3. Interruptible Programs

We address the four core areas and issues in the same order as identified in the September 21, 2001 Scoping Memo and Ruling. The first core area involves modifications to existing interruptible programs. We began each section by stating the issue.

3.1. Extend Rolling Blackout Reduction Program

Issue: Should any program scheduled to terminate before December 31, 2001 be extended from its scheduled termination date to December 31, 2002.

The only program due to expire before December 31, 2002 is SDG&E's Rolling Blackout Reduction Program (RBRP). The RBRP permits SDG&E to call on customer-owned emergency backup generators (BUGs) during a CAISO-declared Stage 3 event to reduce demand that must otherwise be met by system resources. The program was authorized for one year, at SDG&E's request, and will expire on May 31, 2002. (D.01-06-009.)

SDG&E proposes an extension through at least December 31, 2002. We adopt SDG&E's recommendation, but extend the program through completion of SDG&E's next rate design proceeding (with completion expected by April 2004). Extension through at least the end of 2003 is recommended by CMTA and CIU for nearly all programs, and is consistent with our extension of all programs below.

SDG&E states that the program has 35 customers from which a load reduction of 73.56 MW could potentially be realized during a Stage 3 emergency. (Joint Comments, October 12, 2002, page 2.) The program has been well received by customers. In fact, SDG&E has no other operational interruptible program that reaches this level of participation and amount of interruptible load.

No Stage 3 events have been called since RBRP was approved. Thus, there is no operating experience to show a need for any program revisions. We authorized a program in June 2002 which we believed reasonable and workable, and believe RBRP as authorized is still such a program. We are persuaded by SDG&E that this program merits continuation.

3.1.1. Program Type, Cost, Pollution, and "Free Riders"

TURN objects to continuation of the RBRP in it current form. TURN argues that the RBRP is not a demand reduction program, is expensive, causes pollution, and includes a high percentage of "free-riders" (i.e., participating customers who had already installed backup generation to meet their own needs). We are not inclined to make changes based on further argument of points that were so recently addressed in D.01-06-009.

For example, in relation to denying capacity payments we said that the RBRP is a demand reduction program, not a program for providing generation capacity to the grid. (D.01-06-009, mimeo., page 9.) TURN disagrees, and asserts that the program is being offered under the guise of a load reduction program when it is actually a supply-side generation program. TURN presents no new or compelling evidence or argument which convinces us to modify our view.

TURN asserts that the RBRP energy payment of $0.20/kWh is excessive. TURN proposes, however, that all programs (with the possible exception of OBMC) be combined into one new commercial/industrial program beginning in 2003. TURN's new program stresses energy payments (to focus on pay-for-performance results) rather than capacity or reservation payments (which are paid whether or not the customer is asked to perform). TURN proposes energy rates of up to $1.00/kWh for the first 20 hours, and up to $0.50/kWh for the next 130 hours, along with a capacity payment of $20/kW-year. TURN does not convincingly show, however, why the existing RBRP energy payment of $.20/kWh is excessive compared to the prices in its proposed replacement program.⁶

We agree with TURN that price is an important matter. As a result, when authorizing RBRP we declined to adopt either the proposed capacity payment or proposed interconnection charge, and substantially reduced the proposed energy payment. We considered several competing factors in reaching that decision and are not persuaded by TURN to revisit that issue now. If the price at any time is either too high or too low, a rapid adjustment mechanism is available. (D.01-06-009, mimeo., page 11.)

We agree with TURN that there are environmental concerns. We addressed those concerns by requiring customers to be responsible for compliance with all federal, state and local laws and regulations, including those regarding air quality. (D.01-06-009, mimeo., page 5.) Further, we conditioned approval of Schedule RBRP on environmental dispatch. (D.01-06-009, mimeo., page 12.) We are not persuaded by TURN that further mitigation is necessary or that this concern justifies program termination on May 31, 2002.

TURN correctly asserts that we primarily want customers to participate in existing programs (e.g., BIP, OBMC), and that we said continuation of RBRP beyond one year should be based on assessment of experience in Summer 2001. TURN states that experience in 2001 clearly demonstrates that RBRP is not needed in the future. We disagree. The experience in Summer 2001, and the lack of operation of RBRP, do not by themselves support continuation or cancellation of this program. We believe that the general need for interruptible programs, such as RBRP, has not ended and that it is reasonable to continue RBRP along with other programs.

Finally, we considered the "free rider" problem in our rejection of the proposed capacity payment. (D.01-06-009, mimeo., page 9.) We authorized payment for the costs of incremental operation, but not fixed capacity costs, in part to address this concern. We declined to provide either a capacity payment or interconnection payment because we did not intend to use the program to facilitate installation of new BUGs. We noted that customers already have incentives to install BUGs. (D.01-06-009, mimeo., page 9.) We concluded that there is little or no overlap between BUGs used to meet the customer's own emergency needs and those that will be used in the RBRP to maintain system reliability. (D.01-06-009, mimeo., page 14.) TURN fails to present any compelling evidence or argument that justifies revisiting that issue now.

3.1.2. Relation to OBMC

ORA proposes that all existing programs be consolidated into one or two programs by 2003. For example, ORA suggests that the RBRP be consolidated with OBMC, and that circuit exemption from rotating outages rather than cash be the incentive for participation.

We generally decline to consolidate existing programs (as discussed more below). We also note that different customers respond to different incentives. Cash and outage exemption are not interchangeable benefits to all customers. It is reasonable to have both, depending upon the program to which the customer subscribes.

Further, according to SDG&E, many RBRP participants are small, and share a circuit with many other customers. Small RBRP participants are not eligible for OBMC, as now designed, without several customers presenting a joint circuit plan. The benefit of RBRP may be lost, in part or whole, without this coordination, but neither ORA nor any other party present a compelling proposal that would ensure the same benefits are secured after program redesign.

Thus, we are persuaded to continue the program without change through conclusion of the next rate design proceeding, as we do with all other programs below.

3.2. Duration of All Programs

Issue: Should programs scheduled to terminate on December 31, 2002 be extended, and, if so, should megawatt and total program dollar limits adopted in D.01-04-006 be modified.

At the commencement of this proceeding, all interruptible programs were scheduled to expire on March 31, 2002. We agreed with the majority of Phase 1 parties, however, that the need for these programs was unlikely to end by March 31, 2002. We stated that we could not extend these programs indefinitely, but decided to extend the expiration date to December 31, 2002, with both capacity and expenditure limits. (D.01-04-006, mimeo., pages 20, 78-81.) We resolved to again consider extensions, program redesign, and program limits in Phase 2, as necessary.

Respondent utilities state that there is insufficient information at this time to assess the value of interruptible programs for 2003 and beyond. Utilities propose that each utility be instructed to prepare a report in August 2002 evaluating customer participation in existing programs, stating an estimate of costs, and recommending ratemaking treatment. They suggest that parties be given an opportunity to comment on each report, and that they will simultaneously submit advice letters to continue programs they believe to be necessary. If the Commission disagrees with utilities' assessments, utilities recommend that there be an informal effort to reach resolution by September 30, 2002 and, absent resolution by that date, that the Commission open a limited proceeding to review issues on program continuation.

We decline to adopt this approach. Reports submitted in August 2002 cannot contain much data, if any, regarding actual Summer 2002 experience. This is the case whether they are submitted at the end of the month, or at the beginning, as CLECA recommends.

Moreover, respondent utilities' proposal lacks adequate information about the necessary schedule (e.g., amount of time needed to pursue informal resolution, to initiate and conduct a limited proceeding, to inform customers of the results, to permit subscription to new programs). We agree with CMTA that a "'wait and see' approach is simply not compatible with the need for forward planning by most businesses," and that waiting until the third quarter of 2002 to make a decision on these programs jeopardizes participation, even if programs are extended. (CMTA Reply Comments, November 16, 2001, page 2.)

We seek an approach that permits parties to provide better information and recommendations, and that allows adequate time for the Commission to make informed, reasonable decisions. Moreover, now is the time to consider an approach that allows us to once again integrate interruptible programs with the comprehensive review of rates and rate design that occurs in each utility's general rate proceeding.

