Technical Conference: Increasing Real-Time and Day-Ahead Market Efficiency through Improved Software

The California ISO has taken several measures to incorporate the uncertainty due to Wind and Solar variability into the unit commitment and dispatch algorithms used in market operations. A hybrid approach that recognizes the need to incorporate uncertainties in operation of Day-ahead and real-time markets is presented. The hybrid approach is based on leveraging different forecasting techniques related to supply, demand, system flexibility, and deliverability requirements for both day-ahead and real-time applications. Incorporation of different requirement uncertainties into specific optimization constraints is described. Many research efforts have been conducted on stochastic unit commitment (" UC ") and robust optimization because of the uncertainty introduced by renewable integration. MISO is collaborating with Alstom and University of Florida to explore the possibility of applying robust and/or stochastic optimization based unit commitment to address real world operational issues caused by uncertainty. The project team developed a robust optimization based prototype on MISO Look-Ahead Security Constrained Unit Commitment (LAC) and configured the optimization model based on the discussion with MISO operations. LAC is used by MISO operators primarily to make decisions on the commitment of fast-start resources. The team is also implementing recently proposed " Unified Stochastic and Robust Unit Commitment " approach. The next step is to expand the prototype to Reliability Assessment Commitment so that the commitment of slow-start resources can ensure enough on-line capacity and fast-start resources available to cover the uncertainties in real time. This presentation will describe the NYISO's experience replacing the legacy Lagrangian Relaxation (LR) with a Mixed Integer Programming unit commitment. We will describe the driving forces and technical hurdles associated with the migration while highlighting risks and rewards the proof of concept (POC) allowed us to take. Details of this presentation include; describing a novel approach to offloading computationally intensive workloads to a cluster of high performance Linux machines, various iterations of POC tests and feedback improving the next POC, added benefits to the NYISO market efficiency, and technology decisions which will allow more open and low cost exploration of advanced MIP, complex modeling and solution methodologies. The California ISO is in the process of enhancing the contingency model in its market optimization to handle the post contingency corrective actions. With this enhancement the market optimization advances from a pure preventive mode to a preventive-corrective mode, where both pre contingency dispatches and post contingency re-dispatches are co-optimized to meet the reliability standards. With the mandatory system …