program funding decisions. In addition, we adjust Joint Staff's trajectory of natural gas savings to reflect the same 10-year time period (2004-2013) that it presents on the electric side. We also adjust the savings goals for SoCalGas to better reflect its proportion of statewide natural gas sales, so that each utility's long-term (cumulative) savings goal is now roughly proportionate to its share of those sales. The effect of these adjustments is a reduction of 28 Mths (from 472 to 444 Mths) in cumulative savings goals over the 10-year period for all three natural gas utilities combined. Most of this reduction (22 Mths) is reflected in our adopted savings goals for SoCalGas. (See Attachment 9.)

In terms of how best to express energy savings goals, we agree with Joint Staff that establishing per capita usage reduction goals using future forecasts of per capita usage is problematic, since the calculation of energy savings based on such goals is particularly vulnerable to forecasting errors. We therefore prefer to express savings goals in terms of annual and cumulative GWh, peak MW, and Mth savings levels for each of the IOUs. To the extent that such goals need to be expressed in terms of per capita usage reductions, they should be described relative to a single base year of usage, as the Joint Staff proposes. Our adopted annual and cumulative savings goals are presented in Tables 1A through 1D, by IOU service territory. Table 1E presents the goals for all four IOUs, combined.

For the three electric IOUs combined, today's adopted savings goals reflect the expectation that energy efficiency efforts in their service territories should be able to capture on the order of 70% of the economic potential and 90% of the maximum achievable potential for electric energy savings over the 10-year period, based on the most up to date study of that potential. ²⁷ Table 4 presents the share of incremental needs met by electric energy efficiency programs if these long-term goals are met. As indicated in that table, energy efficiency programs are projected to meet 55% to 59% of the IOU's incremental electric (GWh) energy needs between 2004 and 2013, including those savings produced by programs funded through the $232 million PGC authorized by the Legislature. When electricity savings associated with this minimum program funding level are removed from the baseline forecast, achieving the recommended goals would enable the IOUs to meet 36%-45% of projected increases in electricity usage over the next decade with increased investment in energy efficiency.

For natural gas, our adopted savings goals are designed at this time to capture approximately 40% of the maximum achievable potential identified in the Xenergy study. This level of expectation recognizes the fact that natural gas program funding levels have dropped substantially over the last five years, and that ramping up those efforts to meet the full savings potential may take more time than on the electric side. It also recognizes some uncertainty over the level of achievable savings in the non-core sector. Nonetheless, today's adopted natural gas savings goals represent substantial "stretch goals" by anyone's standards: They reflect an increase in savings by 244Mth over the 210 Mth in savings that would be achieved if current funding levels and program effectiveness (therms per dollar) remained constant. In other words, today's adopted goals for natural gas energy efficiency represent a 116% increase in expected savings over the next decade, relative to continuation of the status quo.

In sum, we believe that today's expectations for energy efficiency savings over the next decade are appropriately aggressive and in keeping with the objectives of the Energy Action Plan. At the same time, they recognize that there may be some practical limits to effectively increasing program funding and ramping up programs to capture the full economic potential of energy efficiency at this time, particular with respect to natural gas savings.

We will use our adopted savings goals primarily on a prospective basis for resource procurement and program planning. More specifically, during each program cycle for energy efficiency, we expect the program administrator(s) (which may or may not be the IOUs²⁸) to demonstrate that their proposed level of program activities and funding is consistent with these goals. In doing so, the program administrator(s) should exclude projected savings associated with customers not included in the calculation of savings potential (e.g., resale cities and self-generation). Similarly, when documenting program accomplishments, savings by customers not included in the calculation of savings potential should be removed from the calculation of savings, in order to ensure consistency between the basis for establishing the goals and the assessment of whether those goals have been met.

We note that none of the saturation studies, lists of high efficiency measures, or estimates by service territory distinguish between consumption by low-income households and non-low income households. It is therefore reasonable to count the savings achieved from energy efficiency measures installed under the IOUs' LIEE program toward these goals, as PG&E, TURN and others recommend. Accordingly, program administrator(s) should include the reported savings from LIEE measures when reporting energy savings accomplishments. As SESCO points out, we will need to ensure that those reported savings use ex ante assumptions²⁹, such as estimated useful lives, unit savings, etc., that are consistent with ex ante assumptions we may utilize in assessing the performance of energy efficiency measures offered under the non-low income program. In a separate phase of this proceeding we will be adopting evaluation, measurement and verification (EM&V) protocols for post-2005 energy efficiency programs. As part of that process, we will work to ensure that measures installed under the LIEE program that count towards our adopted savings goals will be subject to the same EM&V protocols that apply to measures offered under the non-low income program.