3.2.1. Extend to Next General Rate Case or Similar
Proceeding

CMTA and CIU recommend extension of programs through at least December 31, 2003, with capacity and dollar limits modified consistent with expected conditions. In particular, CMTA requests that the Commission provide some badly needed certainty by promptly extending existing programs through 2003.

We agree, and extend programs through the date of the final decision in the rate design phase of each utility's next GRC or similar proceeding. This is an extension through 2003 or early 2004.⁷

We do this because electricity supply and demand issues are sufficiently unpredictable that an extension of interruptible programs, with updated limits, is reasonable. We expect estimation of supply and demand to become somewhat more predictable and stable when adequate new supply is added to California's resource base, the role of multiple state agencies and other entities is clarified, the remarkable conservation achieved in 2001 is or is not secured for the long term, and the profoundly dysfunctional electricity market is permanently reformed. Extension through the rate design decision in each utility's next GRC or similar proceeding provides time for some, if not all, of these events to unfold. It also permits examination of interruptible rates and rate design in the context of each utility's overall rates and rate design.

We agree with CLECA that there has not been a large subscription to the new programs we authorized in 2001. This is at least in part because it takes time to market programs, and for customers to make informed decisions. A reasonable extension will provide an opportunity to pursue marketing of stable programs for Summers 2002 and 2003.

Interruptible programs serve as a type of insurance policy against uncertainty. They function to provide statewide grid reliability, and reduce the probability of experiencing rotating outages or catastrophic system collapse. Some level of interruptible programs will probably always be desirable, as long as prices are reasonable for customers and ratepayers. As TURN says, "interruptible programs are insurance policies that need to match insurance premium payments to the value of the item being insured." (TURN Reply Comments, November 16, 2001, page 2.) We will have the opportunity in each GRC or similar proceeding to assess need, program design, rate design, rate levels and other factors.

3.2.2. Modified Capacity and Dollar Limits

We previously authorized up to 5,000 MW and $500 million per year for costs related to interruptible programs and curtailment priorities. We stated that we may reduce capacity and dollar limits going forward based on monthly reports or other information. (D.01-04-006, mimeo., page 81.)

The monthly reports filed for results through December 31, 2001 show total subscribed interruptible load of about 1,420 MW (at the 5% level for OBMC, and minimum Demand Bidding Program (DBP) response). CEC recommends a planning goal of 2,500 MW for demand responsiveness programs in 2002. We adopt this recommendation, and set a goal of 2,500 MW through the next GRC or similar proceeding. We reduce capacity and dollar limits accordingly, in the same proportion as previously authorized. The result is:

INTERRUPTIBLE PROGRAM

AND CURTAILMENT PRIORITY LIMITS

UNTIL EACH UTILITY'S NEXT GRC

OR SIMILAR PROCEEDING

UTILITY	INTERRUPTIBLE PROGRAM LIMIT (MW)	TOTAL ANNUAL PROGRAM DOLLAR LIMIT ($ MILLION)
PG&E	1,000	100.0
SCE	1,375	137.5
SDG&E	125	12.5
TOTAL	2,500	250.0

The monthly reports show no expenditures for curtailment priority limits, with the exception of about $1.5 million for SCE in 2002. This is a sufficiently small component that we do not separately adjust the SCE total. As already authorized, a respondent utility may file and serve an application, as needed, to adjust these capacity and dollar limits.⁸ (D.01-04-006, mimeo., page 80, and Ordering Paragraph 17, as renumbered by D.01-04-009.)

The planning goal of 2,500 MW is reasonable given that current interruptible programs provide about 1,500 MW. The goal allows for successful program marketing to substantially increase existing demand responsiveness capacity by about 1,000 MW. Moreover, according to CEC, capacity of 2,500 MW will provide California system operators with about 5% of Summer 2002 projected load (44,000 to 47,000 MW) to be immediately responsive to necessary system conditions. No party disputes the reasonableness of having 5% of demand responsive to system conditions.

We clarify that we do not view this 5% planning goal as a system resource to be dispatched by system operators, but as a margin against uncertainty and risk in Stage 2 and 3 emergencies. The majority of customers have told us that they do not want to be considered part of California's electricity resource base. (D.01-04-006, mimeo., page 30.) They do not expect to be routinely interrupted. Rather, their business is conducting their business, not buying and selling electricity, nor constantly monitoring the electricity market to make decisions about curtailing operations. An interruptible program as a system resource is generally a second best solution. The first best solution is adequate supply of safe and reliable electricity at just and reasonable rates.

Parties may wish to propose rates for interruptible programs in upcoming rate design proceedings that differentiate payment based on customer role. Payment may differ depending upon whether the customer provides insurance against occasional risks, or is a resource upon which California may call each and every year to meet demand (e.g., equivalent to a peaking resource). Payment for insurance may consider the likelihood of that insurance being used (i.e., expected experience). Payment to a customer providing interruptible load that is equivalent to a system resource might reflect the duration of the commitment (e.g., one year or 10 years), similar to payments to qualifying facilities.

3.2.3. Utility Reports

We decline to direct that utilities file a special report in August 2002. Rather, we direct that utilities continue to file and serve regular reports monthly. (D.01-04-006, Ordering Paragraph 4.) These reports will continue to inform the Commission and parties about program progress and costs. They will also provide useful information to parties as we transition ongoing assessment of interruptible programs and tariffs to GRCs or similar proceedings, and may assist with the framing of data requests in those proceedings.

The reports, however, will not be necessary after conclusion of those proceedings. Unless directed otherwise by subsequent order, each utility may terminate the filing and service of its monthly reports effective the date that the final decision is mailed in the next GRC or similar proceeding that addresses interruptible tariffs.

Further, we order utilities to meet with Energy Division staff to improve these monthly reports going forward, to the extent that existing reports fail to adequately provide necessary information consistent with our orders. For example, each report must state megawatts subscribed, program costs, and program revenues, among other things. PG&E and SDG&E should include an estimate of program revenues from existing rates, as does SCE. SCE and SDG&E should include a table with megawatts subscribed, as does PG&E. SDG&E should include program cost information in table form, as do PG&E and SCE. Utilities should agree with ED staff on a common report and table format.

We also directed that utilities report any other relevant information the Commission should know to be reasonably informed. This should include the amount in authorized memorandum accounts. Current monthly reports do not include this information. We understand this to mean the balances are zero. Utilities should, going forward, specifically report a zero, or other, balance in these memorandum accounts.

The reports should also specifically state the total subscribed megawatts and incurred annual costs compared to the megawatt and dollar limits. Further, the reports should contain information, when relevant, on energy supply and other information utilities otherwise proposed to include in their August 2002 report. Utilities should agree with Energy Division on the subjects and timing of this additional information.

3.2.4. Program Consolidation

TURN recommends that all current programs end by December 31, 2002. TURN proposes that after 2002 all interruptible programs be combined into a single commercial/industrial interruptible program that relies heavily on a "pay for performance" incentive. According to TURN, that incentive would be accomplished by a relatively large energy payment, and a minimal capacity payment. TURN suggests not deferring consideration of program structure until utilities file reports (e.g., in August 2002 and after), but that the Commission decide now that any future program structure will be based on a pay for performance model.

TURN makes a reasonably strong case for its program structure. We are persuaded by CEC, CMTA, and others, however, that the most reasonable approach is to offer a portfolio of options, including various pricing structures. As CEC says, "most customers are not in favor of a `one demand responsiveness program fits all' philosophy." (CEC Proposals, November 9, 2002, page 13.)

ORA also recommends consolidating programs into one or two going forward, seeking to promote customer understanding, increase participation, and reduce administrative burden. We are not persuaded by ORA that program consolidation would necessarily result in those benefits any more than would a portfolio of options.

Nonetheless, TURN, ORA and others may propose alternative program structures as appropriate in each utility's GRC or similar proceeding. We will not prejudge those outcomes in this decision, but will reach decisions in those proceedings based on the best information presented at that time.

