8. Conclusion

With each major storm event and subsequent investigation, it is a challenge to balance the desire to respond immediately and specifically to the unique circumstances that arise against the need to carefully review each event and avoid a crisis management response. We firmly believe that it is of little value to adopt standards that apply to situations of limited duration or that are unlikely to repeat themselves. This proceeding is no exception. Our primary objective in PG&E's TY 2003 GRC is to ensure that PG&E continues to provide utility service at the lowest reasonable rates, maintaining a high level of customer service and satisfaction, and a safe working environmental for its employees. In this phase of the GRC, we are reviewing PG&E's overall reliability performance and storm response to determine whether PG&E has met this level of service and whether additional standards or metrics are necessary to ensure that PG&E continues to provide this level of service. Our detailed review was focused mainly on controversies which arose between PG&E and other parties and a comparison of PG&E's performance to previously-established performing standards.

The December 2002 storms consisted of a series of four severe storms that occurred within a period of nine days. These storms severely tested PG&E's facilities and staff and highlighted many weaknesses in PG&E's organization. While PG&E maintains that its overall reliability performance was reasonable, it admits that its performance during the December 2002 storms requires improvement, especially in the area of outage communications, and has identified several initiatives designed to prevent similar situations from occurring in the future. Other parties also proposed various measures designed to improve PG&E's performance.

While we approve the majority of the PG&E/ORA Agreements, we do not adopt the PG&E/CUE proposal for a performance incentive mechanism. Based on the record in this case, the PG&E/CUE proposal is not in the ratepayer's interest. There is a lack of record evidence supporting a need for improved reliability during normal conditions as well as a lack of evidence that the incentive mechanism will benefit ratepayers. We find that the incentive proposal is not likely to result in achieving our basic regulatory objective of maintaining the lowest reasonable rates consistent with safe, reliable, and environmentally sensitive utility service because it would place ratepayers at a significant risk of paying for the same level of reliability two or three times.

Under cost of service ratemaking, our objective is to adopt a revenue requirement that allows the utility to provide high quality service at just and reasonable rates. Adopting a revenue requirement necessarily includes a presumption of a certain service level. While we support PBR-style incentives in concept, the incentives must be consistent with and not jeopardize our other regulatory goals. We must also avoid using incentives as a substitute for the utilities' statutory obligation to provide high quality service, especially in monopolistic utility markets. In this case, we find that the combination of traditional cost of service regulation and the proposed PG&E/CUE incentive mechanism is likely to result in ratepayers paying twice for the same level of reliability.

Although we decline to adopt the proposed incentive mechanism, we believe that it is necessary to provide additional guidance to PG&E management regarding the level of reliability we consider adequate. In the TY 1996 GRC, the Commission found that in the absence of any statute or Commission rule defining statutorily acceptable performance, and recognizing that reliability is strongly tied to costs, statutory reliability should be defined as the level of reliability historically accepted as reasonable, as measured by indices in use at the time. In the TY 1999 GRC, the Commission found that PG&E had been providing "adequate" service for the period from 1996-1998 and set a revenue requirement sufficient to allow PG&E to continue to meet that level of service. "Adequate" service was defined in D.00-02-046 as service "continuously meeting and exceeding public demand for the firm's output" and "in compliance with all laws, regulations, and public policies that govern public utility facilities and operations." The standard we set in this case is based on historical performance, adjusted to reflect adopted revenues and reasonable expectations of performance, adjusted to reflect adopted revenues and reasonable expectation of performance.

In the proposed decision, we put forward a proposal to adopt a minimum reliability threshold to the average reliability level achieved during the 1999-2002 period as reflected in the SAIDI, SAIFI and MAIFI measurements. In its comments, PG&E requests that the Commission change the phrase "minimum reliability threshold" to "target reliability," arguing that setting a minimum reliability threshold is inconsistent with how the Commission has historically evaluated the level of reliability for PG&E. PG&E apparently misunderstands our intent. In today's decision, although we find PG&E's performance reasonable, we note several concerns that have been raised regarding PG&E's performance, how this performance is measured and reported, how it compares with other utilities' performance, and whether or not customers are satisfied with the performance. Given these concerns, we find that it is necessary to adjust our standards. We are not, as PG&E suggests, "simply reaffirming the principles of statutory reliability as described in D.96-09-045 and affirmed in D.00-02-046." (PG&E Opening Comments, p. 5) Instead, we adopt a new approach, and set a minimum standard for performance that we expect PG&E to meet. While the "adequate standard" used D.00-02-046 evaluated the full range of performance during the period from 1996-1998 and did not identify a specific numerical value, the standard we adopt here consists of a single reliability performance threshold for each adopted measure of reliability (e.g., SAIDI and SAIFI).

