BAL-001-TRE-1

1. BAL-001-TRE-1 Primary Frequency Response in the ERCOT Region Overview Jan 10, 2014 Future Ancillary Service Team Workshop By: Sydney Niemeyer – NRG Energy

2. Requirements 1 & 2 • Identify and post events that will be measured. • Measure each generator’s performance.

3. Requirements 3 – 5: Maintain Interconnection Frequency Response R3. The BA shall calculate the Interconnection minimum Frequency Response (IMFR) in December of each year for the following year, and make the IMFR, the methodology for calculation and the criteria for determination of the IMFR publicly available. M3. The BA shall demonstrate that the IMFR was calculated in December of each year per Requirement R3. The BA shall demonstrate that the IMFR, the methodology for calculation and the criteria for determination of the IMFR are publicly available.

6. Requirements 3 – 5: Maintain Interconnection Frequency Response

7. Requirements 3 – 5: Maintain Interconnection Frequency Response R5. Following any FME that causes the Interconnection’s six-FME rolling average combined Frequency Response performance to be less than the IMFR, the BA shall direct any necessary actions to improve Frequency Response, which may include, but are not limited to, directing adjustment of Governor deadband and/or droop settings. M5. The BA shall provide evidence that actions were taken to improve the Interconnection’s Frequency Response if the Interconnection’s six-FME rolling average combined Frequency Response performance was less than the IMFR, per Requirement R5..

8. ERCOT Primary Frequency Response – BAL-003-1 2014 Bias and Frequency Response Measure Page 1

9. ERCOT Primary Frequency Response – BAL-003-1 2014 Bias and Frequency Response Measure – Page 2

10. R6. Each GO shall set its Governor parameters as follows: R6.1. Limit Governor dead-bands within those listed in Table 6.1, unless directed otherwise by the BA. M6. Each GO shall have evidence that it set its Governor parameters in accordance with Requirement R6. Examples of evidence include but are not limited to: • Governor test reports • Governor setting sheets • Performance monitoring reports M6.1 The GO shall have evidence that it set the Governor dead-bands as required in Table 6.1 in Requirement R6. Table 6.1 Governor Deadband Settings

11. R6.2 Limit Governor droop settings such that they do not exceed those listed in Table 6.2, unless directed otherwise by the BA. Table 6.2 Governor Droop Settings

12. Proportional Droop From The Dead-band For digital and electronic Governors, once frequency deviation has exceeded the Governor dead-band from 60.000 Hz, the Governor setting shall follow the slope derived from the formula below. where MWGCS is the maximum megawatt control range of the Governor control system. For mechanical Governors, droop will be proportional from the dead-band by design.

13. Requirements 7 & 8 • Operate with governor in service. • When on-line and released for dispatch. • Report when governor out of service.

14. R9 & R10 • Generator Performance measure for Initial response. Measures response during 20 to 52 seconds after the the beginning of the event. • Uses ERCOT data. Time period allows for some data delay and time to deliver response. • Generator Performance measure for Sustained response. Utilizes the frequency value at 46 seconds after the beginning of the event and measures generator’s best response during the 46 to 60 second time period. • Measures utilize Per Unit scaling, Actual Response divided by Expected Response with minimum compliance of 0.75 P.U. rolling monthly average on both measures.

15. Pre-event Frequency Initial Measure “R9” Expected Primary Frequency Response “EPFR” Sustained Measure “R10” Expected Primary Frequency Response “EPFR” Frequency @ t(+46) for “Sustained” measure R10. Pre-event Average MW Event Average Frequency for the “Initial” measure R9.

18. When the generator is in a down ramp at the time of the event, the post event performance will be impacted by the AGC sent to the generator that the generator has yet to deliver. The BAL-001-TRE measurements adjust for this pre-event down ramp.

19. Adjustments To Measures • Steam Turbines: • Initial inlet pressure less than rated pressure. • If a generator slides pressure, response will be less. Adjustment compensates for the reduced pressure. • Steam Flow adjustment. • At outputs greater than 50%, backpressure due to steam already flowing through the turbine, additional steam flow has resistance to increase flow. This is small reduction in expected performance. • For flows less than 50%, the adjustment increases the expected performance.

20. Adjustments To Measures • Steam Turbines: • The change in inlet steam pressure caused by “responding” to the change in frequency reduces the expected response. • Combustion Turbines: • The change in mass flow through the turbine caused by the change in grid frequency (turbine speed) reduces the expected response.

21. BAL-001-TRE • R9 & R10 Compliance • Based on average performance of past 12 months with a minimum of 8 event participation. If fewer than 8 events are measured in 12 months, the time period continues until 8 events are measured. • Some infrequently used generators will never get measured. • Generator must have a minimum of 5 MW or 2% of HSL margin to be measured during an event. • Increases the importance of keeping the HSL accurate as possible.

22. Generator either “Off Line”, below LSL or less than 5 MW or 2% margin during these events. If the generator with margin responded in the wrong direction, the minimum P.U. score is zero.

25. Equipment limits during events can result in low measures. The rolling 12 month average reduces the need to adjust for these events.

28. Summary • All generators participate equally. • Each generator has to do less with full participation. • Measures are designed to account for known limitations of delivery of Primary Frequency Response. • Properly implemented, each generator should easily meet the minimum performance requirements.