3.2.5. Cost Benefit Analysis

ORA also recommends that utilities be ordered to submit cost-effectiveness analyses in future reports (e.g., the August 2002 suggested by utilities). ORA suggests that utilities use the Commission's Demand Side Management (DSM) Standard Practice Manual for Cost Benefit Analysis, often used to analyze energy efficiency programs. We decline to adopt this recommendation.

The DSM Manual is very useful for its intended purpose, but appears not as useful here. PG&E alleges, for example, refinements to the DSM Manual would be necessary to reasonably evaluate the efficacy of an air conditioner cycling program with 100% versus 50% cycling capability, or the value of a curtailment program with daily, weekly, monthly or annual limitations.

Each party must support proposals it makes in GRCs and other proceedings regarding interruptible programs. We encourage each party to employ the best cost-effectiveness analysis and tools available, but will not specify a single approach.

3.3. Bill Limiter

Issue: Should the bill limiter provision currently reflected in the interruptible program tariffs of SCE terminate on March 31, 2002.

3.3.1. Background

Bill limiters for SCE Schedule I-3 and I-5 interruptible customers were first adopted in SCE's 1992 GRC decision. (D.92-06-020, 44 CPUC2d 471, 528.) The purpose was to mitigate the bill impact of transferring Schedule I-3 and I-5 customers of record on December 31, 1992 to Schedule I-6 on January 1, 1993, given the lower level of interruptible credit in Schedule I-6. According to SCE, the bill limiter capped these customers' bills to a total of no more than 15% in 1993, and 30% in 1994, above what would have otherwise been their I-3 or 1-5 bills based on December 1992 rates.

Legislation adopted in 1993 prohibited reductions in interruptible credit levels during 1995 and 1996. (Public Utilities Code Section 743.1.) Public Utilities Code Section 743.1 was amended in 1994 to extend the prohibition through 1999. It was amended again, in 1996, to continue the prohibition against reductions through March 31, 2002.⁹

According to parties, there are approximately 100 customers representing about 200 MW of load subject to the bill limiter. SCE states that in its 1995 GRC, the bill limiter reduced revenues from eligible customers by about $25 million annually, and rates from all other large power customers (Schedules TOU-8 and I-6) were increased by an equivalent amount. SCE says the annual revenue deficiency for 2002 is about $54 million, with about $35 million for the nine-month period from April 1, 2002 through December 31, 2002.

3.3.2. Termination

SCE seeks clarification of whether or not the bill limiter provision expires on March 31, 2002, or December 31, 2002 (the sunset date for Schedule I-6, as extended by D.01-04-006). We direct that the bill limiter expire on the effective date of this order.

3.3.2.1. Public Utilities Code Section 368(a)

CIU asserts that D.01-04-006 extends all interruptible programs, and all components of those programs including the bill limiter, through December 31, 2002. This is incorrect.

The bill limiter provision is Special Condition 14 of Schedule I-6, which states in relevant part: "This Special Condition expires on January 1, 1999." A footnote further explains: "This scheduled change [i.e., expiration on January 1, 1999] is suspended due to the rate freeze mandated by Assembly Bill 1890 and implemented through Public Utilities Code Section 368(a)."

Public Utilities Code Section 368(a) provides that rates shall remain at certain levels until the earlier of March 31, 2002, or the date on which certain Commission-authorized costs are fully recovered. Under current conditions, the earlier date will be March 31, 2002.

Nothing about our order in April 2001 (D.01-04-006) extending interruptible programs in general, or I-6 specifically, disturbed Public Utilities Code Section 368(a), or its application to I-6. The I-6 tariff filed by SCE pursuant to D.01-04-006 contained the bill limiter through the date of the rate freeze implemented by Public Utilities Code Section 368(a). No protests were filed on this point. Those tariffs became effective five days after filing, unless suspended by the Energy Division Director. (D.01-04-006, Ordering Paragraph 1.) The Energy Division Director did not suspend the tariffs for non-compliance on this point. The I-6 tariff became effective with the bill limiter provision expiring pursuant to Public Utilities Code Section 368(a). This is fully consistent with D.01-04-006.

3.3.2.2. Bill Limiter Has Served Its Purpose

Through March 31, 2002, the bill limiter will have been in effect for nine years and three months. This is enough time to have served the purpose of mitigating bill impacts caused by transferring Schedule I-3 and I-5 customers to Schedule I-6. In fact, assuming an average savings by eligible customers of $25 million per year for 9.25 years (through March 31, 2002, based on 1995 GRC estimates), these customers have enjoyed reduced rates of $231.25 million. If there have been approximately 100 customers over this period, each customer has enjoyed an average of about $2.3 million in reduced rates.

As CLECA and others point out, rates for customers have increased over time, even with the bill limiter. Nonetheless, the bill-limited total charges have been reduced compared to what they would otherwise have been. This has eased the transition, and met the legislative goal.

3.3.2.3. Bill Limiter and SCE/CPUC Settlement Agreement

CIU states that permitting the bill limiter to continue through March 31, 2002, but expire December 31, 2002, would be consistent with the approach taken in the recent Settlement Agreement between SCE and the Commission, approved by U.S. District Court Judge Ronald S.W. Lew. (United States District Court, Central District of California, Western Division, Case No. 00-12056-RSWL(Mcx).) In support, CIU says two primary purposes of the Settlement Agreement are to: (1) avoid instability and uncertainty for ratepayers, the State of California and SCE, and (2) protect customers from the potential impact of further volatility in electricity prices. (CIU Proposals October 12, 2001, page 6, footnote 3, citing Settlement Agreement, Recital F.) CIU concludes that extending the bill limiter is consistent with these purposes. We disagree.

We do not accept that ending the bill limiter causes instability, uncertainty and volatility in electricity prices. The bill limiter has consistently been subject to termination, beginning in 1995, then deferred to 1996, to 1999, and finally to on or about March 31, 2002. This information has always been available to customers. Finally implementing the planned termination does not cause instability, uncertainty or volatility. Moreover, CIU does not convincingly explain how deferring the end of the bill limiter by about nine months makes any material difference in this result.

CIU also asserts that the Settlement Agreement says "continuation of retail rates that produce revenues in excess of SCE's current costs [meets certain goals]...without further retail rate increases." (CIU, Proposals October 12, 2001, page 6, footnote 3, citing Settlement Agreement, Recital E.) According to CIU, extension of the bill limiter end date would be consistent with this goal.

We do not accept that ending the bill limiter is a rate increase. Rather, it implements a transition within one rate schedule that has been known for a long time. No rates are increased, only a transition occurs. The goal has always been to transition Schedule I-3 and I-5 customers to I-6. A transition period of more than nine years is sufficient.

Moreover, even if the transition is incorrectly viewed as a rate increase, we note that the Settlement Agreement specifically prevents rate decreases. It does not, however, prevent rate increases. (Settlement Agreement, Section 2.2(a).)

Finally, CIU claims SCE represents the parties' intent to be that: (1) rates not be increased, (2) the Settlement Agreement not change any rate, and (3) the Settlement Agreement expressly preserves the rates SCE is already charging. (CIU Proposals October 12, 2001, page 6, footnote 6, citing SCE' Reply Brief in Support of Entry of Stipulated Judgment at page 10.) The Settlement Agreement does not increase or change any rate, and it preserves the rates SCE is charging. In particular, the Settlement Agreement does not increase or change any Schedule I-6 rate, including the bill limiter. All terms and conditions of Schedule I-6 are preserved with the Settlement Agreement.

3.3.3. Limited Opt-Out

The Schedule I-6 tariff adopted and approved pursuant to D.01-04-006 leaves no doubt the bill limiter expires on or about March 31, 2002. The Settlement Agreement does not in any way disturb this conclusion. As a result, we direct that the bill limiter end on the effective date of this order.

Nonetheless, assuming some customers may have been confused, we permit SCE to offer a 15-day opt-out period for customers subject to the bill limiter. This opt-out period should begin within 30 days of the date of this order. These customers may opt-out of their interruptible tariff effective the date the bill limiter ends, or effective with the beginning of their next billing period, similar to the opt-out authorized in D.01-04-006. They may also opt-out, as may any interruptible customer, during the next annual opt-out, in November 2002.

