Challenges of Planning and Pricing an Electric System
Speaker: Greg Roberts (Georgia Power)
Abstract: Georgia Power is a regulated utility which provides electric service to its 2.5 million customers. To provide this service, the company must plan an electric system that is capable of delivering clean, safe and reliable power every minute at a price that is affordable to all customers. The planning and pricing of our service provides numerous challenges in the engineering, analytical, economic and regulatory arenas.
Southern Power’s Renewable Hedging Program
Speaker: Jeff Baker (Southern Power)
Abstract: With over 46,000MW of electric generating capacity and 1,500 billion cubic feet of natural gas consumption, Southern Company is one of America’s premier energy company’s serving over 9 million customers. Southern Company provides clean, safe, reliable and affordable energy. It is known for excellent customer service, high reliability, and affordable pricing below the national average. Southern Power is Southern Company’s premier wholesale energy company. Southern Power invests in clean energy solutions for the customers it serves and consists of approximately 1,700MW if Solar capacity, 1,590MW of Wind capacity, and 9,300MW of Natural Gas capacity. Today we will discuss how Southern Power is using data analytics to optimize its renewable hedging program. Specifically, we will demonstrate how Southern Power’s Commercial Optimization and Trading organization has utilized data analytics and machine learning to optimize Day Ahead and Congestion trading across the SPP, ERCOT, and CAISO markets.
Biography: Jeff Baker is a Project Manager in Southern Power’s Commercial Optimization and Trading department. He joined Southern Company in 2007 as an Analyst in Southern Company’s Fleet Operations and Trading organization. Throughout his tenure at Southern Company Jeff has applied analytics, mathematical modeling, and optimization in projects ranging from Energy Trading, Unit Commitment, and Transmission Planning. Jeff attended Samford University where he received a BS in Mathematics, Georgia Tech where he received an MS in Applied Mathematics, and the University of Alabama at Birmingham where he received a PhD in Applied Mathematics.
Transmission, Monitoring, Diagnostics and Visualization & Generation Asset Fault Signature Identification
Speaker: Clifton Black (Southern Company)
Abstract: To capitalize on the wealth of existing data available within the Southern Company enterprise, Southern Company Research has embarked on an effort to develop the requirements for a Transmission Monitoring, Diagnostics, and Visualization (TMDV) application. The TMDV application will help system operators and engineers from other transmission functions define and implement mitigating measures to enhance system performance, efficiency, and reliability. There are significant challenges in applying analytics to provide actionable information to stakeholders across the transmission system. These include, but are not limited to, (1) the sheer volume of data to be analyzed (2) the significant variability of the grid usage on a daily, weekly, and yearly basis makes the basis for understanding what is “normal” versus “abnormal” a challenge for automated systems and (3) the challenge to acquire and integrate the variety of data types from disparate sources. TMDV will facilitate the integration of these data, many of which formerly existed in silos, and allow stakeholders to extract insights and information for decision support information in uniquely configured portals by end-users. Major research efforts are also underway, in the generation sector, to develop fault signatures for many generation type assets such as pumps, motors, combustion turbines, etc. The goal is to automatically diagnose specific fault types before failures occur. The talk will provide a summary of the TMDV and generation asset fault signature identification efforts.
Biography: Clifton Black joined Southern Company in 2006 and is currently a Principal Engineer in the Research and Development department. He is responsible for research in Grid Operations, Planning and Visualization. His technical focus areas include situational awareness, data analytics, power system analysis and resiliency. Clifton leads internal research efforts and manages collaborative partnerships with EPRI and other external organizations. He actively participates in various technical forums with papers and presentations at conferences in national and international venues. Clifton attended the University of Alabama (Tuscaloosa) where he received the BS, MS and PhD degrees in Electrical Engineering.
Computational Issues in ISO Market Models
Speaker: Richard P. O’Neill (FERC)
Abstract: The ISO market problem is to operate an efficient sustainable auction markets for energy and ancillary services that observe the physics within a fixed time window. The Nirvana solution to this problem is a stochastic dynamic unit commitment AC optimal power flow model. Today we are nowhere near solving this problem. Current approach is a day-ahead market and real-time market. Day-ahead market must be solved in three hours or less. It has 24 hourly periods that are settled financially and additional 12 period to avoid end effects. Offers are mitigated and units are efficiently committed and AC feasibility is checked iteratively. Real-time market must be solved in fifteen minutes or less. It has a look ahead of about four hours. Only the first period is settled financially. Offers are mitigated and a limited number of units are efficiently committed and AC feasibility is checked iteratively. We will address the areas of improvement and lines of research to improve these markets.
