Ancillary Service Methodology - CIMView

Register Here During GCPA : Register

Ancillary Service Methodology

Table of Contents

The Gist

Every year ERCOT looks at historical data to determine how much Ancillary Services (reserves, or ‘AS’)  it needs to procure over the coming year.  Currently it does this for four reserves; Regulation Up (REGUP) , Regulation Down (REGDN) , Non-Spinning Reserves (NSRS, or NonSpin), and Responsive Reserves (RRS).  The current (as of 1/1/2022)  methods for determining these reserves are as follows:

REGUP  

For each month take the larger of A) the 95th percentile of REGUP deployments for the same month of the previous two years and B) the 95th percentile of the positive net load changes for the same month of the previous two years.

REGDN

For each month take the larger of A) the 95th percentile of REGDN deployments for the same month of the previous two years and B) the 95th percentile of the negative net load changes for the same month of the previous two years.

For both REGUP and REGDN ERCOT will adjust amounts throughout the year based on how much wind and solar gets added.  The adjustment tables which detail MWs of REGUP and REGDN change per 1000MW of added Wind, and same -different table- for Solar, are included in the methodology document .

Also, if the CPS1 score (a metric for determining how well frequency is in control) drops below certain thresholds ERCOT will increase the amount of REG it purchases.

NSRS

NonSpin is based on analysis of four hour blocks of the same month in the previous three years.  The main component is based on “Net Load Uncertainty”.  For each 5 minute interval in an hour the net load is determined by subtracting the uncurtailed total wind and solar output from the load.  The highest of those 12 net loads for an hour is compared to the 6 hour forecasted value of the net load and the difference is the Net Load Uncertainty.  The main component of NonSpin is either the 85th percentile (for low risk intervals) or the 95th percentile (for high risk intervals) of the net load uncertainty.

A second component of the NSRS is calculated as the 75th percentile of the historical intra-day forced outages of thermal resources (since midnight) for six hour blocks of each month in the last three years.

A third component is a variable (up to 1000MW) quantity of NonSpin that ERCOT may buy if it feels an upcoming day is high risk.

Finally, Like REG, NonSpin amounts will be adjusted throughout the year based on the amount of additional installed wind and solar capacity.  Tables detailing these adjustments (per 1000MWs of additional installation are in the Methodology document.

New for 2022: 

  1. Basing Net Load Uncertainty on a 6 hour forecast (previously 4 hour forecast)
  2. The forced outage compnent
  3. The variable additional 1000MWs (introduced mid year 2021)

RRS

RRS is based on the amount that ERCOT studies are needed to maintain frequency during sudden shocks to the  system based on expected load, solar and wind patterns.  Over the peak hour blocks at least 2800MW of RRS will be purchased.  At all time at least 40% or 1,420MW of RRS, whichever is greater, must be provided by resources using Primary Frequency Response (PFR, usually conventional generation with governors).  The most RRS that can be provided by Fast Frequency Response resources (usually batteries) at any given time is 450MW. 

New for 2022: The 2800MW floor over peak.

Important Actions

  • 1/1/22 Implemented
  • 2/24/22 ERCOT clarifies that the additional 1000MW of NonSpin is not just for high variability days but also for high risk days (e.g. high winter load)

Reference

2022 AS Methodology (ERCOT)

Meeting Notes

2/24/22 WMWG

11/29/21 TAC

11/3/21 WMS

10/25/21 WMWG

Feedback?

Send any corrections/clarifications to notes@cimview.com