SCE, CIU, CLECA, and others, are concerned that termination of the bill limiter on or about March 31, 2002 may cause customers otherwise subject to the bill limiter to convert to firm service, thereby causing the system to lose up to 200 MW of interruptible load. We are comfortable with these customers making this choice based on their needs, ability to continue to be interrupted, and rate levels. Existing interruptible discounts are still attractive compared to firm service rates. We seek to have a base of interruptible load available for Summer 2002 upon which system operators can reasonably rely. This opt-out opportunity will allow these few customers to re-evaluate their situation. A bill-limited customer who elects to opt-out may enroll in any other program (e.g., BIP) on a current and going-forward basis without restriction.

3.4. Aggregation of More Than Two Circuits for
OBMC

Issue: Is it necessary or feasible to develop a tariff option for aggregation of more than two circuits with a single lead customer for the purpose of participation in the OBMC program (D.01-06-087, Ordering Paragraph 3).

On June 19, 2001, Wolfsen, Inc. (Wolfsen) petitioned for modification of the OBMC program. Wolfsen proposed allowing a single customer to aggregate its load on up to 15 circuits for purposes of OBMC participation. We granted the petition in part, by permitting aggregation of load on two circuits, and directed respondent utilities to hold a workshop to develop a more complex OBMC circuit aggregation program for Commission consideration through a petition for modification. (D.01-06-087, page 8 and Ordering Paragraph 3.) Respondent utilities suggested that this matter be included in the Phase 2 workshops. The Assigned Commissioner agreed. (Scoping Memo, page 2.)

No party now affirmatively advocates allowing a single customer to aggregate its load on more than two circuits for participation in OBMC. Respondent utilities state that they do not believe further aggregation should be incorporated into the OBMC program without a showing of significant benefit to the overall system, and to ratepayers at large. We remain concerned that aggregation of more than two circuits could pose administrative and tracking problems. (D.01-06-087, mimeo., page 7.) Therefore, we decline to direct development of a tariff option for aggregation of more than two circuits with a single lead customer for the purpose of participation in the OBMC program.

3.5. Alternate Workweeks

Issue: Should the 10-day baseline for purposes of participation in the OBMC program recognize alternate workweeks, as proposed by Cal Steel (D.01-06-087, page 14).

On June 21, 2001, respondent utilities jointly petitioned for modification of the OBMC program, including changes to the calculation of the 10-day baseline. In response, Cal Steel proposed that similar days for purposes of the baseline be grouped as (1) weekends and holidays, (2) mid-week full operation days, and (3) mid-week scheduled reduced operation days.

We granted respondent utilities' petition with slight change. We declined to adopt the recommendation of Cal Steel, but invited parties to revisit the issue in a workshop. (D.01-06-087, mimeo., pages 11-15.) At utilities' suggestion to promote efficiency, the Assigned Commissioner included the issue in Phase 2 workshops, and as an item in the list of Phase 2 issues. (Scoping Memo, page 2.)

Cal Steel offered nothing in Phase 2 to support its proposal. Respondent utilities do not support Cal Steel's recommendation. Neither CMTA, nor any other party, proposes separating weekday baseline measurement into mid-week full operation days and mid-week scheduled reduced operation days.

Therefore, absent need for this change demonstrated and supported through a specific proposal, we decline to consider the Cal Steel alternative workweek proposal further. We consider other proposed changes to OBMC below, including various modifications to baseline measurement.

3.6. Other Modifications

Issue: Should other modifications and consolidations be adopted.

Parties propose several other modifications and consolidations of interruptible programs. In the following sections, we evaluate: (1) modification of SLRP, (2) modifications of OBMC, (3) interruptible program proposals suggested by the CCPCFA, (4) SDG&E's EAEI program, (5) interruptible program proposals made by the CEC, and (6) SCE's petition for modification of D.01-04-006 regarding changes to the FSLs of existing interruptible customers.

3.6.1. SLRP

PG&E and SDG&E propose a slight modification to the SLRP tariff regarding non-compliance. The current tariff provides in part that:

"...the energy usage during the on-peak period for the four weekdays following a curtailment, unaffected by program operations and excluding holidays, will be evaluated and cannot exceed the customer's posted baseline amount."

PG&E and SDG&E propose that it be revised to state (the change is underlined):

"...the energy usage during the on-peak period for the four weekdays following a curtailment, unaffected by program operations and excluding holidays, will be evaluated and cannot exceed the customer's posted baseline amount by more than 15%."

According to PG&E and SDG&E, the proposed modification relates to load shifting for customers that do not have 12 months of interval data, or for customers whose current year's consumption varies by more than 5% from the previous year's same month consumption. They assert that this modification is necessary to keep non-compliance rules consistent for those with and without interval meter data history.

No party opposes this modification. The proposal is in the public interest to maintain consistency, and is adopted.

3.6.2. OBMC

Several changes to OBMC are proposed. We generally decline to adopt these proposals. OBMC was recently modified to provide significant additional flexibility. (D.01-06-087, mimeo., pages 11-15.) These modifications have not been tested, given lack of OBMC implementation since they were adopted. With one exception (regarding monthly requirements discussed below), we agree with SCE that no further changes in baseline calculation should be adopted unless and until the current revised baseline methodology has been tested and found deficient.

3.6.2.1. Similar Days

The 10-day baseline is now measured by using the immediate past 10 similar days. Similar days are either business days, or weekend days and holidays.

CMTA proposes that past similar days be defined as "days when the customer's business was in operation." According to CMTA, this does not require the utility to differentiate between weekends, holidays, and two types of weekdays (such as the Cal Steel proposal), but only requires that utilities use the customer's past similar days of electrical usage.

We are not persuaded by CMTA to complicate OBMC baseline calculations in this way. Customer usage may vary for any number of reasons, with variations reflecting more than just whether or not the customer's business was in operation. Further, the definition of "in operation" may be subject to many interpretations.

Moving away from clear, objective criteria will result in increasingly individually tailored baselines. This will be relatively more difficult for utilities to administer, and we are not persuaded that the advantages outweigh the disadvantages. Thus, we decline to adopt CMTA's recommendation.

3.6.2.2. Temperature Correction

CMTA proposes that a temperature correction be built into the calculation of similar days for customers whose loads are significantly affected by changes in ambient temperature. According to CMTA, if the customer's load during the 10 similar days is 10 MW and a 15% OMBC reduction is ordered, the customer must reduce load to 8.5 MW. CMTA says, however, that if the day the OBMC is called is hotter than the 10 similar days, and the customer's actual load is 12 MW when OBMC is called, a reduction to 8.5 MW requires about a 30% reduction in load.

We decline to adopt CMTA's proposal. We agree with PG&E that a temperature adjusted baseline calculation eliminates the benefit of the customer knowing with certainty their targeted maximum load level for each curtailment event. That is, a customer would not know the temperature correction factor until after the event.

Further, temperature adjustment unreasonably complicates the baseline calculation, and significantly increases the burden on the utility. We are not persuaded that the potential benefits of a temperature adjustment outweigh the disadvantages.

3.6.2.3. Stage 1 and 2 Days

CMTA also proposes elimination of Stage 1 and 2 days from the past 10 similar days for calculating the baseline. Otherwise, CMTA asserts, OBMC customers have no incentive to voluntarily reduce operation during Stage 1 and 2 days since it will make compliance during a Stage 3 day more difficult.

We decline to adopt this proposal. Excluding Stage 1 and 2 days allows customers to maintain a relatively high OBMC baseline. This could result in a customer providing less load reduction during Stage 3 than during Stages 1 and 2. At the same time, the customer might participate in what can be lucrative interruptible programs during Stages 1 and 2.

Further, the likely contiguity of Stage 1, 2 and 3 days means that eliminating Stage 1 and 2 days from the definition of similar days could result in a participant's baseline being calculated from "similar days" that are potentially weeks removed from current conditions. The rationale of adopting a 10-day baseline compared to other periods, such as one year before, is to maintain some reasonable relationship with current conditions. We are not persuaded that the benefit of eliminating Stage 1 and 2 days, if any, outweighs the disadvantage of the probable distancing of the baseline calculation from current conditions.

