From: owner-ros@lists.ercot.com on behalf of Breitzman, Paul [PaulB@gplops.org]
Sent: Thursday, July 07, 2005 5:59 PM
To: 1 2004 ROS and others
Subject: July 14, 2005 ROS Meeting Agenda -Item 10

Follow Up Flag: Follow up
Flag Status: Blue



At the July 14 meeting, the ROS is expected to vote on a recommendation regarding PRR 586 - SCE Performance and Regulation Cost Re-allocation.  This is a serious matter to which the PDCWG has devoted four special meetings, not just on the specific impact of PRR 586 but on a whole range of issues involving frequency control, regulation deployment, and QSE performance.



Given the nature of the discussion on this PRR at the June 16 ROS, I have serious concerns that the ROS can reach consensus on a recommendation at the July 14 meeting unless we can organize and focus the discussion.  Accordingly, I propose that the ROS discussion of this matter should address the following questions in sequence:



1.  Does the ROS agree with the PDCWG that there is a frequency reliability issue throughout the day and not just across the 0600 and 2200 ramp windows?



2.  If the ROS agrees that there is a reliability issue throughout the day, should PRR 586 be implemented to address SCE performance on a stand alone basis, or as part of a wider set of issues?



3.  If the implementation of PRR 586 should be done as a part of a wider set of issues, does the ROS endorse the following specific initiatives?



a. Primary frequency control in ERCOT must be improved.  Either by a new ancillary service, or by enforcement of the frequency response standard in PRR 468, it is vital that ERCOT be able to rely on adequate primary frequency control.



b. Secondary frequency control by ERCOT must be improved by the use of a realistic variable bias, or by the use of alternative mechanisms with similar effect. With such a change, coupled with adequate frequency response from QSEs, ERCOT can achieve better frequency control that is less burdensome on QSEs.



c. Clarity is needed as to  what are the appropriate sources for Responsive Reserve Service (RRS), the requirements for the providers of RRS, and how RRS is to be used to maintain system frequency.  It needs to be established to what extent RRS is a ten minute ramp deployed service, and to what extent RRS must be available on unloaded units with an active governor.  It may be necessary to also discuss limits on LAAR participation and deployment.



d. Clarity is needed on the obligation of Ancillary Service providers, particularly providers of regulation service, to maintain units on AGC.



4.  If the foregoing improvements establish conditions under which willing QSEs can more easily minimize SCE, and a mechanism such as that established under PRR 586 is a part of a package of system improvements, what modifications are reasonable to make to PRR 586?



a. Should a differentiation be made for 'good' SCE and 'bad' SCE that is based on actual system frequency.



b. Should expected governor response (QSE bias) be removed from the SCE equation when measuring compliance under PRR 586.



c. What is an appropriate QSE deadband, either in MWs or percentage of scheduled generation, that would provide an umbrella to well performing QSEs against the cumulative effect of minor uncontrollable errors.



d. What is an appropriate ERCOT deadband, either in MWs or Hz, that would establish limits wherein no QSE errors would be counted.



e. What is the appropriate allocation of the total cost of Regulation Service that is covered by the PRR 586 re-allocation mechanism, and the remaining allocation that is left solely to Loads.



By this sequence of questions, I am trying to build on what progress was made in discussing this matter at the June 16 ROS.  By looking at the larger picture, I also hope we can address the concern of those who feel QSE performance should not be addressed while other items exist that also contribute to frequency excursions and excessive regulation requirements.  I look forward to discussing this with all of you next week.