In response to comments on the draft decision, we clarify that only actual installations should be counted towards these goals, and not commitments. That means, for example, that the savings reported for PY2006 will reflect measures actually installed during calendar year 2006 (January through December), regardless of whether the commitments to install those measures were made in PY2006 or in prior program year(s). This will require some changes to current reporting requirements, so that while commitments are still tracked for each program year, only the actual installations are counted toward our adopted goals.SoCalGas raises the issue of whether our adopted goals represent savings that are "gross" or "net" of free riders. Gross savings count the energy savings from installed energy efficiency measures irrespective of whether or not those savings are from free riders, i.e., those customers who would have installed the measure(s) even without the financial incentives offered under the program. Gross savings are adjusted by a net-to-gross ratio to produce net savings, that is, to remove the savings associated with free riders. It is our understanding that the savings modeled in the potentials studies are net of free riders in the near-term, but that they become equivalent to gross savings as the net-to-gross ratio approaches 1.0 over the longer-term. Hence, we clarify that the savings goals we establish today through 2008 are net of free riders. We will revisit the issue of whether the savings goals for the outer years (2009-2013) truly reflect gross savings potential when we next update our savings potential studies.

We recognize that there may need to be some differences between the near-term numerical goals and the savings levels associated with the program portfolios developed during the upcoming PY2006-PY2008 program cycle. Nonetheless, we expect the program administrator(s) to clearly describe in their filings how both electric and natural gas energy efficiency program activities and associated savings will be ramped up over time, how program savings yield ratios will be improved, or other actions will be taken to meet the longer-term numerical goals presented in Tables 1A-1D.

For this purpose, we encourage the program administrator(s) to aggressively develop program design options during the next program cycle that will address major barriers to energy efficiency deployment. We expect program administrator(s) to submit for our consideration an analysis of a wide range of promising options to remove barriers to rapid energy efficiency deployment, including on-bill financing of energy efficiency measures. In doing so, program administrator(s) should look to the practices used in other states to resolve the ratemaking, cost allocation and consumer protection issues raised by the parties in this proceeding regarding on-bill financing. As Joint Staff points out in its March 26, 2004 report, concerted efforts by program administrator(s) and the CEC to develop and support new building and appliance standards beginning in 2008 can also contribute significantly to meeting our savings goals.

PG&E, SDG&E and SoCalGas recommend that the goals we establish today be recalibrated during each energy efficiency funding cycle to take into account the existing supply portfolio "so that ratepayers do not procure redundant resources."³⁰ We disagree with the underlying premise reflected in this statement; namely, that the reasonableness of energy efficiency savings goals must be considered in the context of the IOUs' plans to dispatch existing or procure additional supply-side resources. Rather, the converse is the case, based on the policies clearly articulated in the Energy Action Plan and by this Commission. Those policies dictate that cost-effective conservation and energy efficiency are first in the IOUs resource loading order-that is, energy efficiency is evaluated for cost-effectiveness and procured before supply-side resources are to be factored into the procurement plan.

We therefore need to ensure that the energy efficiency savings goals adopted in this proceeding are fully reflected in the IOUs' resource acquisition and procurement plans so that ratepayers do not procure redundant supply-side resources over the short- or long-term. To this end, our upcoming decisions in R. 04-04-003 concerning the long-term procurement plans and 2005/2006 ongoing procurement authorizations of PG&E, SCE and SDG&E will be made in full recognition of the aggressive energy savings goals we adopt today. For the procurement plans that will be filed in 2006 and during subsequent procurement plan cycles, or for any updating to the long-term procurement plans required by the Commission before then, PG&E, SDG&E and SCE shall incorporate the most recently-adopted energy savings goals into those filings.

More generally, in any application or other filing in which PG&E, SCE, SDG&E or SoCalGas present projections of supply-side resource needs, pipeline or transmission needs, propose new facilities or otherwise utilize projections of energy demand, they must demonstrate that such filings are fully consistent with and reflect today's adopted energy savings goals, or updates to these goals as adopted by the Commission. We note that in our current natural gas rulemaking, R.04-01-025, the IOUs have submitted natural gas demand forecasts over the 2006 and 2016 period, along with information on their infrastructure requirements for meeting those forecasts. Since the gas demand forecasts for PG&E, SDG&E and SoCalGas will be reviewed in their respective Biennial Cost Allocation Proceedings (BCAPs), they also need to reflect the natural gas energy savings goals adopted in today's decision in their BCAP filings.