3.6.2.4. Real Time Profile Option

CMTA urges incorporation of a real time profile option for baseline measurement. In this way, according to CMTA, customers can successfully respond based on actual conditions. Customers whose load substantially increases due to short-term high temperatures, for example, and who would otherwise find it extremely difficult to reduce usage from a 10-day baseline, may still participate and provide benefit to the system.

We decline to adopt this recommendation. While a real-time basis has some advantages, it presents its own set of problems. For example, OBMC customers with a real time baseline would have an incentive during the later part of a Stage 2 event to ramp-up their load, thereby reducing the burden of a subsequent 5%, 10% or 15% reduction.

CMTA states that "gaming" can be addressed in any number of ways, but offers no proposals. We decline to make an already relatively complex system more complex by adopting an optional method of baseline calculation, and then developing additional terms and conditions to prevent abuses.

CMTA generally says we "should refrain from making abrupt changes in the programs..." (CMTA Reply Comments, November 16, 2001, page 2.) We think this suggestion applies here.

3.6.2.5. Days to Exclude From Baseline

If its other recommendations (e.g., temperature adjustment, real time load profile) are unlikely to be adopted, CMTA says it is willing to simplify and narrow the scope of its proposals to achieve some incremental improvements in OBMC baseline calculation. The current 10-day baseline calculation allows (a) 15-day ramp-up and ramp-down adjustments, (b) exclusion of up to 10-days when those days are provided in advance to the utility, and (c) exclusion of up to two days permitted after the fact. As an alternative to its other proposals, CMTA recommends that the number of days which a customer may exclude from the 10-day baseline under options (b) and (c) be increased from 10 and two days, respectively, to 15 and 10 days, respectively. Because, according to CMTA, this will not address all problems, CMTA also suggests that once per year each customer be allowed to select a baseline measured by either (a) the past 10 similar days or (b) real time. (Petition for Modification of D.01-06-087 filed on November 9, 2001.) We decline to adopt these proposals, and deny CMTA's petition for modification.

The design of each program must consider the amount of flexibility to give customers. The final balance seeks to meet as many competing needs and interests as possible between participating customers, the utility and other ratepayers. There is only a limited amount of flexibility that can reasonably be permitted here, however, since OBMC is designed to replace firm service interruptions. There must be actual, measurable and dependable load reductions when OBMC is called, or the program has limited or no value. If OBMC is not dependable, additional firm service customers must be interrupted at the time of system need.

The original OBMC program did not permit excluding any days when calculating the baseline. Flexibility was added in June 2001, permitting exclusions of up to 10 and two days in options (b) and (c), respectively. (D.01-06-087.) No OBMC events have occurred since June 2001. There is insufficient actual experience and data to justify program modification.

CMTA asserts that its suggested changes are "modest" and will not impair the reliability of the OBMC program, while providing customers with needed flexibility. We are not persuaded. CMTA does not, for example, construct a hypothetical case to show that the increased days of exclusion from the baseline measurement can or will result in the same or similar operation over a sample period. We decline to make the recommended modifications without supporting data or example. The existing balance between competing interests is, and remains, reasonable.

3.6.2.6. Minimum 30-Minute Notice

OBMC participants are now provided no less than 15 minutes, and possibly up to 29 minutes, to reduce load after notification. In its comments, CMTA recommends that customers be provided notice at least 30 minutes before being required to reduce load under the OBMC program. In the alternative, CMTA recommends that the penalty for failure to reduce load for the first half-hour of the outage be eliminated if the customer still achieves the required reduction on the circuit for the first full hour, and for the remainder of the outage. (October 12, 2001 Comments, page 6.)

In its petition for modification, CMTA says that, as a compromise, it proposes during the first hour of the OBMC event that compliance be measured over the full hour rather than in half-hour increments. (November 9, 2001 Petition, page 5.) In subsequent hours, compliance is measured in half-hour increments. CMTA asserts that the customer will still need to meet the required reduction in the first hour, but will not be penalized if it fails to be in full compliance during the first 30 minutes. CMTA claims this is a modest measure to give customers more flexibility without impairing overall load reduction. Further, CMTA says it is reasonable in light of limited experience and the unproven nature of advance notification by the utility.

We adopt neither proposal, and deny CMTA's petition for modification. We acknowledge that customers may have difficulty achieving 5%, 10% or 15% reductions from their baseline with limited advance notification. The balance that must be struck, however, is between the benefit provided to the OBMC customer (i.e., exclusion from rotating outage), and the benefit that the OBMC customer must provide to the system (i.e., real time load reduction when needed). We are not convinced that we should disturb the existing balance.

An entire advance notification infrastructure is now in place, and several methods of advance notification are available before outages occur. (See D.01-09-020, mimeo., page 26 for a complete discussion.) For example, the CAISO provides forecasts both 48 and 24 hours in advance of expected rotating outages. The CAISO provides frequent updates to the public during periods of forecast electricity emergencies. Executive Order D-38-01 requires that the CAISO notify utilities and public agencies one hour in advance of any firm load curtailment. Each utility is in turn required to notify the public and the media no less than one hour in advance of any reduction in electricity output, including the time and location where the anticipated blackout will occur. Individual customers may also receive energy alerts regarding Stage 1, 2, and 3 emergencies from the State of California. As experience is gained, this advance notification system will become a powerful tool to inform customers.

Utilities also maintain direct notification paths with OBMC and other customers by several means (e.g., by customer account representatives using telephones, electronic mail, pagers). As SCE says, under expected circumstances, OBMC customers will have received both Stage 2 and Stage 3 warnings by electronic mail and pager in advance of potential rotating outages.

As a result, the notification infrastructure now in place gives us confidence that the period of time during which an OBMC customer should develop a reasonable expectation of a rotating outage will be in excess of 30 minutes. We agree with SCE that further expansion of the specific notice requirement to OBMC customers, or measuring results over the first full hour, could have the unreasonable effect of delaying load relief, and necessitating increased firm load curtailments.

3.6.2.7. Monthly Interruptible Contract Requirements

In comments, CIU proposes that interruptible customers be permitted to participate in OBMC after meeting their interruptible contract obligations, whether on a daily, weekly, monthly or annual basis. For example, CIU says an interruptible customer could participate in OBMC after completing the full six-hour per day load reduction required by the interruptible tariff. If there is any conflict between programs, CIU proposes the customer be required to first meet its interruptible obligation.

In its petition, CIU limits the proposal to the interruptible customer first meeting its monthly interruptible obligation. CIU states this revised proposal results from discussions at the workshop, and a compromise between differing parties.

PG&E responds to CIU's petition with qualified support, and states that once it has reviewed operational issues it will be prepared to report back to the Commission. No further report was made, and we are not persuaded that those operational issues, if any, cannot be resolved.

SCE supports CIU's petition with one exception. SCE states that for customers who are participants in both I-6 and OBMC, there is the potential for a monthly simultaneous I-6/OBMC event in which the customer satisfies the final increment of its 40-hour monthly requirement under I-6 but the event continues. Until the 40^th hour, SCE says the customer would have been required to reduce load from the lower of its FSL¹⁰ or OBMC baseline. Under CIU's proposed modification, however, SCE asserts that at the end of the 40^th I-6 hour the customer would be permitted to adjust its load to the OBMC compliance level. SCE says that for a customer with a very low FSL, this circumstance could result in the customer remaining in compliance while significantly increasing its load during the critical emergency. We agree with SCE's concern.

For example, a customer with an FSL of zero, but an OBMC baseline of 40 MW, might be able to increase its load from zero MW (FSL) to as much as 38 MW (a 5% reduction from the OBMC baseline) and still be in full compliance. This could occur even though the Stage 3 emergency continues, and would necessitate even greater firm load curtailments in the form of rotating outages for other customers. Such an adverse system impact is inconsistent with the fundamental goals of these programs. We believe, as does SCE, that the incidence of this particular circumstance is likely to be infrequent, but the potential system impact if it does occur could be significant.

As a result, we adopt CIU's petition with the modification proposed by SCE. During an overlapping event, the present provision remains in effect and the customer will be required to continue to reduce load to the lower of FSL or OBMC baseline during the entire length of that particular OBMC event. For all subsequent OBMC events during that month (or year, after the annual requirements are met), the customer may reduce load from the OBMC baseline.

