Pricing in markets with non-convexities with reference to electricity markets
Speaker: Dr. Panagiotis Andrianesis (Boston University)
Abstract: The problem of pricing in markets with non-convexities lies at the interface of economics, operations research and market design. It has attracted renewed interest in the context of electricity markets, most notably day-ahead wholesale electricity markets where unit commitment costs and capacity constraints give rise to non-convexities. To address this problem, various pricing schemes that lift the price above marginal cost and/or provide side-payments (uplifts) have been proposed in the literature. We critically review several pricing schemes, employing exact analysis for a two-supplier setting, with asymmetric capacities and asymmetric marginal and fixed costs. We derive closed-form expressions for the price, uplifts, and profits that each scheme generates that enable us to analytically compare these schemes along these three dimensions. Our analysis identifies trade-offs between the market outcome characteristics that are weighed differently by each scheme. Further, we extend analytical comparisons to the case of more than two suppliers, we discuss the case of price-elastic demand, and we add to extant numerical comparisons.
Bio: Panagiotis Andrianesis is a graduate of the Hellenic Army Academy, also holding a Bachelor degree in Economics (2004) from the National and Kapodistrian University of Athens, and a Diploma degree in Electrical and Computer Engineering (2010) from the National Technical University of Athens. For the work he carried out in the context of his Diploma Thesis, he was awarded the IEEE Antennas and Propagation Society Pre-Doctoral Research Award. In 2011, he received his M.Sc. degree in Production Management and Industrial Administration from the University of Thessaly. He continued his studies towards a Ph.D. degree in the area of design and analysis of electricity market mechanisms, which he earned in 2016. He has been a consultant and research associate of ECCO International Inc. for more than 5 years. Currently, he is a postdoctoral associate in the Division of Systems Engineering, at Boston University, also affiliated with the Information and Data Science (IDS) Research Group, Electrical and Computer Engineering. His research interests include power system economics, market design and analysis, optimization, distributed algorithms, and applied mathematics.
Tight-and-Cheap Conic Relaxations for the AC Optimal Power Flow and the Optimal Reactive Power Dispatch Problems
Speaker: Miguel Anjos (Polytechnique Montreal)
Abstract: The classical alternating current optimal power flow problem is highly nonconvex and generally hard to solve. Convex relaxations, in particular semidefinite, second-order cone, convex quadratic, and linear relaxations, have recently attracted significant interest. The semidefinite relaxation is the strongest among them and is exact for many cases. However, the computational efficiency for solving large-scale semidefinite optimization is lower than for second-order cone optimization. We propose a new conic relaxation obtained by combining semidefinite optimization with the reformulation-linearization technique, commonly known as RLT. The proposed relaxation is shown to be stronger than the second-order cone relaxation and nearly as tight as the standard semidefinite relaxation on standard test cases with up to 6515 buses, with the time to solve the new conic relaxation being up to one order of magnitude lower than for the chordal approach to solve the standard semidefinite relaxation. An extension to optimal reactive power dispatch will also be presented. This is joint work with C. Bingane and S. Le Digabel.
Bio: Miguel F. Anjos is Full Professor in the Department of Mathematics and Industrial Engineering of Polytechnique Montreal, where he holds the NSERC-Hydro-Quebec-Schneider Electric Industrial Research Chair on Optimization for the Smart Grid, and the Inria International Chair on Power Peak Minimization for the Smart Grid. He received the B.Sc. degree from McGill University, the M.S. from Stanford University, and the Ph.D. degree from the University of Waterloo, and is a Licensed Professional Engineer in Ontario, Canada. His research interests are in the theory, algorithms and applications of mathematical optimization. He is particularly interested in the application of optimization to problems in power systems management and smart grid design. He is the Founding Academic Director of the Trottier Institute for Energy at Polytechnique, which he led from its inauguration in May 2013 until August 2016. Under his leadership, the Institute published several White Papers on the Canadian energy landscape. He is a former Editor-in-Chief of Optimization and Engineering, and serves on several other editorial boards. He was elected to three-year terms on the Council of the Mathematical Optimization Society and as Program Director for the SIAM Activity Group on Optimization, and to a two-year term as Vice-Chair of the INFORMS Optimization Society. He has served on the Mitacs Research Council since its creation in 2011. His allocades include a Canada Research Chair, the Méritas Teaching Award, a Humboldt Research Fellowship, the title of EUROPT Fellow, and the Queen Elizabeth II Diamond Jubilee Medal. He is a Senior Member of IEEE and a Fellow of the Canadian Academy of Engineering.
