Informs Annual Meeting Phoenix 2018

INFORMS Phoenix – 2018

TB44

3 - Optimal Energy and Regulation Market Bidding Strategies for ESS-integrated Wind Power Producers Duehee Lee, KonKuk University, Kwangjin Gu, Seoul, Korea, Republic of, Seungwoo Son, Sini Han, Jae Hyung Roh We present the optimal bidding strategy for wind power producers (WPP)s in the day-ahead (DA) energy and regulation markets. WPPs integrate the energy storage system (ESS) with wind farms to mitigate the imbalance penalty in the real-time (RT) market and to earn extra profits in the regulation market. Our bidding strategies in the DA market is designed and represented as offer curves by considering the optimal operation of the ESS for 24 hours and wind power and RT price forecasts. The ESS operation is determined by solving the multi-stage stochastic optimization problem through the progressive hedging algorithm. We show that our bidding strategy can have higher profits than a near-sighted strategy. Distributed Optimization for Power Systems Sponsored: Energy, Natural Res & the Environment/Electricity Sponsored Session Chair: Ross Baldick, University of Texas at Austin, Austin, TX, 78712, United States Co-Chair: Subhonmesh Bose, University of Illinois-Urbana Champaign, Urbana, IL, 61801, United States 1 - Market Mechanism Design for Horizontal and Vertical Coordination of Power System Expansion: Distributed Optimization Approach Sambuddha Chakrabarti, 2000 Pearl Street, Room 113, Austin, TX, 78705, United States In this work, we will consider long term generation & transmission expansion & investment coordination problems, where there are multiple Transmission System Planners (TSPs) & Generation Companies (GenCos). Each agent acts only to maximize its own utility. To attain the optimal social surplus for the bigger geographical region, we will use Distributed Stochastic Optimization algorithm to design a market mechanism. We will demonstrate our method with numerical simulations for the Nordic grid. 2 - Transmission Expansion Planning via Distributionally Robust Optimization Antonio J. Conejo, The Ohio State University, Department of Integrated Systems Engineering, 210 Baker Systems Building, Columbus, OH, 43210, United States This presentation addresses the transmission expansion planning problem under long- and short-term uncertainty. Long-term uncertainty pertains to changes across years, whereas short-term uncertainty pertains to changes within a year. This problem is formulated as a distributionally robust optimization model, and is solved via a tailored implementation of the primal Benders’ decomposition algorithm. The effectiveness of the proposed algorithm is illustrated through a realistic case study. 3 - Convex Relaxations of the Network Flow Problem Under Cycle Constraints with Application to Electric Power Systems Madi Zholbaryssov, University of Illinois at Urbana Champaign, Urbana, IL, 61801, United Statesu, Alejandro Dominguez-Garcia We consider a variation of the minimum-cost network flow problem (NFP) with additional non-convex cycle constraints on nodal variables; this problem has relevance in the context of optimizing power flows in electric power networks. We propose one approach to tackle the NFP that relies on solving a convex approximation of the problem, obtained by augmenting the cost function with an entropy-like term to relax the non-convex constraints. We show that the approximation error can be made small enough for practical use. An alternative approach is to solve the NFP without the cycle constraints and solve a separate optimization problem in order to recover the actual flows satisfying the cycle constraints. 4 - Distributed Optimization for Solving Nonconvex Optimal Power Flow Kaizhao Sun, Georgia Tech, Atlanta, GA, United States Motivated by the problem of coordination between ISO markets, we study distributed optimization algorithms for the AC optimal power flow problem. We first give an overview of the field of distributed optimization with an emphasis on applications to OPF. Then we propose distributed algorithms for solving the convexified as well as the nonconvex versions of the AC OPF problems with provable convergence guarantees. Numerical results show promising performance. n TB44 North Bldg 227C

5 - Robustness of Primal-dual Dynamics and its Implications for Online, Data-driven Optimization Jorge Cortes, PhD, University of California-San Diego, La Jolla, CA, 92093, United States Primal-dual strategies are used extensively in the design and analysis of distributed feedback controllers and optimization algorithms in many applications. In large-scale optimization, an aggregate objective along with the local computability of the constraints make the dynamics amenable to distributed implementation. Here, we discuss progress on establishing stability and robustness properties of primal-dual dynamics, paying special attention to input-to-state stability, and illustrate the implications for online optimization and data-driven implementations with performance guarantees. 6 - Prosumer-based Decentralized Power Supply Sleiman Mhanna, University of Sydney, School of EIE, EIE Building J03, The University of Sydney, 2006, Australia, Gregor Verbic, Archie Chapman Traditionally, most optimization models used in power systems are computed in a centralized fashion. However, the large increase in the penetration of distributed energy resources (DERs) on the low voltage side of the distribution network will put the centralized approaches under strain. The talk will focus on distributed algorithms for solving the problem of prosumer orchestration. The problem is computationally challenging as it explicitly considers the nonconvex AC power flow constraints and because the DER models require the use of mixed-integer variables to model them accurately. n TB45 North Bldg 228A Enhancing Power Grid Efficiency Through Mathematical Programming Sponsored: Energy, Natural Res & the Environment/Electricity Sponsored Session Chair: Yu Zhang, Santa Cruz, CA, 4592921, United States Co-Chair: Yihsu Chen, UC Santa Cruz 1 - Mechanism Design for Demand Response in Electricity Markets Sepehr Ramyar, University of California-Santa Cruz, Santa Cruz, CA, United States, Yihsu Chen Contrary to the current practice of trading demand response in wholesale power markets, we investigate the application of mechanism design theory as a new framework for designing a demand response market. The proposed mechanism is truth-revealing and would eliminate information asymmetries in the current market design. It also eases computational burden of the ISO, and guarantees participation of prospective agents. We show how the properties of the proposed mechanism can complement the existing regulatory context and eliminate opportunities for manipulation by market participants. 2 - Asynchronous Large-scale Decentralized Unit Commitment Paritosh Ramanan, Georgia Institute of Technology, Atlanta, GA, 30324, United States The unit commitment problem for power networks is a critical and a computationally challenging problem especially for large-scale power systems. We propose an asynchronous decentralized solution to the unit commitment problem that is computationally more efficient and is driven by privacy preserving valid inequalities. We benchmark our algorithm against a state of the art synchronous and the centralized method. In both cases, we demonstrate that our method is highly scalable, improves solution times and provides a competitive and stable solution quality. 3 - Optimal Load Shedding via Mixed-integer Bilinear Programming Yu Zhang, UC Santa Cruz, 1156 High St SOE2, Santa Cruz, CA, 95064, United States, Atif Maqsood, Keith Corzine In this work we propose a novel design of optimal load shedding schedules for power distribution networks with multiple load zones. A mixed-integer bilinear optimization problem is formulated, which aims at minimizing the system-wise load shedding cost while achieving a desired reliability for industrial and residential users. Numerical results corroborate the effectiveness and merits of the proposed approach.

290