In addition, proposals for a risk/reward mechanism for energy efficiency should consider using the cumulative savings goal in a particular year as a threshold for performance, subject to a reasonable uncertainty band around the numerical levels. For example, if SDG&E were the program administrator and the uncertainty band is 15%, for SDG&E to qualify for earnings on its 2006 programs, it would need to show that its programs saved at least 280.5 GWh in 2006 plus or minus 15 %, or between 238 and 323 GWh per year.³¹ We will consider how best to link today's adopted savings goals with the performance basis of a risk/reward mechanism when we address proposals for such mechanisms in a later phase of this proceeding, and in the context of the portfolio of programs being implemented at that time. We will also need to consider at that time how to treat the cost of performance incentives, e.g., whether to include those costs in program cost-effectiveness calculations. We agree with PG&E that it is premature to adopt a position on this issue in today's decision.³²

With respect to updating our savings goals, we agree with Joint Staff and the workshop participants that energy savings forecasts should be updated every three years, in concert with a three-year program implementation and funding cycle ("program cycle").³³ Our use of the term program cycle in today's decision refers to both (1) how many years energy efficiency program(s) will be approved for implementation and (2) how many years of funding will be authorized for the approved program(s). For a three-year program cycle beginning in 2006, for example, the program administrator(s) would submit proposals for energy efficiency activities and funding levels for Commission approval during 2005 for PY2006-PY2008, and then again during 2008 for PY2009-PY2011, and so on. We recognize that there are evaluation and planning components for each program cycle, but we address in this decision only the timeframe for the implementation and funding years, as defined above.

Today's adopted energy savings goals will apply to the PY2006-PY2008 program cycle without further updates. In preparation for the subsequent program cycle (for PY2009-PY2011), Energy Division and CEC staff should jointly prepare recommendations for adjustments to our adopted savings goals, as appropriate, based on updated savings potential studies, accomplishment data, changes to CEC mandatory efficiency standards and other evaluation studies and factors they deem appropriate. These studies will continue to be funded out of PGC collections. The administration of savings potential and other evaluation studies, i.e., who contracts for and manages them, will be addressed in a separate decision on energy efficiency administrative structure in this proceeding.

As the IOUs and others point out, the calculation of avoided costs and the specific metrics to be used in evaluating cost-effectiveness require further consideration for resource planning purposes. We are currently addressing avoided cost issues in R.04-04-025, and the outcome of that proceeding will clearly feed into future cost-effectiveness evaluations of energy efficiency. We are also addressing the issue of what metric to adopt as the performance basis for energy efficiency resource programs in a separate phase of this proceeding, and will also be developing updated policy rules on cost-effectiveness and other issues in the coming months.

Nonetheless, the adoption of energy savings goals does not need to await the outcome of these efforts. As described in its reports, Joint Staff has taken reasonable steps to account for uncertainties in avoided cost and energy price forecasts, and to evaluate factors that could bias the analysis of statewide savings potential in either direction. We believe that Joint Staff has also taken a reasonable approach to combining cost-effectiveness metrics for this particular application. Joint Staff's screening process first eliminated all measures that did not pass the total resource cost (TRC) test. Next, Joint Staff compared those measures that did pass the TRC screening against the levelized cost of specific supply projects that can meet the same need. Finally, Joint Staff calculated the economic potential of energy efficiency based on the energy efficiency measures that passed both screenings.

Hence, SESCO's concerns that Joint Staff used levelized costs in place of the TRC and other tests of cost-effectiveness we have used for program evaluation in the past are unfounded. Nonetheless, we clarify that today's decision does not adopt Joint Staff's screening methodology for the purpose of evaluating the cost-effectiveness of individual measures or programs, or prejudge our consideration of what policy rules to adopt with respect to cost-effectiveness testing for future funding cycles.

Finally, with respect to rate impacts, we will adopt Joint Staff's recommendation that the program administrator(s) submit their estimates of the rate impacts of their proposed program portfolio in each program cycle. This showing should include a calculation of the net rate impacts, that is, taking into account the savings of the programs over the measure lives. In addition, as NRDC and others recommend, the program administrator(s) should submit estimates of bill impacts. The program administrator(s) should work with Joint Staff to develop a consistent format and input assumptions for presenting this information in their program plan applications.