Tight Piecewise Convex Relaxations for Global Optimization of AC Optimal Power Flow
Speaker: Russell Bent (LANL)
Abstract: Since the alternating current optimal power flow (ACOPF) problem was introduced in 1962, developing efficient solution algorithms for the problem has been an active field of research. In recent years, there has been increasing interest in solutions based on convex relaxations to the ACOPF. In this talk, we will discuss how the convex, quadratic (QC) relaxation of the ACOPF is strengthened through extreme point formulations and piecewise, convex functions. We combine these relaxations with bound tightening and an adaptive, multivariate partitioning algorithm to progressively tighten these relaxations. Computational results show that this approach reduces the best-known optimality gaps for some hard ACOPF cases.
This is joint work with Kaarthik Sundar, Harsha Nagarajan, Carleton Coffrin, Sidhant Misra, and Mowen Lu.
Bio: Russell Bent received his Ph.D. degree in computer science from Brown University in 2005. Since then he has been a scientist at Los Alamos National Laboratory, NM, USA. He is currently in the Applied Mathematics and Plasma Physics Group (T-5), where he leads LANL’s inter-organizational Advanced Network Sciences Initiative (ANSI)—https://lanl-ansi.github.io/. ANSI is an interdisciplinary initiative that enables fundamental and applied research to address long-term challenges in critical infrastructure design, operation, and security. The primary philosophy of ANSI is that combining insights from Theoretical Physics, Applied Mathematics, Computer Science, and Applied Engineering can result in novel computational methods that address a variety of emerging challenges in infrastructure networks. Dr. Bent is the principal or co-principal investigator for numerous DOE projects in infrastructures systems with focuses on improving robustness of infrastructure systems to extreme events, increasing resilience of distribution networks, modeling interdependencies between systems, managing disasters that impact critical infrastructure, modeling smart grid technologies, and developing methods for mixed-integer, non-linear optimization. He is the lead developer for the software POD, A Global Solver for Nonconvex MINLPS (https://github.com/lanl-ansi/POD.jl) and the software GasModels.jl, a toolbox for modeling natural gas systems (https://github.com/lanl-ansi/GasModels.jl) He is the author of one book, Online Stochastic Combinatorial Optimization, and over 80 peer reviewed journal and conference publications.
A feedback approach to real-time power system operation
Speaker: Saverio Bolognani (ETH)
Abstract: We propose a unified control approach to the real-time operation of a power system. By formalizing relevant ancillary services as an optimization problem on the manifold of the grid steady states, we can employ a feedback scheme to steer the power system towards safe and efficient working points that satisfy the grid operational constraints. Stability of the interconnection between power grid dynamics and the flow induced by the proposed optimization scheme can be guaranteed via an ad-hoc singular perturbation analysis of the closed-loop system. Preliminary simulations show how the proposed approach can be used to provide essential services like secondary frequency control, economic re-dispatch, voltage regulation, and line congestion control.
Bio: Saverio Bolognani received his Ph.D. degree in Information Engineering from the University of Padova, Italy, in 2011. He was a visiting graduate student at the University of California at San Diego, and later a Postdoctoral Associate at the Laboratory for Information and Decision Systems of the Massachusetts Institute of Technology in Cambridge (MA). He is currently a Senior Researcher at the Automatic Control Laboratory at ETH Zurich, Switzerland. His research interests include the application of networked control system theory to power systems, distributed control and optimization, and cyber-physical systems.
Development of Market Clearing Software for Future Industry Challenges
Speaker: Yonghong Chen (MISO)
Abstract: The core of the electricity market clearing software is large scale security constrained unit commitment and security constrained economic dispatch problems. Through the development of advanced modelling and the application of state of the art optimization solvers, MISO has successfully maintained consistent market clearing system performance. This presentation discusses the R&D initiatives at MISO to prepare market clearing software for future market enhancements. These initiatives cover areas of advanced resource modeling, advanced mathematical formulation to improve computational performance as well as price efficiency, computational research on better interacting with existing commercial solvers and the development of high performance computing based next generation optimization engine under the ARPA-E HIPPO project.
