Pacific Gas & Electric (PCG)

Table 1. Values of v1 and v2 for Percentiles of the CL in Specified Ranges

Thus, for example, if for a particular Voltage Class the percentile of the historical CL is 55%, this says that the CL is located at the 55 percentile of all bootstrap means in the "boot" column. From Table 1, v1=6, and v2=8.

•

Test 3:    At least two out of three consecutive annual index values fall outside the UWL or LWL on the same side of the CL.



•

Test 4:    Six or more values are consecutively increasing or consecutively decreasing.

        Therefore, Test 1 is designed to detect a short term change or jump in the average level. Tests 2 and 4 are looking for long term changes. Test 2 will detect a shift up in averages or a shift to a lower level. Test 4 is designed to detect either a trend of continuous increase in the average values or continuous decrease. Test 3 is designed to assess changes in performance during an intermediate period of three years. If Test 3 is satisfied, the evidence is of a decline (or increase) in Availability over a three year period. Together the four tests allow the ISO to monitor the availability performance of a Voltage Class for a PTO.

        If none of these tests indicates that a change has occurred, performance shall be considered to be stable and consistent with past performance. If one or more of these tests indicates a change then Availability performance shall be considered as having improved or degraded relative to the performance defined by the control chart. Table 4.3.1 provides a summary of the performance indications provided by the tests. The control chart limits may be updated annually if the last year's Availability performance indices did not trigger any of the four tests. If none of the four tests are triggered, the new limits will be constructed including the last year's data.

        The control chart limits may be modified each year to reflect the number of Transmission Line Circuits in service during that year if necessary. However, it is suggested that unless the number of lines changes by more than 30% from the previous year, the use of the median number of lines should continue. Consider an example. Suppose after the control chart has been prepared for a Voltage Class, next year's data arrive with the number of lines 30% higher than the median used in the past. New limits will be generated in order to assess the Availability performance for that year.

        For the special case where only one Transmission Line Circuit has a Forced Outage^(IMS) in a Voltage Class during a year, the assessment process for Index 2 is as follows. If Index 2 for this Transmission Line Circuit does not trigger any of the four tests, no further action is necessary. If, however, one or more of the tests are triggered, then limits for this Transmission Line Circuit for that year should be recalculated based on the historical data for this Transmission Line Circuit alone using an individual and moving range control chart. The only test warranted here is Test 1. More information

64

on the individual and moving range control charts can be found in the user manuals of the statistical software approved by the ISO for use in creating the control charts

        If the ISO deems that the Availability Measure Targets should be modified, they shall be modified in accordance with Section 8.0, "Revision of ISO Maintenance Standards," of this document.

65

Table 4.3.1 Performance Indications Provided by Control Chart Tests