3.6.2.8. Lead Customer

CMTA proposes that the "lead customer" concept be deleted from OBMC, and that all customers on a circuit who agree to participate in OBMC bear responsibility in proportion to their individual loads. In support, CMTA asserts that very few, if any, customers on shared circuits participate in OBMC. The lead customer concept is one obstacle to greater participation, according to CMTA.

Under the lead customer approach, one customer must notify other customers on the circuit of the OBMC event. Further, the lead customer is responsible for circuit compliance, along with administration of penalties. CMTA says these duties should be the responsibility of the utility. PG&E and SCE oppose CMTA's proposal.

We decline to adopt CMTA's recommendation. OBMC is designed to provide overall system benefits with the expectation that the largest customers on a circuit agree among themselves how to meet program requirements. In exchange, the entire circuit receives the substantial benefit of an exemption from rotating outages. If the lead customer is unable or unwilling to shoulder liability on behalf of the entire circuit, the lead customer may agree with other customers on an allocation of responsibility. Customers may formalize their agreement by contract, providing maximum flexibility, while retaining program responsibility with customers. We think this is reasonable.

On the other hand, requiring all customers to participate in proportion to their load unnecessarily and unreasonably limits program flexibility. Customers may now agree to any allocation of load reductions, along with compliance and penalties. We are not convinced that this flexibility should be removed.

Further, CMTA's proposal would convert a circuit level program to a customer level program. This would require the utility to offer each participating customer on the OBMC circuit an individualized load reduction plan, but create a circuit-wide exemption from rotating outages. The utility would incur additional costs and burdens for compliance measurement and contract administration, including sorting out liabilities and potentially settling liability disputes. This unreasonably increases the burden on the utility not contemplated within the OBMC program.

The OBMC program places the burden "on the customer to demonstrate that the proposal is realistic, workable, measurable, and enforceable." (D.01-04-006, mimeo., page 37.) Several customers on a circuit, however, may propose that the utility administer the program, including allocation of penalties for noncompliance. If the plan is reasonable, we would expect the utility to agree. CMTA may consider developing a standardized plan for utility administration of OBMC, and work with each utility to approve such plan. We decline, however, to adopt CMTA's proposal to vacate the lead customer concept, or to limit the possible types of OBMC plans to one requiring all customers to participate in proportion to their individual load.

3.6.2.9. Costs Allocated to Large Customers

TURN proposes that OBMC program costs be allocated only to large power customers. In support, TURN asserts that program benefits accrue solely to those few industrial customers who participate (receiving an exemption from rotating outages) and not the rest of the system as a whole. We disagree.

Industrial customers are not the only beneficiaries of OBMC. Rather, the entire system benefits by having OBMC circuits reduce load by prescribed amounts. The amounts are generally equivalent to the reduction in system load sought by rotating outages (e.g., 5%, 10% or 15%). Moreover, these reductions are required for the entire duration of the system rotating outage (e.g., several hours), and are not limited to the duration of a rotating outage on one circuit or block (e.g., 60 to 90 minutes). We are not persuaded by TURN to limit OBMC cost recovery to only the industrial class.

3.6.3. CCPCFA Comments

By letter dated October 30, 2001, CCPCFA offers comments on capacity payment interruptible load programs. It also lists nine steps to create a framework for actualizing "the low-cost peaking insurance that we need to ensure system performance through next summer."

We appreciate CCPCFA's proposals, and will endeavor to work with CCPCFA, as with all state and federal agencies, to benefit California. CCPCFA's suggestions, however, are not sufficiently specific to adopt without additional development. We comment on a few of CCPCFA's concepts.

3.6.3.1. Existing Programs and Load Aggregators

CCPCFA proposes that the Commission:

"create and fund through...[utility] rates an interruptible program that had a capacity payment either to an individual customer or an aggregator of customers in exchange for the right to interrupt load on short notice at a specified number of hours per year." (CCPCFA letter to Commissioner Wood dated October 30, 2001.)

All three respondent utilities already have Commission-approved interruptible programs that provide a capacity payment in exchange for the right to have the load interrupted a specified number of hours per year. We generally decline to authorize the use of load aggregators, with limited but specific possible exceptions (e.g., energy efficiency programs, air conditioner cycling programs). Utilities largely offer the same products to the same customers as aggregators. We are not persuaded that the existing system needs to be duplicated.

PG&E reports that it incurs significant ongoing administrative costs with each customer participating in load reduction through load aggregators. PG&E bases this on its experience coordinating with load aggregators in the CAISO's demand response program. We are hesitant to add another layer of cost on an already burdened electricity system absent clear evidence that these additional costs are necessary and reasonable.

The proposal also does not adequately develop how we would monitor and supervise load aggregators. The existing dysfunctional electricity market has provided opportunities for abuse of ratepayers and other participants. Great care must be taken to avoid creating additional opportunities for abuse.

3.6.3.2. Program Potential

The CCPCFA says its proposed program has the potential to create a large pool of interruptible megawatts that can function as peaking resources. CCPCFA says it has received proposals totaling in excess of 2,000 MW that would benefit from such an approach.

We are encouraged by this opportunity, but are dubious of the potential. First, the majority of businesses have told us they want and need to conduct their business, not become part of California's electricity resource base. Second, to the extent some businesses are willing and able to become the equivalent of a reliable, dependable peaking resource that can be called upon every year to satisfy summer peak load, this can be achieved within the range of existing programs (e.g., BIP, RBRP, DPB). Third, CCPFA does not state at what price the 2,000 MW may become available. Without price information, we cannot evaluate whether or not this opportunity is reasonable and worth pursuing. Finally, it is not clear how much, if any, of this 2,000 MW is new capacity, or simply repackaging megawatts already subscribed under existing programs.

3.6.3.3. Cycling of Air Conditioners

The specific example given by CCPCFA involves using aggregators for satellite-directed cycling of air conditioners. We agree there is merit in this idea. In fact, we said in April, 2001:

"The CEC estimates that 14,000 MW of air conditioning load (28% of total load) occurs during the state's summertime peak demand of 50,000 MW...

This is a potentially vast, untapped source of interruptible electricity. Properly partnered with companies such as Comverge, respondent utilities and ratepayers can enjoy benefits with the providing company taking the financial risk. This opportunity needs further exploration.

Therefore, we order PG&E and SDG&E to explore reasonable options for implementing air conditioner cycling, and other electric motor interruption, programs targeted to residential and commercial customers...

PG&E...and SDG&E shall each file and serve an advice letter no later than May 1, 2001. The advice letter shall analyze and report on alternatives, and seek approval of the most reasonable alternatives, including proposed tariffs for implementation...

We caution PG&E and SDG&E that we are convinced one or more air conditioning cycling programs should be approved in each service area. This is the opportunity for PG&E and SDG&E to propose what each believe are the best options for their areas. That is, the advice letters of PG&E and SDG&E should seek approval of the options that each utility finds most reasonable." (D.01-04-006, mimeo., pages 35-36; also see Ordering Paragraph 1 and Attachment A, Item 2.3.4.)

On May 1, 2001, PG&E filed Advice Letter 2110-E. On January 14, 2002, the Energy Division conducted a workshop to discuss PG&E's proposal. The workshop included considering third party (aggregator) implementation of a pay for performance air conditioner cycling program. In a subsequent decision or resolution we will address PG&E's Advice Letter and air conditioner cycling program.

On May 1, 2001, SDG&E filed Advice Letter 1320-E, in which it proposed an EAEI program. Below, we discuss SDG&E's Advice Letter and its EAEI program, which includes air conditioner cycling.

3.6.3.4. CCPCFA Loan And Repayment

The CCPCFA proposes that the Power Authority loan money to qualified parties. The loan would be used to finance installation of necessary equipment and payment of incentives. Concurrently, CCPCFA apparently envisions the Commission approving a program to collect money from customers for repayment of the Power Authority loan. CCPCFA asserts that the result would be peaking capacity available for use by the CAISO or other entity.