Biography: Richard P. O’Neill is the Chief Economic Advisor at the Federal Energy Regulatory Commission. From 1988 to 2000, he was the Chief Economist and Director of the Office of Economic Policy. From 1986 to 1988, he was the Director of the Commission’s Office of Pipeline and Producer Regulation. Led the Chairman’s teams to develop policy and restructure the natural gas market and the electric power market. He led the Chairman’s team to develop oil-pipeline rate index by benchmarking to actual industry costs. He led the Chairman’s program to increase the efficiency of the market software resulting in savings of at least $5 billion/year due to the introduction of mixed integer software. Led the group that developed the transmission switching software that when fully implemented has the potential to having of at least $10 billion/year. He created the initial design for the ARPA-E optimal power flow software prize.
From 1978 to 1986, he directed oil and gas analysis, including the development of software systems, oil and gas resource analysis, energy modeling systems, analysis of natural gas markets, and oil and gas forecasting at the Energy Information Administration.
From 1973 to 1978, he taught and did research in computer science and applied mathematics on the computer science and business faculty of Louisiana State University. From 1969 to 1973, he taught and did research in the areas of operations research and statistics on the business school faculty of the University of Maryland.
He has a B.S. in chemical engineering, an MBA and a Doctorate in operations research (with minors in mathematics, statistics, economics and accounting) all from the University of Maryland. He is a Fellow of INFORMS and in 2016 received the International Association of Energy Economics Outstanding Contributions to Energy Economics Award.
He has worked with many countries, states, the World Bank, energy companies and computer companies in the development of mathematical software, energy modeling, forecasting, regulation, privatization, restructuring and market design.
His published work has appeared in academic and professional journals and books in the areas of Applied Mathematics, Optimization, Operations Research, Management Science, Computer Science, Energy, Electrical Engineering, Economics, and Law with over 3000 Google citations.
Wholesale Electricity Market Pricing – Experiences and Challenges
Speaker: Tongxin Zheng (ISO New England)
Abstract: Market clearing price in the wholesale electricity market is commonly determined by solving a security constrained optimal power flow problem. It is generally constructed by constraint shadow prices using the marginal pricing concept, and supports the short-term market equilibrium. The integration of renewable resources, distributed energy resources and energy storage, however, introduced new complexities into the power system operation as well as electricity market design and operation. These complexities include uncertainty, nonconvexity and time-coupling. The current pricing scheme may not be efficient under this new environment. This talk will focus on ISO NE’s practice on the electricity market pricing, and discuss the pricing challenges of renewable integration. A multi-period multi-stage pricing and settlement scheme is also discussed as a potential solution to these challenges.
Biography: Tongxin Zheng is currently the Technical Director at Independent System Operator of New England Inc. (ISO-NE). He manages both research and development projects for the regional wholesale electricity market design and operations, and collaboration projects with the research community. He provides technical consultation to market and system operations, and oversees the development of the market clearing engine and the market simulation software. He received his Ph.D in electrical engineering from Clemson University.
Decentralized Coordination in Energy Prosumer Networks
Speaker: Santiago Grijalva (Georgia Institute of Technology)
Abstract:
Energy prosumers are economically motivated subsystems that can consume, produce or store electricity. They emerge naturally from the deployment of distributed energy resources (DERs) such as solar, wind, demand response, energy storage, electric vehicles, etc. Energy prosumers include homes, buildings, microgrids, as well as larger subsystems. Connected through communication networks, and equipped with distributed computing, prosumers become “energy aware” and can locally and intelligently optimize their energy utilization. However, in order to maintain the functionality, security, and resilience of the overall electricity grid, they must coordinate with each other in a decentralized and provable manner. This talk will describe architectures and methods for massively scalable decentralized coordination among energy prosumers at the control and optimization (scheduling) time scales. We will present novel results on algorithmic development and their performance on DER-rich networks, demonstrating scalability.
