IEEE TRANSACTIONS ON POWER SYSTEMS, VOL. 29, NO. 3, MAY 2014 1481 CPS1 Compliance-Constrained AGC Gain Determination for a Single-Balancing Authority Héctor Chávez, Member, IEEE, Ross Baldick, Fellow, IEEE, and Julija Matevosyan, Senior Member, IEEE Abstract—The integration of non-dispatchable generation has increased the need for ancillary services to maintain frequency control performance, and the determination of adequacy to ensure reliability has become an important operating concern. This work determines the AGC gain of a single balancing authority intercon- nection in order for system frequency to comply with NERC con- trol performance standards, ensuring secondary control adequacy. The particular case of wind expansion is considered, and a simu- lation of ERCOT is presented to evaluate the performance of the formulation. Index Terms—Power system dynamics, power system simula- tion, secondary frequency control, wind power generation. I. INTRODUCTION A CCORDING to the North American Electricity Re- liability Corporation (NERC), “secondary frequency control maintains the minute-to-minute balance throughout the day” [1]. Also, the NERC defines the Control Performance Standard 1 (CPS1), which “provides with a frequency-sensitive evaluation of how well the demand requirements were met” [2]. Although the NERC guides do not explicitly define a performance standard for secondary frequency control (SFC), this work will consider CPS1 compliance as an SFC adequacy criterion. The NERC also defines automatic generation control (AGC) as “the most common means of exercising secondary control [1],” so this work will consider secondary reserves (SR) as the system capacity that is reserved for and deployed by the AGC system to maintain the minute-to-minute demand-generation balance. In a North American context, SR are widely known as “regulation.” The impact of non-dispatchable generation (NDG) integra- tion on SR adequacy has fostered several integration studies, summarized in [3]. These studies are mainly based on estima- tions of net load imbalance (NLI) statistics, where NLI is de- fined in this work as the MW imbalance due to the combined effect of NDG output variability and the 5-min forecast uncer- tainty (from the operation of a real time market) that SFC must regulate. Sudden losses of generation are meant to be controlled Manuscript received April 29, 2013; revised August 07, 2013; accepted September 19, 2013. Date of publication October 25, 2013; date of current ver- sion April 16, 2014. This work was supported in part by the National Science Foundation under Grant ECCS 1065224. Paper no. TPWRS-00523-2013. H. Chávez is with the Universidad de Santiago de Chile, Ingenieria Electrica, Santiago, Chile. R. Baldick is with the Department of Electrical and Computer Engineering, University of Texas at Austin, Austin, TX 78712 USA. J. Matevosyan is with ERCOT, Operations Planning, Taylor, TX 76574 USA. Digital Object Identifier 10.1109/TPWRS.2013.2285727 by primary frequency response, so NLI does not account for such contingencies. Statistics on NLI are then used to determine capacity requirements for SR. The various estimation method- ologies, not necessarily consistent, lead to different outcomes, as noted in [3]. These “static” analyses do not guarantee CPS1 compliance per se, because:  the AGC system must be properly defined (tuned) to track net load, and  the generation driven by AGC must be capable of fol- lowing AGC instructions. Thus, the adequacy analysis of SFC must consider the dynamics of AGC. Some works have established conditions for the AGC system to comply with CPS1, but they are based on dynamic determi- nations of AGC gains and SR that may not be implementable in AS electricity markets. The work [4] defines fuzzy rules to tune the gain of a proportional-based AGC. The need for more (or less) SR is dynamically (minute-to-minute) determined by the CPS1 value of the twelve previous months, leading to a time-varying AGC gain and dynamic SR requirements. Normal markets require that SR be available one day in advance for the market to execute, so dynamic requirements are not immedi- ately applicable to such markets. Other works establish dynam- ical, CPS1-constrained AGC gain algorithms [5], [6], presenting similar limitations. However, these works suggest that the AGC structure needs to be updated for CPS1 compliance under in- creasing NDG scenarios. A simulation-based determination of SR is presented in [7]. That report simulates frequency dynamics and determines SR needs based on CPS1 compliance and the use of fast storage. A key finding of the report is that the AGC structure needs to be revised for the system to improve CPS1 and take advantage of the fast storage for high NDG penetration scenarios. The re- port considers the scaling of an existing one day time series of NDG power output. However, this approach may be insuffi- cient to represent statistics of annual or monthly operation. Al- though a similar report on the PJM system [8] considers a longer term simulation and more representative CPS1 statistics, it does not provide a criterion to determine AGC gains to comply with CPS1, and, instead, provides qualitative observations or estab- lishes adequacy by trial and error. The paper [9] has the same limitation. The work presented in [10] shows, by a statistical derivation, the negative impact of wind integration on CPS1 performance. A simulation of a two-area system based on the IEEE RTS-96 system is presented to show the predicted effect on CPS1. Al- though the paper proposed ways to improve CPS1 (increase pri- mary reserves, increase the forecast accuracy of wind power, 0885-8950 © 2013 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.