A CCPCFA financed loan is intriguing. Absent better understanding of the flow of funds, responsibilities, costs and benefits, however, we decline to participate in a CCPCFA loan arrangement at this time.

3.6.3.5. Capacity Payment

Finally, the CCPCFA proposal includes a capacity payment, but no additional energy payment. Our current portfolio of programs offers this option. We generally think there is room in a portfolio of choices for several types of programs, including different compensation possibilities (e.g., capacity payments, energy payments, exclusion from rotating outage).

3.6.3.6. Conclusion

We appreciate the CCPCFA's thoughts and suggestions. We will continue to develop our programs with their comments in mind.

3.6.4. SDG&E EAEI Program

SDG&E proposes cancellation of its proposed EAEI program. We agree.

SDG&E first proposed the EAEI program on May 1, 2001, in compliance with our order for SDG&E to explore and propose reasonable options for implementing air conditioner cycling and other electric motor interruption programs. (D.01-04-006, mimeo., page 36, and Ordering Paragraph 1, Attachment A, Item 2.3.4; Advice Letter 1320-E.) The proposed program involves residential and commercial air conditioner cycling, curtailment of small commercial lighting, and curtailment of domestic hot water heaters.

As SDG&E points out, however, SDG&E was also ordered to conduct a Residential Demand-Responsiveness (Smart Thermostat) Pilot Program through December 31, 2004. (D.01-03-073.) This pilot project utilizes internet technology to adjust residential heating and air conditioning thermostats.

We are persuaded by SDG&E that it should be allowed to focus its efforts on one program at this time. The Smart Thermostat Pilot Program is underway, with customers being recruited and thermostats being installed. SDG&E states that the Smart Thermostat Pilot Program will be completely operational by Summer 2002. By comparison, the EAEI program has not been approved, a vendor contract has not been awarded, customer recruitment has not begun, and reaching the goal of 5,000 operational switches is unlikely for Summer 2002. SDG&E reports that focus group studies show more interest in Smart Thermostat than EAEI. Limited resources should be devoted to the program with the greater potential for customer participation.

Withdrawal of the proposed EAEI program means we will not direct SDG&E to pursue commercial air conditioner cycling, curtailment of small commercial lighting, and curtailment of domestic hot water heaters. We will continue to pursue residential air conditioner cycling through the Smart Thermostat project. We now think it best to defer pursuit of commercial air conditioner cycling in SDG&E's area, however, until we have more data. Also, we accept SDG&E's assertion that only small load reductions are likely from commercial lighting and domestic hot water heaters. Many commercial customers have already adjusted their lighting load. As a result, there is a reduced potential MW load base from which to secure savings, the amount of "free ridership" with this program would likely be high, and additional savings are better pursued through further light fixture replacement. Finally, the relatively high load diversity of domestic electric hot water heaters reduces the potential for savings.

Withdrawal of the EAEI program also means we permit customers to override the cycling signal. That is, in the Smart Thermostat Pilot Project the customer may override the cycling/curtailment signal, while that is not allowed in EAEI. We are persuaded to pursue the Pilot Program for now, and review whether or not to permit customer override in the future based on review of the Pilot Program results.

TURN argues that both TURN and the Commission agreed that the interruptible program of choice would be air conditioner cycling, not the Smart Thermostat or other program, citing D.01-04-006 in support. To the contrary, the Commission authorized the Smart Thermostat Pilot Program in March 2001. In April 2001, we directed that SDG&E file an Advice Letter proposing an air conditioner and other electric motor cycling program. We always contemplated reviewing the proposed program before its adoption. While we expressed enthusiasm for an air conditioner cycling program in our April 2001 order, we must now consider the proposed program in relation to other alternatives, costs and benefits. We are persuaded that a better use of limited resources is to allow SDG&E to first implement the Smart Thermostat Pilot Program, and rely on results of the pilot program before pursing a competing program.

3.6.5. CEC Proposals

CEC makes several proposals for interruptible programs. First, CEC proposes a non-bypassable surcharge of $0.001/kWh, assessed on all customers who receive distribution service. Second, CEC proposes replacing the current DWR DBP with a renewed, modified VDRP.¹¹ Third, CEC proposes modifications to the current BIP.

CEC's program proposals generally seek to lower the minimum load drop requirement from 100 kW to 50 kW in order to encourage participation by smaller customers. CEC also recommends the use of load aggregators to facilitate program participation. For its modified BIP, CEC proposes changing the performance measurement from a "fixed" amount (reflective of FSL) to a "variable" amount (reflective of more current usage calculated on a moving 10-day baseline) to provide flexibility and promote participation.¹² Also for its modified BIP, CEC proposes adding an additional energy incentive payment of $0.10/kWh for incremental load reductions beyond the committed load reduction, and $0.35/kWh for incremental load reductions distinct from committed load reductions.

We decline to adopt the surcharge or modified VDRP, but we authorize a two-year pilot program to test the merits of a modified BIP.

3.6.5.1. Surcharge

CEC says it proposes a surcharge because utilities are unhappy with cost recovery treatment authorized in D.01-04-006. Further, CEC reports that utilities are concerned with CEC program proposals if additional costs are funded through memorandum account balances with deferred recovery. We are not persuaded that these reasons justify a surcharge.

We have adequately addressed cost recovery, and specifically considered and rejected funding through a surcharge. (D.01-04-006, D.01-07-029.) CEC fails to convince us that our prior decisions should be revisited or reversed.

CEC's proposed surcharge would generate an additional amount of approximately $200 million per year. CEC fails to show that funds collected in current rates are inadequate to fund existing programs, or are insufficient to fund expanded programs. In fact, SDG&E specifically states that the proposed surcharge "would be excessive for SDG&E programs as they are currently designed." (Reply Comments, November 19, 2001, page 11.)

Utilities may seek increases in total annual program dollar limits as needed. (See, for example, D.01-04-006, mimeo., page 80.) Utilities have made no such request, and make no convincing showing that current limits are inadequate.

We also agree with TURN that collecting up to an additional $200 million per year without a specific purpose must be done with great caution. We decline to adopt CEC's proposed surcharge.

3.6.5.2. Replace DBP With VDRP

CEC asserts that the DWR DBP is an unfunded, moribund program and recommends revival of a modified VDRP in its place. We decline to adopt this proposal.

DWR suspended the DBP on December 15, 2001, but states that the program will be available in June 2002, or in the event of a Stage 2 or 3 emergency, whichever occurs first. Thus, the program will be available as needed no later than June 2002 to the same extent that it has at any time been available. A replacement program is unnecessary.

Further, the cost consequences of CEC's proposal are not clear. Implementation costs can be significant because a modified VDRP program with reduced minimum load drop requirements would be open to a substantially larger pool of customers. These costs may exceed potential benefits. We are not inclined to make these changes without more and better information.

At the request of the Governor, parties and the Commission undertook great effort to develop, authorize and implement the DBP. Normal procedures were waived, and the program was implemented expeditiously. CEC fails to convincingly show that DPB should now be replaced with a different program. To the extent program details should be changed, parties may file a petition for modification of D.01-07-025, fully explaining the reasons and proposing specific replacement language. (Rule 47 of the Commission's Rules of Practice and Procedure.) To the extent funding and cost recovery should be considered further, parties may participate in other appropriate proceedings to accomplish that goal.

3.6.5.3. Pilot Test of Modified BIP

Parties raise several concerns that lead us to decline global modification of BIP. Nonetheless, we are sufficiently intrigued with CEC's proposals that we authorize a pilot program.

For example, concerns arise with CEC's proposal to expand the customer base. CEC identifies four groups as potential participants in an expanded program.¹³ CEC fails to convincingly show why most of these customers are not already participating, or cannot participate, in existing programs.

Further, just as with the proposed revival of VDRP, the cost-effectiveness of a modified BIP is unclear. Reducing the minimum load requirement will produce a significantly larger pool of candidates. As PG&E points out, many small customers may be ill equipped to participate in capacity reduction interruptible programs because of the nature of their businesses or hours of operation. There must be reasonable confidence in a customer's willingness and ability to participate before incurring substantial costs for implementation and reservation (capacity) payments.¹⁴ Before doing so, we need better information.