Biography: Santiago Grijalva is the Georgia Power Distinguished Professor of Electrical and Computer Engineering and Director of the Advanced Computational Electricity Systems (ACES) Laboratory at The Georgia Institute of Technology. He is a leading researcher on decentralized power system control, smart grids, energy internet, and future sustainable electricity systems. Dr. Grijalva is a Member of the Federal Smart Grid Advisory Committee of the National Institute of Standards and Technology (NIST). The Committee advises the US on Smart Grid Interoperability and Grid Modernization. He joined Georgia Tech in 2009 and he has been the principal investigator for research projects under Department of Energy (DOE), Advanced Research Projects Agency-Energy (ARPA-E), Electric Power Research Institute (EPRI), Power System Engineering Research Consortium (PSERC), National Science Foundation (NSF), several DOE National Laboratories, and numerous industry sponsors. He has published widely on power systems and smart grid. From 2013 to 2014, he served at the National Renewable Energy Laboratory (NREL) as founding Director of the Power System Engineering Center (PSEC), supervising about 50 staff and a $20 million/year research budget. From 2002 to 2009 he was with PowerWorld Corporation developing commercial grade software for transmission system planning and optimization. Dr. Grijalva has been the recipient of Fulbright, University of Illinois, and Organization of American States Fellowships, and recipient of several awards including the Georgia Tech ECE Outstanding Faculty Award, Great Minds in STEM National Achievement Award, and IBM Faculty Award. Dr. Grijalva’s MS and PhD degrees in Electrical and Computer Engineering are from the University of Illinois at Urbana-Champaign.
Operation Under Uncertainty: Forecasting and Optimization
Speaker: Lang Tong (Cornell University)
School of Electrical and Computer Engineering Cornell University, Ithaca, NY 14850
Abstract: A key challenge of operating interconnected power systems is to achieve efficiency and reliability under uncertainty. To this end, two fundamental problems arise. The first is the ability to forecast future operating conditions in the presence of renewable generation and stochastic load with behind-the-meter distributed resources. The second is optimizing system resources for operation and planning under such uncertainty.
In this talk, we first present a machine learning approach to probabilistic forecasting market operations. This technique provides estimates of joint probability distributions of power flows, locational marginal prices, and network congestion. We then present approaches to interchange scheduling that address some of the weaknesses of the current coordinate transaction scheduling (CTS) framework.
Biography: Lang Tong joined Cornell University in 1998 where he is now the Irwin and Joan Jacobs Professor in Engineering and the Cornell site director of the Power Systems Engineering Research Center (PSerc). He received the B.E. degree from Tsinghua University, Beijing, P.R. China in 1985, and PhD degree in EE from the University of Notre Dame, Notre Dame, Indiana in 1991. He was a Postdoctoral Research Affiliate at the Information Systems Laboratory, Stanford University in 1991.
His current research focuses on energy systems and smart power grid. In particular, his group investigates data analytics, system optimization, and market issues associated with renewable energy, storage, and the electrification of transportation systems. He is part of the Engineering and Economics of Electricity Research Group. Lang Tong is a Fellow of IEEE. He received the 2004 Best Paper Award (with Min Dong) from the IEEE Signal Processing Society, the 2004 Leonard G. Abraham Prize Paper Award from the IEEE Communications Society (with Parvathinathan Venkitasubramaniam and Srihari Adireddy), and the 1993 Outstanding Young Author Award from the IEEE Circuits and Systems Society. He is a coauthor of seven student paper awards, including two IEEE Signal Processing Society Young Author Best Paper Awards (Qing Zhao in 2000 and Animashree Anandkumar in 2008) for papers published in the IEEE Transactions on Signal Processing. He was named as a 2009 Distinguished Lecturer by the IEEE Signal Processing Society. He was the recipient of the 1996 Young Investigator Award from the Office of Naval Research.
Scenario-based Dispatch in Power Systems with Adaptable Level of Risks
Speaker: Le Xie (Texas A&M)
Abstract: A central challenge towards low carbon power system is the increasing level of uncertain resources that arise from both supply and demand sides. These uncertain resources may have a large range of variation, and are typically not well characterized in terms of their probability distribution. A theoretically rigorous and practically implementable approach to solving chance-constrained economic dispatch is introduced based on the recent progress of scenario approach-based optimization. This approach is shown to provide theoretical guarantee of the level of risks (i.e. maximum probability of violating the constraints in posterior test) based on the size of the historical samples used in the optimization. Furthermore, it offers the system operator an option of remove certain scenarios in exchange of quantifiable risk increment. This approach is critically appraised in view of other existing methods (robust, stochastic, and other chance-constrained dispatch methods). Joint work with Profs. M. C. Campi, S. Garatti, P. R. Kumar, and Mr. S. Moddaresi, X. Geng, and H. Ming.