Bio: Dr. Yonghong Chen is a Consulting Advisor at MISO Market R&D. She has played key roles in the startup and development of MISO markets in the past 16 years. She was in the winning team of 2011 INFORMS Franz Edelman Award for Achievement in Operations Research and Management Science. She is currently leading several initiatives at MISO on computational research and future market design.
Optimization for Resilient Transmission/Distribution Systems
Speaker: Santiago Grijalva (Georgia Tech)
Abstract: Smart and sustainable electricity systems require the deployment of significant amounts of distributed energy resources (DERs) such as solar PV, wind, energy storage, and demand response. Besides being a clean source of energy and having costs that continue to decrease, DERs exhibit properties that can contribute to higher grid resilience. The integration of large amounts of DERs including resilience considerations present several optimization challenges for operations and planning.
This talk will describe recent developments on large-scale integration of DER including optimal DER scheduling on AC 3-phase unbalanced distribution systems, integrated transmission and distribution (T/D) optimization, incorporation of grid resilience modeling against weather events, and extensions to optimal grid design.
Adaptive Building Load Control to Enable High Penetration of PV in Residential Neighborhoods
Speaker: Dong Jin (Oak Ridge National Lab)
Abstract: The high variability in solar photovoltaic (PV) power production causes voltage variations temporally and spatially on distribution feeders and substations. To tackle this problem, we propose absorbing most of the PV power generation locally by building loads such as heating, ventilation and air conditioning (HVAC) units to minimize the impact on the grid and reduce the need for large energy storage devices. On/off HVAC units are widely available in residential buildings in addition to many small to medium size commercial buildings. We formulate a mixed integer optimization problem to optimally dispatch an aggregate of on/off HVAC units to consume most of the PV power generation while maintaining occupants comfort and hardware constraints. Numerical results show that by assigning the proper number of aggregated HVAC units, the proposed mechanism achieves good PV tracking performance without jeopardizing occupants comfort. To deal with the uncertainty of solar PV generation, a distributionally robust chance constrained (DRCC) model was formulated to ensure that PV generation is consumed with a desired probability for a family of probability distributions. Furthermore, an Open-DSS based simulation platform is used to investigate the voltage impact on the distribution feeder of increased penetration of PV systems while applying the proposed optimal scheduling mechanism.
Bio: Dr. Jin Dong is a member of ORNL’s Building Envelope & Urban Systems Research Group in the Energy and Transportation Science Division. He received his PhD degree in Electrical Engineering from the University of Tennessee, Knoxville in 2016. Since joining ORNL in 2016, his work has focused mainly on control-oriented building energy modelling and new optimal control algorithms for improving the energy efficiency of buildings as well as actively engaging them in response to grid needs. This is especially important in power systems with high penetrations of wind and solar. He has been involved in the development of transactive building load control, adaptive building modelling and virtual storage devices projects sponsored by the DOE Solar Office and Building Technology Office (BTO). In his work, he uses methods from a variety of fields including controls, optimization, statistics, and machine learning. Towards the goal of building intelligent Grid-Interactive Efficient Buildings (GEB), behave and interact with the power grid as any distributed energy resources (DERs) do, his research interests lie in the intersection of smart buildings, building-to-grid integration, optimal control and optimization, multi-agent system and game theory, machine learning and artificial intelligence.
Failure Localization via Tree Partition
Speaker: Steven Low (Caltech)
Abstract: Cascading failure in power systems propagates non-locally, making the control and mitigation of outages hard. In this work, we characterize line failure localizability on transmission networks described by DC power flow equations. Specifically, we show that a network can be uniquely partition into subgraphs, we call cells. Two cells are connected either by a cut edge or by a cut vertex. If a line within a cell is tripped, the impact of this failure is contained within this cell. If a line connecting two cells is tripped, on the other hand, the impact propagates globally across the network, affecting the power flow on all remaining transmission lines. This characterization suggests that it is possible to improve the system robustness by switching off certain transmission lines, so as to create more, smaller cells, thus localizing line failures and making the grid less vulnerable to large-scale outages.
Joint work with Daniel Guo, Chen Liang, Alessandro Zocca, and Adam Wierman
Bio: Steven Low is the Gilloon Professor of the Department of Computing & Mathematical Sciences and the Department of Electrical Engineering at Caltech, and holds Honorary Professorship in Australia and China. Before that, he was with AT&T Bell Laboratories, Murray Hill, NJ, and the University of Melbourne, Australia. He was a co-recipient of IEEE best paper awards and is an IEEE Fellow. His research on communication networks has been accelerating more than 1TB of Internet traffic every second since 2014. He was a member of the Networking and Information Technology Technical Advisory Group for the US President’s Council of Advisors on Science and Technology (PCAST) in 2006. He received his B.S. from Cornell and PhD from Berkeley, both in EE.
