Incentive Programs, Optimization, Linear Programming, Peak Load Contribution
Increasing efforts have been dedicated recently towards the development of advanced system controls to optimize Central Energy Facility (CEF) operations in order to reduce energy consumption, and, consequently, energy cost. Reduction of electric consumption is beneficial for both customers and the Regional Transmission Organization (RTO) managing the power grid. Therefore, RTOs have setup Incentive-Based Demand Response (IBDR) programs, such as Economic Load Demand Response (ELDR), and Price-Based Demand Response (PBDR), such as Peak Load Contribution (PLC), to incentivize customers to lower their electric consumption or shift their electric loads. These strategies also reduce the need for the RTO to commission additional or even invest in new power plants during peak hours. The ELDR program allows customers to choose when and by how much to curtail their electric consumption in response to market prices. The customer is then compensated for the amount of power curtailed at the real-time Locational Marginal Prices (LMP). PLC charge, which prompts customers to shave or shift their peak load consumption, is a demand charge structure based on a customer’s contribution to the demand peaks which occur in a region or a zone managed by an RTO at certain hours over a base period. Charges associated with PLC are significant and a customer is billed, in addition to the regular energy consumption and demand charges, a monthly charge during the billing period, based on their PLC during the base period in the prior year. Given the diversity of assets within a CEF, the challenge becomes how to efficiently run the facility and allocate assets while responding to market prices in the IBDR programs and minimizing cost due to PLC charges. In this work, a hierarchal approach for optimizing CEF operations with integrated IBDR and PBDR programs is developed. The approach is focused on the ELDR program and the PLC charge structure in the Pennsylvania, Jersey, Maryland (PJM) RTO region. However, it can be extended to accommodate other programs in different regions. Given predicted CEF loads, day-ahead and/or real-time LMP, PLC charges, and energy rates, the optimization problem is solved over a horizon into the future using a linear programming framework. Since PLC Coincidental Peaks (CP) are not known in advance, the optimization problem uses an hourly mask of projected CP hours, which can be either entered by the user or predicted based on the status of the region. The developed approach allows for an optimal allocation of assets to guarantee the curtailment commitment in the ELDR program, in addition to minimizing the customer’s PLC during projected CP hours. Furthermore, it is adaptive as it updates asset allocation based on feedback from the ELDR market and any changes in the projected CP hours. In this paper, a case study of the implementation of the developed approach at Kent State University (KSU) is presented, which shows the validity of the proposed solution.