Biography: Dr. Le Xie is an Associate Professor and Eugene Webb Faculty Fellow in the Department of Electrical and Computer Engineering at Texas A&M University. He received B.E. in Electrical Engineering from Tsinghua University in 2004, S.M. in Engineering Sciences from Harvard in 2005, and Ph.D. in Electrical and Computer Engineering from Carnegie Mellon in 2009. His industry experience includes ISO-New England and Edison Mission Energy Marketing and Trading. His research interest includes modeling and control in data-rich large-scale systems, grid integration of clean energy resources, and electricity markets.
Dr. Xie received the U.S. National Science Foundation CAREER Award, and DOE Oak Ridge Ralph E. Powe Junior Faculty Enhancement Award. He was awarded the 2017 IEEE PES Outstanding Young Engineer Award. He was recipient of Texas A&M Dean of Engineering Excellence Award, ECE Outstanding Professor Award, and TEES Select Young Fellow. He is an Editor of IEEE Transactions on Smart Grid, and the founding chair of IEEE Power and Energy Society Subcommittee on Big Data & Analytics for Grid Operations. He and his students received the Best Paper awards at North American Power Symposium and IEEE SmartGridComm.
OIL for District-Energy Systems
Speaker: Michael Chertkov (Los Alamos National Laboratory)
Abstract: We discuss how Optimization, Inference and Learning (OIL) methodology is expected to re-shape future demand-response technologies acting across interdependent energy, i.e. power, natural gas and heating/cooling, infrastructures at the district/metropolitan/distribution level. We describe hierarchy of deterministic and stochastic planning and operational problems emerging in the context of physical flows over networks associated with the laws of electricity, gas-, fluid- and heat-mechanics. Then we proceed to illustrate development and challenges of the physics-informed OIL methodology on examples of:
a) Graphical Models approach applied to a broad spectrum of the energy flow problems, including online reconstruction of the grid(s) topology from measurements;
b) Direct and inverse dynamical problems for timely delivery of services in the district heating/cooling systems;
c) Ensemble Control of the phase-space cycling energy loads via Markov Decision Process (MDP) and related reinforcement learning approaches.
Biography: Dr. Chertkov’s areas of interest include statistical and mathematical physics applied to energy and communication networks, machine learning, control theory, information theory, computer science, fluid mechanics and optics. Dr. Chertkov received his Ph.D. in physics from the Weizmann Institute of Science in 1996, and his M.Sc. in physics from Novosibirsk State University in 1990. After his Ph.D., Dr. Chertkov spent three years at Princeton University as a R.H. Dicke Fellow in the Department of Physics. He joined Los Alamos National Lab in 1999, initially as a J.R. Oppenheimer Fellow in the Theoretical Division. Since 2001 he is a technical staff member at LANL. Since 2012 Dr. Chertkov is advising SkolkovoTech — new graduate school in Moscow/Russia. He also has an adjunct professor affiliation with the Department of Industrial & Operations Engineering of the U of Michigan, Ann Arbor. Dr. Chertkov has published more than 180 papers. He is an editor of the Journal of Statistical Mechanics (JSTAT), associate editor of IEEE Transactions on Control of Network Systems, member of the Editorial Board of Scientific Reports (Nature Group), a fellow of the American Physical Society (APS) and a senior member of IEEE.
Minor Formulation and Relaxation for the Optimal Power Flow Problem
Speaker: Andy Sun (Georgia Tech)
Abstract: I will present recent advances in solving large-scale Optimal Power Flow (OPF) problems. I will present a new reformulation of matrix rank constraint using 2-by-2 minors of the matrix and its application in forming strong convex relaxations based on second-order cone programming (SOCP) for OPF. Extensive experiments show the proposed SOCP relaxations produce very high quality solutions compared to the traditional SDP relaxations and can be orders of magnitude faster. This is joint work with Santanu Dey and Burak Kocuk.