Model predictive control framework for congestion management with large batteries in subtransmission grid.
Speaker: Jean Maeght (RTE, France)
Abstract: RTE will build and put into operation 3 large battery storage systems in 2020 (12MW/24MWh). These batteries, together with intermittent renewable generation curtailment, will be used to manage congestions in 3 small subtransmission zones (63kV or 90kV). A local controller will send orders to the battery and to power plants every 5 seconds, using all the flexibility offered by permanent and emergency ratings. This local controller will not have any forecast and will not be able to manage preventive actions, so a higher level scheduler will be in charge of security analysis (N-1 analysis), battery preventive actions, pre-discharging the battery for forthcoming congestions. Morever, this higher level scheduler will be in charge of computation of capacity tunnels; these capacity tunnels will to share the use of the batteries with other services when there are no congestions.
Bio: Jean Maeght received the M.S. degree in mathematics in 1997 from both Orsay (now Paris Saclay) and Toulouse Universities, France. In 2000, he received the Ph.D. degree in mathematical optimization applied to medical imaging from Toulouse University. He has been with Artelys French company for 8 years, as a Consultant specializing in applications of optimization to various industrial and economic fields. In 2008 he joined the French Transmission System Operator RTE (Reseau de Transport d’Electricite) in the R&D Department
He is focusing on optimization methods for power systems: coordination, optimal power flow, operational planning, grid development, economical studies. Jean has been involved in many European Research Projects (Twenties, iTesla, e-Highway2050, Osmose, POEMA) and bilateral cooperation projects with academic labs. As in France PhD students may work on their PhD directly in companies, Jean is supervising several PhD students within RTE, currently on Global Optimality for OPF and on Battery Management algorithms.
Real Time Voltage Control of Active Distribution Systems
Speaker: Sakis Meliopoulos (Georgia Tech)
Abstract: Active distribution systems with distributed energy resources and microgrid clusters exhibit voltage issues. In many cases, voltage swings well above permissible levels have been observed. These voltage irregularities create abnormal power factors at the distribution/transmission interface and propagate the issues to the bulk power system. As larger percentages of distribution systems become active, these phenomena will affect the operation and voltage stability of the bulk power system.
Voltage issues in active distribution systems can be solved via real time control in a way that the active distributed resources in distribution can assist voltage control and stability on the transmission system. We describe an automated system for monitoring an active distribution system and all the distributed energy resources via synchronized measurements and a distributed state estimator. This system provides the real time model of the active distribution feeder at speeds up to once per cycle. The real time model is used to automatically set up a multi-period optimal power flow whose solution provides: (a) controls to levelize the voltage profile along the active distribution system and (b) a reduced dynamic model of the feeder with limits of available reactive power at the head of the distribution circuit. The last reduced model is incorporate into transmission level Voltage/VAr control for the operation of the bulk power system and for characterizing its voltage stability. The proposed system makes the entire active distribution feeder an asset for the bulk power system.
The overall system will be demonstrated with power hardware-in-the-loop simulations of feeder models from National Grid (NG) and the Public Service Company of New Mexico (PNM), and a field demonstration at NG.”
Optimal Adaptive Approximations of the Power Flow Equations
Speaker: Daniel Molzahn (Argonne National Laboratory, Georgia Tech)
Abstract: The power flow equations model the relationship between voltages phasors and power flows and are therefore at the heart of many optimization and control problems relevant to electric power systems. The nonlinearity of the power flow equations results in a variety of algorithmic and theoretical challenges. Accordingly, solution algorithms for power system optimization and control problems often rely on linearizations of the power flow equations. Typical linearizations are computed using “one-size-fits-all” approaches that do not exploit knowledge of the system parameters or forecasts of load demand and renewable generation. In contrast, this presentation describes algorithms for computing “adaptive linearizations” that are specifically tailored to a given power system and operating range of interest. These adaptive linearizations are constructed in order to minimize a selected error metric relative to the actual nonlinear power flow equations over the relevant operating range. Numerical results demonstrate that the adaptive linearizations can result in significantly reduced errors relative to typical linearization techniques.