In their initial comments CUE and TURN raise concerns regarding the reliability performance standard, arguing that adoption of the historical average is inconsistent with the findings reached in rejecting the CUE/PG&E joint testimony. CUE and TURN argue that the reliability standard for PG&E should be consistent with the reliability improvements being funded by ratepayers.

TURN notes that the Commission's proposed decision on the GRC settlement would approve distribution revenues for O&M that are 98% of the level requested by PG&E. TURN argues that ratepayers have a right to expect that the work they are funding be performed and the associated reliability improvements are achieved. TURN suggests that an appropriate measure would be an average of the TURN and ORA forecasts that incorporated the funding requested by PG&E. TURN recommends that the Commission adopt the average of the TURN/ORA reliability projections for SAIDI and SAIFI as the basis for the determining the reasonableness of PG&E's performance. This average would result in SAIDI benchmarks of 157.5 for 2005, 153.9 for 2006, and 152.2 for 2007 and SAIFI benchmarks of 1.2-1.3 for 2005 and 2006, and 1.1-1.2 for 2007. CUE recommends that the Commission set standards equal to the TURN forecast, resulting in SAIDI benchmarks of 161.1 for 2005, 157.8 for 2006, and 154.5 for 2007; and SAIFI benchmarks of 1.25 for 2005 and 1.15 for 2006 and 2007.

PG&E appears to agree with TURN and CUE on the limited point that the standards should take into account the funding provided to the utility, by stating that the Commission should continue to consider the "reliability triumvirate" of customer expectations, company performance, and associated funding when setting reliability standards, but disagrees with respect to the appropriate SAIDI and SAIFI levels.

The CUE proposal to use the TURN forecast as the adopted reliability standard is consistent with the findings reached in rejecting the CUE/PG&E joint testimony, and we adopt it. Adequate service should be defined as the service deemed reasonable, consistent with adopted revenues. Although PG&E argues that "the fact that the Settling Parties set forth specific amounts for certain categories of costs is not intended to limit PG&E's management discretion to spend funds as it sees fit and consistent with its obligation to serve," we agree with CUE and TURN that ratepayers (and the Commission) should be able hold PG&E to a standard of reasonableness consistent with the funding level being provided to PG&E. Ratepayers should also be able to assume that PG&E would spend the adopted revenues in a manner consistent with the activities and expected results described in PG&E's revenue requirement requests.

It is for this reason, among others, that we find it necessary to adopt not just a target, but a minimum standard. This may be a departure from the standards set in the last rate case, but it is a necessary departure. We find that the record supports adoption of the following minimum performance standards:

We find that meeting or exceeding this reliability level should establish a rebuttable presumption of reasonableness applicable to the utility's SAIDI and SAIFI performance. Conversely, failure to achieve the levels deemed reasonable should create a prima facie case of a violation of this reliability standard. We do not believe it is necessary to adopt a deadband or similar adjustment to allow for year-to-year weather variations for three reasons: (1) there is no automatic penalty mechanism, and the utility will have the opportunity to demonstrate why the threshold was not met in any subsequent investigation or penalty phase; (2) we are basing our performance standard in part on a four-year average that already reflects significant weather variations and excludes major events; and (3) PG&E's TY 2003 GRC forecast includes programs and funding that are expected to improve reliability beyond the average levels achieved from 1999 through 2002.

We note that PG&E's CAIDI performance has not improved since the 1996 period and, in fact, has declined in recent years. We do not set performance targets for CAIDI at this time but will consider doing so as part of PG&E's next GRC if PG&E's performance in this area continues to decline.

As stated above, we direct PG&E to prepare a value of service study prior to its next GRC. The updated value of service information will inform the Commission regarding PG&E's customers' desire for and willingness to pay for increased reliability.

Although we find that PG&E has provided adequate service during normal conditions, based on the record in this proceeding, we also find that PG&E's outage communications during the December 2002 storms do not reflect a reasonable level of service. PG&E has admitted that its storm response needs improvement, but maintains that, on an overall basis, its response to the December 2002 storms was reasonable. We disagree. We believe that while the record demonstrates that the outages and damages caused by the storms were reasonable considering the severity and the back-to-back nature of the storms, given the many outage communication and call center problems that occurred during the storms, we cannot find that PG&E's storm response was reasonable. In particular, PG&E concedes that its method for addressing single customer outages failed, resulting in single customer outages being unrecorded and unresolved. PG&E also admits that calls from emergency personnel were handled in a manner that resulted in police and fire personnel standing by hazardous conditions for excessive periods of time during the storms. Given this evidence, we cannot find that PG&E's overall storm response was reasonable. However, based on the fact that PG&E has admitted its deficiencies and begun implementing remedial measures, we do not find that any sanctions or penalties are necessary. We also note that none of the parties requested sanctions or penalties related to PG&E's storm response.