Biography: Andy Sun is an assistant professor in the H. Milton Stewart School of Industrial and Systems Engineering, Georgia Institute of Technology. He received doctoral degree from Massachusetts Institute of Technology. His research interests are in the areas of robust and stochastic optimization, large-scale non-convex optimization, and mathematical modeling of electric energy systems. Dr. Sun’s research has won several awards, including the INFORMS Georgia B. Dantzig Doctoral Dissertation Award, the INFORMS Junior Faculty Interests Group (JFIG) Paper Competition, and the INFORMS ENRE Best Publication in Energy. Dr. Sun has been working with several leading utility companies such as the ISO New England and the Southern Company on robust power system operations. Dr. Sun is an IEEE Senior Member.
Robust Security Constrained ACOPF via Conic Programming
Speaker: Antonio Conejo (Ohio State University)
Abstract: This seminar presents a technique to solve efficiently a robust security-constrained optimal power flow (R-SCOPF) problem. A bilevel max-min optimization model is proposed to find the worst contingencies while representing AC power flow equations and considering corrective actions. The objective is to minimize the total generation cost while satisfying operation and security constraints. The upper-level problem allows representing contingencies using binary variables. The lower-level problem represents the AC optimal power flow problem using a second-order cone relaxed formulation. Duality is then used to merge the upper- and lower-level problems into a single-level one. The resulting mixed-integer second-order conic problem can be efficiently solved. Two case studies are presented as applications of this technique.
Biography: Antonio J. Conejo received the M.S. degree from MIT, Cambridge, Massachusetts, US, in 1987, and the Ph.D. degree from the Royal Institute of Technology, Stockholm, Sweden, in 1990. He is currently a professor at the Integrated System Engineering and the Electrical & Computer Engineering Departments, The Ohio State University, Columbus, Ohio, US. He has published over 190 papers in SCI journals and is the author or co-author of books published by Springer, John Wiley, McGraw-Hill and CRC. He has been the principal investigator of many research projects financed by public agencies and the power industry and has supervised 19 PhD theses. He is an IEEE Fellow for contributions to analytical techniques for power system scheduling. His research interests include control, operations, planning, economics and regulation of electric energy systems, as well as statistics and optimization theory and its applications.
AC Optimal Power Flow with Robust Feasibility Guarantees
Speaker: Daniel K. Molzahn (Argonne National Labs)
Abstract: Optimal power flow (OPF) is an important problem in the operation of electric power systems. The solution to an OPF problem provides a minimum cost operating point that satisfies both engineering limits and the power flow equations corresponding to the network physics. A variety of applications, such as distribution grid optimization and transmission grid security assessment, call for network models that use the non-linear “AC” power flow equations. Additionally, with growing penetrations of renewable generation, OPF problems are increasingly influenced by the forecast uncertainty and short-term fluctuations that are inherent to many renewable energy sources. Thus, reliable and efficient operation of power systems requires the solution of non-convex “AC OPF” problems that incorporate uncertainty.
This presentation describes a recently proposed iterative algorithm for the AC OPF problem that yields a solution with robust feasibility guarantees. The algorithm is based on the observation that considering uncertainty leads to a tightening of the original, deterministic constraints in order to safely accommodate fluctuations due to uncertain generation. The main challenge in solving the robust AC OPF problem is to guarantee the existence of feasible solutions for all points within the uncertainty set. To overcome this challenge, the proposed algorithm (1) employs convex relaxations of the AC power flow equations to obtain a conservative estimate of the required tightenings and (2) uses a sufficient condition for power flow solvability related to the static transfer stability limits of each line. The presentation provides a detailed description of the algorithm and results showing the efficiency and robustness of the proposed approach. This is joint work with Line A. Roald.
Biography: Daniel Molzahn is a computational engineer at Argonne National Laboratory. Prior to his current position, Daniel was a Dow Sustainability Fellow at the University of Michigan. Daniel 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 application of optimization techniques to electric power systems.
Spectral Characterization of Controllability and Observability for Frequency Regulation Dynamics
Speaker: Steven Low (Caltech)
Abstract: We characterize the controllability and observability of the swing dynamics in frequency regulation using spectral graph theory. In particular, we show that controllability/observability depends on the network graph structure and on the algebraic coverage of buses with controllable loads/sensors. Algebraic coverage captures how buses interact with each other through the network and can be verified using the eigenvectors of the graph Laplacian matrix. This characterization suggests a new approach to optimal placement of controllable loads and sensors in future networks for frequency regulation. We show that, for the IEEE 39-bus test system, a single well chosen bus is capable of providing full controllability/observability. This is joint work with Linqi Guo.