This is joint work with Dr. Sidhant Misra (Los Alamos National Laboratory).
Bio: Daniel Molzahn is a computational engineer at Argonne National Laboratory in the Center for Energy, Environmental, and Economic Systems Analysis (CEEESA). In January 2019, Daniel will be an assistant professor in the Electrical and Computer Engineering Department at the Georgia Institute of Technology. Daniel was a Dow Sustainability Fellow at the University of Michigan. He received the B.S., M.S. and Ph.D. degrees in Electrical Engineering and the Masters of Public Affairs degree from the University of Wisconsin–Madison, where he was a National Science Foundation Graduate Research Fellow. His research interests are in the development of optimization and control techniques for electric power systems.
Distributionally Robust Stochastic Dual Dynamic Programming
Speaker: David Morton (Northwestern)
Abstract: We consider a multi-stage stochastic linear program for hydroelectric scheduling under inflow uncertainty. That model lends itself to solution by stochastic dual dynamic programming. In this context, we focus on a distributionally robust variant of the model, inspired by work of Philpott, de Matos, and Kapelevich (2017). Here, the specific realizations in each stage are fixed, and distributional robustness is with respect to the probability mass function governing those realizations. We describe a computationally tractable variant of SDDP to handle thismodel.
Bio: David Morton is the David A. and Karen Richards Sachs Professor and Department Chair of Industrial Engineering and Management Sciences at Northwestern University. He was a Fulbright Research Scholar at Charles University in Prague, and was a National Research Council Postdoctoral Fellow in the Operations Research Department at the Naval Postgraduate School. He also currently directs Northwestern’s Center for Optimization and Statistical Learning http://osl.northwestern.edu.
Coauthor: Daniel Duque
Evaluating the (Computational) State of the Art in Unit Commitment Formulations
Speaker: James Ostrowski (U. Tennessee)
Abstract: This work is a comprehensive overview of mixed integer programming formulations for the unit commitment problem (UC).This has been an especially active area of research the past twelve years, and as a result, the research community has a much better understanding of the underlying structure of the UC problem. However, this understand does not necessarily result in improved solution times. To better understand the relationship between the theoretical quality of the formulation and the result computational performance, we implemented and exhaustively tested much of what is found in the literature using UC instances drawn from both academic and real-world data.
Coauthors: This is joint work with Bernard Knueven and JP Watson, both at Sandia National Laboratory
A Unified Framework for Sequential Decision Analytics in Energy Systems
Speaker: Warren B. Powell (Princeton University)
Abstract: Energy systems offer a variety of forms of uncertainty that have to be accommodated to ensure a reliable source of power. The modeling of these sequential decision problems under uncertainty has lacked the kind of canonical framework long enjoyed by deterministic problems. I will introduce a modeling framework that is completely general, which involves searching over classes of policies rather than deterministic vectors. I then describe two fundamental strategies for creating policies, each of which further divides into two subclasses, creating four classes of policies. We claim that these four classes of policies are universal, in that any method used to solve a sequential decision problem will be drawn from this set.
The four classes of policies are illustrated in the context of several applications in energy systems. An energy storage application is used to demonstrate that each of the four classes of policies might be best depending on the characteristics of the data.
We then transition to a discussion of uncertainty quantification in an energy context, where we focus on the replication of crossing times in the modeling of wind and prices.
A convergent distributed algorithm for nonconvex nonsmooth optimization with applications in optimal power scheduling
Speaker: Andy Sun (Georgia Tech)
Abstract: We propose a new distributed algorithmic framework for general nonconvex nonsmooth constrained optimization problems. The new algorithm overcomes some intrinsic limitation of the ADMM type algorithms for nonconvex constrained problems by creating a new reformulation and a two-level structure. We prove the global and local convergence of the algorithm. Finally, we provide promising computation results for the AC OPF problem. This is one of the first algorithms with convergence guarantee for a very general class of nonconvex nonsmooth constrained programs.
This is joint work with Kaizhao Sun (Georgia Tech)
Bio: Dr. Sun obtained his PhD degree in Operations Research from the MIT Operations Research Center. He has a broad interest in optimization under uncertainty and nonconvex optimization, with applications in energy systems. His research has won several awards including the NSF CAREER award, the INFORMS best publication in energy, and the George B. Dantzig Dissertation prize.