Thus, we decline to revise the BIP as proposed, but will study CEC's proposed program modifications. A pilot study will permit testing the merits of opening interruptible programs to smaller customers, and measuring response on a "variable" basis (i.e., 10-day baseline) rather than a "fixed" basis (i.e., FSL). It will allow testing cost-effectiveness before full implementation.

We adopt a Pilot Base Interruptible Program (PBIP) to last two years. The adopted principles and details are stated in Attachment A. We number the program in sequence based on programs adopted in Phase 1. (D.01-04-006, Attachment A.)

We adopt CEC's concept that transmission system contingencies may justify calling a localized block of participants. We implement the pilot in the San Jose area based on helping to alleviate regional system constraints in that region.¹⁵ The program will be implemented when the CAISO declares a Stage 2 emergency, or when transmission system contingencies justify calling a localized block of participants. We initially cap the pilot program at 50 MW.¹⁶

We increase the incentives to $8.00/kW-month, and $0.15/kWh for energy reductions in excess of the customer's committed load reduction. Smaller customers need a larger incentive to participate since the benefits are likely to be modest compared to the customer's total operating expenses. This occurs when a customer's electricity expenses are already a relatively small percentage of total expenses, and savings from load curtailment are moderate, as is likely with smaller customers.

To permit reasonable assessment of participation and operation, participants must agree to complete an annual customer survey. PG&E, CEC and the Energy Division should discuss the details of the customer survey. Energy Division will be responsible for preparation of the final survey. PG&E will transmit the survey to participants, compile survey results, and report the results.

CEC's proposed program would be triggered when the CAISO declares a Stage 2 emergency and CAISO operating reserves are less than 5%. CAISO states that it would prefer the single requirement of a Stage 2 system emergency being declared (i.e., reserves are expected to fall below 5%). Otherwise, the CAISO states that it would be required to make additional determinations not associated with its normal staged emergency procedures. We think a single requirement is reasonable.

Some parties express concern about the role of aggregators. CEC acknowledges that its recommendation to use aggregators cannot be implemented until several details are resolved. We agree, and decline to authorize the use of aggregators in the pilot program.

Within seven days of the date of this order, PG&E should file an advice letter including the necessary tariffs to implement the pilot program. The tariffs will become effective 10 days thereafter, unless suspended by the Energy Division Director. Any party who wishes to protest the advice letter for the purpose of seeking tariff suspension must file and serve its protest within nine days of the date of the advice letter, to ensure that the Energy Division Director has time to consider the protest before the tariffs otherwise become automatically effective.

PG&E should file and serve reports monthly, within 15 days after the end of each month, to permit monitoring of this program. The reports shall include details on program initiation and rollout (e.g., training, marketing, recruiting); customers (e.g., participation rates, demographics); identification of barriers to customer participation; costs (e.g., startup, operating); operations (e.g., number of interruptions called during each month, customer compliance, assessed penalties); annual customer survey results; and any other information reasonably necessary to assess the costs and benefits of the program. The monthly reports need to be served only on Phase 2 parties that ask PG&E for copies of the reports.

3.6.6. SCE Petition for Modification Regarding Changes
to FSLs

Finally, regarding other proposed modifications, SCE petitions for modification of D.01-04-006. SCE seeks to clarify that an SCE interruptible service customer may not decrease its FSL during the annual November opt-out period. In its reply to responses, SCE further explains it seeks confirmation that SCE customers currently served on closed interruptible rate schedules¹⁷ cannot decrease their FSLs during the annual November opt-out period reinstated by D.01-04-006. We deny SCE's petition.

3.6.6.1. Background

Beginning in 1998, SCE's interruptible customers were permitted to increase their FSL once per year, normally during a 30-day window beginning each November 1. The increase might be partial or total-that is, to partially or completely "opt-out" of the interruptible program.

On October 19, 2000, we temporarily suspended SCE's annual opt-out opportunity. (D.00-10-066.) In particular, we suspended until March 31, 2001 the portion of SCE's interruptible tariffs that allowed "interruptible customers to either opt-out of the interruptible program or change their firm service levels for a 30-day window" beginning November 1, 2000. (Id., Ordering Paragraph 1.) We lifted the suspension a few months later. (D.01-04-006.)

The issue arises because, according to SCE, approximately six of SCE's nearly 600 interruptible service customers requested a decrease in their FSL during the annual 30-day opt-out period that began November 1, 2001. SCE reports that it denied these requests. SCE now petitions for modification of D.01-04-006 to clarify that decreases are not permitted.

3.6.6.2. Discussion

We used the same language to lift the suspension in April 2001 that we used to apply the suspension in October 2000. That is, for example: "we lift the suspension...[and] allow customers to elect to opt-out or change firm service level...." (D.01-04-006, mimeo., page 17.)

SCE argues that lifting the suspension returned SCE and its customers to the position immediately prior to when the suspension was applied. SCE concludes that only increases in FSL were allowed during the November 2001 opt-out window. This is incorrect.

In support of lifting the suspension, we noted that market conditions had dramatically changed from those that existed in prior years. We permitted customers to make necessary and reasonable changes in FSL so that California would have a base of interruptible load upon which to rely for the difficult period ahead. Among other things, we did this to avoid having to unreasonably rely on penalties to drive customer compliance. (Id., pages 15-16.) The change was not limited to increases in FSL.

We also noted that normal variations in customer operations justified lifting the suspension. We said that businesses and other customers (e.g., universities) grow, modify processes, and make other changes over time. We concluded that:

"It is reasonable to allow customers to periodically reassess their situations and either opt-out or change firm service levels to better match current market and business realities with their abilities to interrupt load." (Id., page 15.)

We also said:

"In addition to this opt-out or readjustment, lifting the suspension means customers may annually reassess and make changes as necessary beginning in November 2001." (Id., page 17.)

SCE filed tariffs pursuant to D.01-04-006. The tariffs refer in Special Condition 3 to "adjustments" rather than "increases" in FSL. This language is consistent with our decision that customers should be allowed to periodically adjust (i.e., increase or decrease) FSL in order to secure a reliable interruptible resource base, to reflect normal changes in customer operations, and to avoid unreasonable reliance on penalties to drive compliance.

SCE says that the only decision language discussing the direction of changes in FSL mentions increases, citing language saying customers may "increase their firm service level as of November 1, 2000." (Id., Attachment A, page 1.) This is only a partial reading of our decision, and is not determinative of SCE's petition.

The cited language reflects the fact that lifting the suspension back to November 1, 2000 involved the issue of whether customers could increase their FSL to avoid penalties that had accrued from October 1, 2000 through January 25, 2001. As SCE points out, no customer sought a retroactive decrease in FSL. Rather, the singular concern raised by interruptible customers was how to get out of the program without penalty, or how to decrease penalties by retroactively increasing their FSLs. The adopted language in this one specific case was not intended to negate our goal of letting customers change their FSL during the following November adjustment window.

SCE asserts that interruptible schedules to which lifting the suspension applied are closed to new customers, and that the Commission confirmed this in D.01-04-006. SCE concludes that since these existing schedules are closed, decreases in FSLs are not allowed.

SCE is correct that several interruptible schedules are closed to new customers. Nonetheless, we allowed existing customers then, and allow those customers now, to remain on those schedules, while at the same time permitting changes in FSL. The fact that such schedules are closed to new customers does not require that we limit existing customers to increases in FSL. Rather, we permit customers to make necessary and reasonable changes, thereby allowing California to have a more reliable base of interruptible load upon which it may reasonably depend.

ORA asserts that SCE's petition should be granted because the cost of discounts used as an incentive to create interruptible load is too high. We agree that these programs are not inexpensive and, to address this concern, have placed megawatt and dollar limits on each utility's programs. We are not persuaded by ORA, however, that cost concerns drive whether or not to grant SCE's petition.

Within 30 days of the date of this order, SCE should notify the approximately six customers whose FSL decrease request was denied in November 2001 that each customer has a 15-day window to now elect to reduce its FSL. The FSL reduction will be effective the same date as it would have been if it had been granted by SCE in November 2001.