Biography: Steven Low is a Professor of the Department of Computing & Mathematical Sciences and the Department of Electrical Engineering at Caltech, and holds visiting 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.
Distributed Feedback Voltage Control under Limited Communication and Limited reactive power
Speaker: Na Li (Harvard University)
Abstract: The rapid growth of power generated from renewable sources demands increasingly sophisticated and fast voltage regulation in distribution networks. Unfortunately, recent research shows that local controllers regulating the voltage through reactive power injection generally fail in achieving the desired regulation unless there is some communication between the controllers. However, the communication resources are still scarce for distribution networks. In this talk, I will discuss the role of communication in voltage control. Specially, I will present several distributed voltage control algorithms under different communication constraints and discuss their performance with respect to several performance metrics. Lastly, I will show one control algorithm that can (i) operate in a distributed fashion where each bus makes its decision based on local voltage measurements and communication with neighbors, (ii) always satisfy the reactive power capacity constraint, (iii) stabilize the voltage magnitude in an acceptable range, and (iv) minimize an operational cost. Joint work with Guannan Qu and Sindri Magnusson.
Biography: Na Li is an assistant professor in Electrical Engineering and Applied Mathematics of the School of Engineering and Applied Sciences in Harvard University since 2014. She received her Bachelor degree in Mathematics in Zhejiang University in 2007 and PhD degree in Control and Dynamical systems from California Institute of Technology in 2013. She was a postdoctoral associate of the Laboratory for Information and Decision Systems at Massachusetts Institute of Technology 2013-2014. Her research lies in distributed optimization and control of cyber-physical networked systems. She received NSF career award (2016) and AFSOR Young Investigator Award (2017). She entered the Best Student Paper Award finalist in the 2011 IEEE Conference on Decision and Control.
Controlling Variance in Power Flows Under Stochastic Injections
Speaker: Daniel Bienstock (Columbia University)
Abstract: In recent years, many authors have considered variants of OPF and PF problems where some of the loads or the generation have stochastic attributes. A typical optimization framework computes an optimal control (with some structure) while stipulating through (for example) chance constraints that line limits are satisfied with high probabilities. However, experiments reveal that the resulting systems may exhibit very large variability, as revealed for example in the variance of power flows. In this talk we describe a number of techniques for modifying the OPF computations so as to significantly diminish such variability. To put more precisely, we study the cost/variance frontier and discuss convex optimization techniques for attaining efficient solutions, including data-driven techniques for producing robust estimates of variances. Joint work with Apurv Shukla (PhD student Columbia University).
Biography: Daniel Bienstock is a professor of Operations Research and Applied Mathematics at Columbia University. He received the PhD from MIT in Operations Research. His research focuses on computational optimization, both from a methodological perspective and in the context of its use in power engineering.
Clearing and Pricing for Coordinated Gas and Electricity Day-ahead Markets Considering Wind Power Uncertainty
Speaker: Jianhui Wang (Southern Methodist University/Argonne National Laboratory)
Abstract: Natural gas and electricity systems are becoming increasingly coupled. Gas-fired units (GFUs) are replacing retired coal-fired power plants, and the power systems are more dependent on the ramping capability provided by GFUs. The GFUs’ power generation capability relies on the availability of gas supply, which is jointly determined by the capacity of gas suppliers and pipeline networks. However, the gas and electricity markets are operated separately. Consequently, the GFUs are forced to “represent” the entire power system to bid in the gas market: they must make forecasts regarding future gas consumption and bear the risk of improper contracts or being unable to meet generation schedules due to occasional insufficient gas supply. When facing larger shares of renewable energy and more-frequent gas network congestions, the current market framework is particularly unreliable and inefficient, as well as economically unfriendly to the investors of the GFU assets. In this presentation, we try to develop a framework that can integrate the two markets. By properly pricing the scarce resources (e.g., gas transmission capacity), the joint market can help allocate the resources more efficiently while satisfying the demands. Moreover, a more forward-looking day-ahead market clearing framework is presented by considering the uncertainty brought by renewable energy. The formulation and algorithm of the proposed joint market model will be presented, as are some case studies.