A mixed-integer optimization model for electricity infrastructure development
Speaker: Valerie Thomas (Georgia Tech)
Abstract: We formulate a mixed-integer program to analyze the decision between centralized and decentralized technologies for new energy infrastructure development. The formulation minimizes the cost of meeting both average and peak power demand in each specified demand node. We demonstrate our methodology with a case study of Rwanda, accounting for existing generation and transmission infrastructure. Thirteen ongoing or proposed projects are considered as potential new centralized generation facilities and the decentralized technology is modeled after a small (∼50 W) solar home system. The case study is repeated using population data at four different resolutions while varying demand levels and decentralized technology cost. A tipping point effect is observed, where the optimal infrastructure tips from being primarily centralized to primarily decentralized under certain combinations of the demand and cost parameters.
Bio: Valerie M. Thomas is the Anderson Interface Professor of Natural Systems at the Georgia Institute of Technology, with appointments in the H. Milton Stewart School of Industrial and Systems Engineering and in the School of Public Policy. Immediately prior to coming to Georgia Tech, Thomas was the 2004-05 APS Congressional Science Fellow. Previously she worked at Princeton University at the Center for Energy and Environmental Studies, the Princeton Environmental Institute, and at the Woodrow Wilson School of Public and International Affairs; and at the Department of Engineering and Public Policy at Carnegie Mellon University. Thomas is Associate Editor of the Journal of Industrial Ecology, board member of the Southeast Energy Efficiency Alliance, and a member of the US DOE/USDA Biomass Research R&D Technical Advisory Committee. She was a Member of the US EPA Chartered Science Advisory Board from 2003 to 2009. Thomas’ research is on the environmental impacts and costs of energy systems, the environmental impacts of biofuels and other products and services, and the effects of policies and technologies on the development of energy systems. Thomas has a PhD in high energy physics from Cornell University and a BA in physics from Swarthmore College. Her Ph.D. thesis work was on the catalysis of proton decay in grand unified theories, and her post-doctoral research was on the verification of nuclear arms control treaties. She was a participant in 1989 Black Sea Experiment on the detection of nuclear warheads, and was one of the founders of the International Summer Symposium on Science and World Affairs, now in its 26th year. She has more than 80 technical publications spanning energy, environment, optimization, physics, and nuclear arms control. At Georgia Tech, Thomas teaches a graduate course in Energy Technology and Policy, and undergraduate courses in Energy, Efficiency and Sustainability, Engineering Economics, and Senior Design.
Coauthors: Yuang Chen, Todd Levin
Unit Commitment with Gas Network Awareness
Speaker: Pascal Van Hentenryck (Georgia Tech)
Abstract: This talk introduces the UC problem with Gas Price Awareness (UCGPA), which schedules a set of generating units for the following day while taking account the physical and economic feedback from the associated natural gas system. The UCGPA estimates the natural gas prices using dual solutions associated with the market-clearing constraints of the gas market, and checks the validity of submitted bids such that all committed bids are profitable with respect to the natural gas price estimation. Experimental results on instances modeling the New England polar vortex events demonstrate the benefits of incorporating gas awareness into the unit commitment decisions.
Bio: Pascal Van Hentenryck is the A. Russell Chandler III Chair and Professor in the H. Milton Stewart School of Industrial and Systems Engineering at the Georgia Institute of Technology. His current research is focusing artificial intelligence, data science, and operations research with applications in energy systems, mobility, and privacy.
Transmission capacity allocation in zonal electricity markets
Speaker: Anthony Papavasiliou (UCL)
Abstract: We propose a novel framework for modeling zonal electricity markets, based on projecting the constraints of the nodal network onto the space of the zonal aggregation of the network. The framework avoids circular definitions and discretionary parameters, which are recurrent in the implementation and study of zonal markets. Using this framework, we model and analyze two zonal market designs currently present in Europe: flow-based market coupling (FBMC) and available-transfer-capacity market coupling (ATCMC). We develop cutting-plane algorithms for simulating FBMC and ATCMC while accounting for robustness of imports/exports to single element failures, and we conduct numerical simulations of FBMC and ATCMC for a realistic instance of the Central Western European system under 768,000 different operating conditions. We find that FBMC and ATCMC are unable to anticipate congestion of branches interconnecting zones and branches within zones, and that both zonal designs achieve similar overall cost efficiencies (0.5% difference in favor of FBMC), while a nodal market design largely outperforms both of them (5.9% better than FBMC). These findings raise the question of whether it is worth for more European countries to switch from ATCMC to FBMC, instead of advancing directly towards a nodal design.
