III. Benchmarking Performance

A. Existing Generating Assets

Existing generating assets shall be benchmarked against their own historical performance. To facilitate this process, the generating asset owner shall, within 60 days of the announcement of the Generation Maintenance Program implementation, submit to the CPUC the necessary data to calculate the monthly Capacity Unavailability Factor (CUF) for at least the past 5 years (60 months). In the event that a unit owner does not have reliable, continuously-recorded CUF data for the 60-month minimum, the CPUC may determine the control chart limits using data for a shorter period or utilize an alternative means for establishing the limits until such time as sufficient CUF data is available for the unit. The CPUC shall periodically review the control chart limits and appropriately modify them when necessary.

New benchmarks shall be calculated each year, by adding in the past 12 months of data to the historical data originally used to calculate the benchmarks.

If the asset owner is a participant in the NERC GADS system, they have the alternative of authorizing NERC GADS to release the data to the CPUC.

B. New Generating Assets

Prior to the commercial date of operation, the generating asset owner shall submit to the CPUC its proposed, pro-forma benchmark for CUF, including a summary description of how the proposed value was derived. If the derivation did not include, as a minimum, comparison with industry class data, the CPUC will make such a comparison with like type units before accepting the proposed benchmark. Once accepted, the pro-forma benchmark becomes the starting point for performance monitoring, until such time as sufficient historical data is available for the unit.

C. Benchmark Calculations

Using the monthly-calculated values for CUF, the Quarterly Mean Value of CUF will be calculated for the historical period. The overall average across all quarters will be the unit specific Centerline for the Control Chart. The monthly-calculated values would also be used to establish the Upper Warning Limit and Upper Control Limit as required for the performance analysis thresholds described below. The Upper Warning Limits and Upper Control Limits are calculated from the historical data using a statistical technique called "bootstrap resampling". This technique is used when data doesn't fit a normal distribution, and is intended to establish an Upper Control Limit such that only .25% of the quarterly CUF values are likely to fall above the Upper Control Limit value, and only about 2.5% of the quarterly CUF values are likely to fall above the Upper Warning Limit value, assuming nothing unusual occurred. New Upper Control Limits, Upper Warning Limits and Centerline Values (benchmarks) will be calculated each year. This information will be provided to the generating asset owner.

The CPUC will use a set of performance analysis thresholds to monitor the actual performance of each generating asset. Activation of any one of the performance analysis thresholds would result in the CPUC initiating additional performance analysis using other data.

D. Performance Analysis Thresholds

The following thresholds shall be used by the CPUC to prompt a more thorough analysis of the generating asset's performance.

1. Current Quarter CUF Value exceeding Upper Control Limit.

2. Two out of the last three Quarterly CUF values exceeding the Upper Warning Limit

3. An adverse trend of CUF as defined by 6 Quarterly CUF Values in a row with increasing values.

4. At least v1 consecutive Quarterly CUF Values above the Benchmark Value.

NOTE: v1 is a derived value obtained during the calculation of the Upper Warning Limits and Upper Control Limits using a statistical technique called "bootstrap resampling".

In addition such performance evaluations could also result from other sources such as:

· CPUC Site Visit Feedback

· Forced outage rates

· Excessive outage extensions

SECTION 3

VERIFICATION AND AUDIT PROCESS

INTRODUCTION 3

I. INITIAL CERTIFICATION 4

II. PERIODIC RE-CERTIFICATION 7

III. NOTICE OF SIGNIFICANT CHANGES 9

IV. INITIAL AUDIT PROGRAM 10

V. RANDOM AUDIT PROGRAM 11

VI. TRIGGERED AUDITS 12

VII. EXAMPLE INITIAL CERTIFICATION REPORT 13