Resource transition model under MISO MIP based Look Ahead Commitment

Abstract

MISO is in the process of implementing Look Ahead Commitment (LAC) to bridge the gap between Reliability Assessment Commitment (RAC) and Real Time Security Constrained Economic Dispatch (RT-SCED) timeframes. This paper introduces two models in LAC pre-processing logic. The first one is anticipatory startup and shutdown model to account for resource megawatt output during resource startup and shutdown. The second one is an infeasibility identification model to identify data conflicts between state estimation, unit commitment plan and offered parameters. The goal of these two modules is to properly estimate non-dispatchable resource outputs and resolve infeasibility from input data so that LAC can achieve good Mixed Integer Programming (MIP) solutions. Index Terms-Ancillary services, Co-optimization, Dispatch and scheduling, Reserve, Unit commitment 1 , Mixed Integer Programming I. NOMENCLATURES r: index for all resources t i : starting time of interval i t 0 : time when LAC case starts to execute IntervalLength(t i ): interval length in minutes for interval i Binary decision variables: CF(r,t i ): commitment decision at interval i SU(r,t i ): startup status at interval i. SU(r,t i ) equals 1 if CF(r,t i-1 ) equals 0 and CF(r,t i ) equals 1. SD(r,t i ): shutdown status at interval i. SD(r,t i ) equals 1 if CF(r, t i-1 ) equals 1 and CF(r, t i ) equals 0. Continuous decision variables: DispatchMW(r,t i ): energy dispatch MW at interval i Input parameters: EcoMin(r,t i ): economic minimum limit in MW EcoMax(r,t i ): economic maximum limit in MW RegMin(r,t i ): regulation minimum limit in MW RegMax(r,t i ): regulation maximum limit in MW BiRampRate(r,t i ): bi-directional ramp rate in MW/Min UpRampRate(r,t i ): single directional up ramp rate in MW/Min DnRampRate(r,t i ): single directional down ramp rate in MW/Min RF(r,t i ): regulation commitment (0/1) surr(r, t i ): startup ramp rate in MW/Min Yonghong Chen, Vijaya Ganugula and John Williams are with MISO, Carmel, IN. Jie Wan and Ying Xiao are with Alstom Grid. Disclaimer: The views expressed in this paper are solely those of the authors and do not necessarily represent those of MISO. sdrr(r, t i ): shutdown ramp rate in MW/Min SEMW(r, t 0 ): resource r state estimator output MW at t 0 Pre-calculated parameters: SUMW(r,t i, t i-k ): anticipatory startup MW at the end of interval i-k if SU(r,t i )=1 (startup curve for startup at t i ) SDMW(r,t i, t i+k ): anticipatory shutdown MW at the end of interval i+k if SD(r,t i )=1 (shutdown curve for shutdown at t i ) SUSDMW(r,t i ): anticipatory startup or shutdown MW at the end of interval i due to startup and/or shutdown from the commitment planIf RF(r,t j )=0, do the following:Step A6: calculate BoundMW(r,t j ) = Max (EcoMin(r,t j ), BoundMW(r,t j-1 )-DnRampRate(r,t j )·IntervalLength(t j ))Step A7: if BoundMW(r,t j ) calculated from Step A6 is above EcoMax(r,t j ), then set MaxLimit(r,t j )= BoundMW(r,t j )In this scenario, this resource can not ramp down into its maximum limit at t j . MaxLimit is adjusted to the lowes...

Stochastic Look-Ahead Commitment: A Case Study in MISO

“…For a given time period, SLAC will calculate an optimal solution (e.g., commitment and dispatches) over the study period that maximizes the expected market surplus over all the operational scenarios it considers. References [1,2] introduced how MISO operations currenlty manage uncertainties through a multi-stage commitment process, reserve products, "headroom" and multiple scenarios. LAC runs every 15 minutes with three hours look ahead, and is the last stage of the commitment process.…”