Most of the rate schedules serve a variety of customers. Segmentation allows the variety of customers within a rate schedule to be further divided into more discrete categories of similar customer types. Segmentation methods include dividing customers by customer categories, usage, climate zone, appliance mix, all-electric versus gas and electric, dwelling type, Standard Industrial Code (SIC), or multi-site customer. Other segmentation methods include identification of homogeneous customer groups with consistent load shapes, identification of costs of service differences on a per-customer basis, and use of statistically valid approaches, such as econometric modeling.
The Workshop Report states that simple segmentation, such as one load profile applied to all customers in a rate category, would facilitate the implementation of direct access, but could result in less accuracy in the assignment of the PX costs because of the larger grouping of customers. A higher degree of segmentation could result in greater accuracy in the assignment of the PX costs, but could complicate and delay the implementation of direct access because of the segmentation process.
The UDCs propose that, initially, the load profile categories should correspond to existing rate categories, with no additional segmentation. The UDCs contend that this is necessary to implement direct access with load profiling on January1, 1998. Although the Workshop Report states that the UDCs are not opposed to the idea of additional segmentation, the UDCs believe that this should be addressed after January1, 1998. According to the Workshop Report, most of the workshop participants agree with the UDCs that the issue of additional segmentation of current rate schedules should be deferred until sometime in 1998.
The UDCs also propose that the issue of which entities should be responsible for developing and evaluating segmentation plans should be deferred without prejudice to a proceeding in 1998. The Workshop Report notes that resolution of this issue is closely tied to the issue of which entities may develop or authorize load profiles.
The UDCs propose that any segmentation should be subject to the following set of criteria:
· customer eligibility must be readily verified to minimize enforcement burdens;
· the possibility for collusive behavior among segment members should be remote;
· segmentation must trade off costs and administrative burden with improved accuracy;
· the load shape resulting from a particular segmentation of a customer class must represent, to a reasonable degree of statistical accuracy, the actual shape for the segment in aggregate; and
· the load profiles resulting from the particular segmentation scheme should be subject to consistent application in determining PX credits and charges, competition transition charges (CTCs) and ISO/PX settlements.
The Workshop Report notes that some of the other participants believe that additional segmentation should occur on January1, 1998. Without additional segmentation, some of the participants are concerned that load profiling errors could be significant, and that these errors could result in an inequitable allocation of costs. Although the impact on individual customers is likely to be small, the CEC believes that such cost differences could have substantial impacts on the cash flows of retailers.
The Workshop Report notes that the CEC believes that the ESPs must be able to pursue their own segmentation schemes in 1998. The CEC believes that the Commission's intent of enabling competition would be undermined if the ESPs are denied this option.
The CEC believes that some additional segmentation will enhance the accuracy of load profiles, and that some level of additional segmentation should occur in early 1998. The CEC points out that some easily identified subgroups already exist in each customer sector, such as inland, higher-cost customers, compared to coastal, lower-cost customers.
The CEC believes that a working group should be started to identify criteria for further segmentation. This working group should also do the following:
· develop a specific test to identify homogeneous customer groups that have load profiles that are statistically different from the class average;
· develop a mechanism for determining the cost of service differences among these groups;
· develop estimates of the incremental metering and sampling costs implicit in additional segmentation;
· identify billing impediments associated with further segmentation;
· identify implications for collection of the CTC resulting from further segmentation, if any; and
· develop a cost/benefit test to ensure that additional segmentation results in an overall net benefit by reducing ISO, PX, and private scheduling coordinator (SC) energy imbalance settlement costs.
Calpine et al. also support the accelerated development of expanded segmentation. At a minimum, segmentation should include the climatic zones which exist within each of the UDC's service territories. Calpine et al. also favor segmentation based upon SICs. Calpine et al. also support the CEC staff's recommendation to permit the ESPs to pursue their own segmentation proposals in 1998.
ORA has examined a variety of stratification variables to determine whether desirable segmentations could be achieved by 1998. These variables include customer usage, climate zone, whether electric space heating is used, service on time-of-use rates, and business type. ORA contends that its load profiling proposal for segmentation is a workable solution for preventing unnecessary costs, and that it would not involve significant system changes that require substantial lead time.
PG&E contends that segmenting the load profiles beyond the rate category level would require additional sample points, entail additional costs, and jeopardize its ability to implement direct access on January1, 1998.
Edison generally supports additional segmentation because it will result in the customers' electricity prices more accurately reflecting their usage characteristics. Edison argues, however, that during the rate freeze period, the additional segmentation provides little or no benefit to customers, because the lower PX charges associated with a lower cost load profile would simply result in a higher assignment of CTC. Edison also points out that some of the proposals to further segment are simply not appropriate because there is not enough cost differential between categories to make segmentation worthwhile.
Edison believes that the interim implementation of load profiles should use only existing rate groups without any finer segmentation. Further segmentation may be worthwhile if it is in accordance with the set of criteria described above.
As discussed earlier, we will require the UDCs to have dynamic load profiles available for customers whose maximum demand falls below 20 kW by July1, 1998. This will relieve some of the inaccuracies that are associated with the use of load profiles. The use of segmentation for dynamic load profiles could further enhance the accuracy of the load profiles by subdividing customers within a rate schedule into more discrete customer load profiles. However, we agree with the UDCs and some of the other commenting parties that the development and agreement of the criteria for dividing customer groups into additional segments is likely to take some time.
Instead of developing segmented load profiles at this juncture, the time will be better spent developing dynamic load profiling. We will, however, defer consideration of segmented load profiles to 1998. We will direct the Energy Division to convene a workshop in conjunction with the UDCs within 100 days from today to discuss the process for segmenting the current rate schedules into more segmented rate categories. The workshop should address the following: the kinds of segmented customer categories that can be created; the kind of criteria that should be used in developing these customer segments; whether the segmented rate categories are justified by the differences in the cost of serving these different customer segments; whether the UDCs are the appropriate entities to develop the segmented customer groups or should others be permitted to develop the load profiles; and the proposed timeline for instituting segmented dynamic load profiles.
At the conclusion of such a workshop, the Energy Division shall prepare a workshop report about the segmentation issues, including its recommendations, a detailed description of any segmentation proposals, and an achievable timetable for implementing the segmented rate categories. This report shall be filed with the Commission's Docket Office within 130 days. Comments to this report may be filed within 21 days of the filing of the report. The Commission will then issue a decision regarding the segmentation issues. If appropriate, the decision will include the procedural schedule for resolving any outstanding issues with respect to segmented load profiles.
This Web page was produced using a Beta version of HTML Transit 3.0