Bio: Anthony Papavasiliou is an associate professor at the Université catholique de Louvain in Belgium and a member of the Center for Operations Research and Econometrics (CORE). Anthony holds the ENGIE Chair and the Francqui Research Professorship at UCL. Anthony has received a PhD from the department of Industrial Engineering and Operations Research at UC Berkeley, and a BSc in Electrical and Computer Engineering from the National Technical University of Athens, Greece. He currently serves as an associate editor for Operations Research and IEEE Transactions on Power Systems.
Stochastic Multi-Period Economic Dispatch at Industrial Scale
Speaker: Jean-Paul Watson (Sandia)
Abstract: We consider the problem of scalable security-constrained economic dispatch, in which a PTDF network representation is utilized and uncertainty in generator contingencies, transmission contingencies, and renewables production is considered. Decomposition is performed via progressive hedging, executed in a modest-scale parallel computing environment. Our base optimization model is that of the PJM Independent System Operator in the US and contains full-fidelity representations of the full range of key ancillary service products. We demonstrate that progressive hedging is able to solve this stochastic, industrial-scale optimization model in operationally relevant time scales. Finally, we analyze the key differences between stochastic and deterministic economic dispatch solutions.
Bio: Dr. Jean-Paul Watson is a Distinguished Member of Technical Staff in the Data Science and Cyber Analytics Department at Sandia National Laboratories, in Livermore, California. He has over 15 years of experience applying and analyzing algorithms for solving difficult combinatorial optimization and informatics problems, in fields ranging from logistics and infrastructure security to power systems and computational chemistry. His research currently focuses on methods for approximating the solution of large -scale deterministic and stochastic mixed-integer and non-linear programs, with applications in the domain of electricity grid operations, planning, and resiliency.
Distributionally Robust Optimization on Power System Resilience
Speaker: Chaoyue Zhao (Oklahoma State)
Abstract: The power grid disruptions caused by extreme weather, although rare, can bring catastrophic impacts to the power industry and the society in general. The evaluation and mitigation of disruption-related risks and impacts are often computationally prohibitive due to the complexity of the power system, uncertainty of weather conditions, and the combinatorial nature of component failures. In this talk, we propose a distributionally robust optimization model to assist power system operations and enhance the power system resilience in face of natural disasters. In our approach, we consider a case where the true probability distribution of system component failures is ambiguous, i.e., difficult to accurately estimate. Instead of assigning a (fixed) probability estimate for each failure scenario, we consider a set of probability distributions (termed the ambiguity set) based on the N-k security criterion and moment information. Our approach considers all possible distributions in the ambiguity set, and is hence distributionally robust. Meanwhile, as this approach utilizes moment information, it can benefit from available data and become less conservative than the robust optimization approaches.
Bio: Dr. Chaoyue Zhao is currently an assistant professor in Department of Industrial Engineering and Management in Oklahoma State University. She received her Ph.D. in Industrial and Systems Engineering from the University of Florida in 2014. Before that, she obtained her B.S. degree in Department of Mathematics from the Fudan University, China, in 2010. She has collaborated with Sandia and Argonne National Labs for her research and worked at Pacific Gas & Electric Company. Her research interests include data-driven stochastic optimization and robust optimization with their applications on power system scheduling and resilience. Her research is funded by multiple federal agencies such as National Science Foundation, Department of Transportation, and Argonne National Laboratory.
Grid Resilience and Its Market Impact
Speaker: Tongxin Zheng (ISO-New England)
Abstract: Grid resilience is commonly referred to a system’s ability to quickly recover from high-impact low-frequency events, and is getting more and more attention from stakeholders as well as policy makers such as DOE, FERC and State Commissions. FERC has requested each ISO/RTO region to provide comments on the definition, metrics and mitigation measures of grid resilience. Improving grid resilience can create wide-spread impact on ISO/RTO, transmission, distribution and generation companies as well as consumers. This talk will focus on the grid resilience’s definition, and its market solutions as well as the computational needs.
Bio: Tongxin Zheng is currently the Technical Director at ISO New England. He manages research and development projects for the regional wholesale electricity market, and research collaborations with the research community. He provides technical consultation on market and system operations to the senior management, and oversees the development of the market clearing engine and the market simulation software. His research interests are power system operation, optimization and electricity market.