Resiliency and Autonomous Optimization
Speaker: Sakis Meliopoulos (Georgia Institute of Technology)
Abstract: Recent technological advances in protection, control and optimization are enabling a more automated power system. Research efforts are focused on integrating these technologies into a seamless and cyber secure infrastructure for autonomous protection, control, operation and optimization. This infrastructure is the basis for accommodating and providing robust solutions to new problems arising from the integration of renewables, increased uncertainty and steeper ramp rates. We discuss the infrastructure from the breakers and control devices at the substation level to the control center. At the substation level we build upon the dynamic state estimation based protection (EBP) and a centralized substation protection to provide real time models and operating conditions to the control center where optimization problems are autonomously formed, solved and return control commands back to the field. We discuss how this infrastructure detects and corrects bad data (self-healing) resulting from equipment malfunctioning (such as failures in the data acquisition systems) or from cyber-attacks and how the bad data are excluded from upstream functions. The end result is the autonomous optimization function is operating on validated data. The approach enhances the resiliency of the system and autonomously optimizes the operation of the system. In the presentation we discuss the autonomous formation of optimization problems and solution methods. Since all functions are model based we discuss an automated creation of the models required at each level of the infrastructure. Standardized model syntax has been developed named SCAQCF (System and Control Algebraic Quadratic Companion Form). We present examples of this standard and its application to the autonomous formation and solution of an optimization problem for the control of distributed energy resources in a distribution circuit.
The proposed approach and infrastructure forms the basis for the next generation of Energy Management Systems. Presently the approach is being integrated in an advanced distribution management system under the DoE ENERGIZE program.
Biography: Sakis (A. P.) Meliopoulos was born on March 19, 1949 in Katerini, Greece. He obtained a Diploma in Electrical and Mechanical Engineering from the National Technical University in Athens, Greece in 1972 and a Master in EE (1974) and a Ph.D. degree (1976) from the Georgia Institute of Technology in Atlanta, Georgia, USA. Dr. Meliopoulos’ first professional association was with Western Electric (1971) in Atlanta, Georgia. After receiving a PhD degree in 1976, he joined the faculty of the Georgia Institute of Technology as an Assistant Professor (1976), Associate Professor (1982-88) and full professor (1989-present). In 2006 Dr. Meliopoulos was named the Georgia Power Distinguished Professor. He is actively involved in education and research for improved safety and electromagnetic compatibility of electric power installations, protection and control of power systems and the application of new technology in these areas. Since 1999 he is the Georgia Tech Site Director of PSERC, an NSF I/URC. Dr. Meliopoulos has pioneered several new analysis and design techniques for bulk power reliability analysis, safety, protection and electromagnetic compatibility of electric power systems. Most well-known is the EPRI transmission reliability program TRELLS (now renamed TransCARE), the GPS-synchronized harmonic state measurement system for transmission systems (first (1993) wide area measurement system on NYPA and still operational), the distributed dynamic state estimation method (SuperCalibrator), the dynamic state estimation based protection (setting-less relay), his invention of the Smart Ground Multimeter, the EPRI grounding analysis programs, the WinIGS (Integrated Grounding System analysis and design), the GEMI (Grounding and ElectroMagnetic Interference) computer code, and the mGrid computer code – a methodology and implementation for precise analysis of multi-wire power systems with distributed energy resources. Dr. Meliopoulos has modernized many power system courses at Georgia Tech, introduced new courses, initiated the power system certificate program for practicing engineers and most importantly he has introduced visualization and animation methodologies that dramatically increase the teaching efficiency of complex power system concepts. Dr. Meliopoulos is a Fellow of the IEEE (1993). He holds 3 patents, he has published two books, a chapter in the Standard Handbook for Electrical Engineers and over 350 technical papers. He has received a number of awards, including the Sigma Xi Young Faculty award (1981), the outstanding Continuing Education Award, Georgia Institute of Technology (twice: 2002 and 2014), three of his papers have received the best paper award (IEEE-PES-SC-1984, IEEE-PES-EC-1987, and IEEE-CSS-HICSS 2002), he received the IEEE PES Prize Paper Award in 2017, he received the 2005 IEEE Richard Kaufman Award and the 2010 George Montefiore international award.
Sampled Subgradient Methods for the Lagrangian Dual in Stochastic Mixed Integer Programming
Speaker: Jeff Linderoth (UW-Madison)
Abstract: We will discuss recent work in improving the performance of classical stochastic (sub)gradient methods when used in the context of solving the Lagrangian Dual of a Stochastic Mixed Integer Program. Enhancements include both sub-sampling and asynchronous versions that run effectively on large-scale, distributed, heterogeneous computing platforms. This is joint work with Cong Han Lim, Jim Luedtke, and Steve Wright
Biography: Jeff Linderoth is a Professor and the chairperson of the department of Industrial and Systems Engineering at the University of Wisconsin-Madison. Prof. Linderoth holds a courtesy appointment in the Computer Sciences department and as a Discovery Fellow in the Optimization Theme at the Wisconsin Institutes of Discovery. Dr. Linderoth received his Ph.D. degree from the Georgia Institute of Technology in 1998. He was previous employed in the Mathematics and Computer Science Division at Argonne National Laboratory, with the optimization-based financial products firm of Axioma, and as an Assistant Professor at Lehigh University. His awards include an Early Career Award from the Department of Energy, the SIAM Activity Group on Optimization Prize, and the INFORMS Computing Society (ICS) Prize. In 2016, he was elected an INFORMS Fellow.
SDDiP: Stochastic Dual Dynamic integer Programming
Speaker: Shabbir Ahmed
Abstract: Stochastic Dual Dynamic Programming (SDDP) is a well-established algorithm for solving large-scale multistage stochastic linear optimization problems arising in hydropower scheduling applications. The presence of nonconvexities, e.g. integer variables to model complex operational constraints, impedes the applicability of SDDP. In this talk, we extend the SDDP approach for solving a large class of multistage stochastic integer optimization problems. We demonstrate the effectiveness of the approach on a nonconvex hydroscheduling case study from Norway.
This talk is based on joint works with Arild Helsuth, Martin Hjelmeland, Andy Sun and Jikai Zou.
Biography: Shabbir Ahmed is the Anderson-Interface Chair and Professor in the H. Milton Stewart School of Industrial & Systems Engineering at the Georgia Institute of Technology. His research interests are in optimization, specifically stochastic and integer programming. He is a past Chair of the Stochastic Programming Society. His honors include the National Science Foundation CAREER award, two IBM Faculty Awards, and the INFORMS Dantzig Dissertation award. He is a Fellow of INFORMS.
On the Rigorous Evaluation of Stochastic Approaches to Power Systems Operations
Speaker: Jean-Paul Watson (Discrete Math and Optimization Department, Sandia National Laboratories)
Abstract: While there is significant recent research on stochastic optimization approaches to power systems operations, e.g., unit commitment and economic dispatch, there are still major impediments to their adoption in practice. In our experience, developed over years of attempting to deploy such approaches, one key issue is accurate evaluation of any proposed approach, relative to existing deterministic operational methodologies. In this talk, we discuss the challenges in such evaluation, and report on a novel methodology addressing what we feel to be deficiencies with current approaches. In the talk, we focus on issues relating to data availability and segmentation, probabilistic scenario generation, and the impact of scenarios on operational performance.
Biography: Dr. Jean-Paul Watson is a member of the Analytics Department, and has been a member of this department or a previous version of this department since starting at Sandia in 2003. 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 programs, with applications in the domain of electricity grid operations and planning. He presently leads projects at Sandia for DOE ARPA-e (stochastic unit commitment), DOE OS/ASCR (optimization for the power grid), and LDRD (quantifiable and rigorous power grid operations and planning). Previously, he developed solutions for real-world stochastic optimization problems in logistics (Lockheed Martin and the US Army) and sensor placement (US Environmental Protection Agency). Additionally, he led the development of programs involving the use of semantic graph technologies for performing geospatial imagery analysis. He is a co-developer of Sandia’s Coopr (https://software.sandia.gov/trac/coopr) open-source software package for modeling and solving optimization problems, and has published over 25 journal articles in the areas of optimization algorithms and their application. Prior to graduate school, he worked at IBM in Austin, Texas working on VLSI design – specifically for the PowerPC family of chips — and at Hughes Information Technology Corporation in Aurora, Colorado working on satellite systems. He received his PhD in Computer Science from Colorado State University in 2003.