Informs Annual Meeting Phoenix 2018

INFORMS Phoenix – 2018

MC46

4 - Data-driven Voltage Control in Distribution Networks Jiafan Yu, Stanford University, Stanford, CA, United States Traditional approaches to reactive power management in distribution networks have focused on designing stable and near-optimal centralized or distributed control schemes under the assumption that network physics constraints are completely known. In this talk, we explore an alternative approach that learns a control procedure by observing streams of measurements from the system. We show that a simple batch learning and optimal control procedure attains comparable performance to existing approaches without requiring any prior knowledge of the network and give some analytical guarantees. We then show how the procedure can be utilized to construct an online learning process for the problem. n MC45 North Bldg 228A Game Theory and Electricity Market Sponsored: Energy, Natural Res & the Environment/Electricity Sponsored Session Chair: Luce Brotcorne, INRIA, Villeneuve D’Ascq, 59650, France 1 - Unit Commitment under Market Equilibrium Constraints Bernard Fortz, Universit libre de Bruxelles, Brussels, Belgium, Luce Brotcorne, Fabio D’Andreagiovanni, Jérume De Boeck We consider an extension of the Unit Commitment problem with a second level of decisions ensuring that the produced quantities are cleared at market equilibrium. In their simplest form, market equilibrium constraints are equivalent to the first-order optimality conditions of a linear program. The UC in contrast is usually a mixed-integer nonlinear program (MINLP), that is linearized and solved with traditional Mixed Integer (linear) Programming (MIP) solvers. Taking a similar approach, we are faced to a bilevel optimization problem where the first level is a MIP and the second level linear. 2 - A Bilevel Optimization Formulation of Priority Service Pricing Anthony Papavasiliou, Universite Catholique de Louvain, Center for Operations Research and Econometri, Voie du Roman Pays 34, Louvain la Neuve, 1348, Belgium, Yuting Mou, Philippe Chevalier Priority service pricing is a promising approach for mobilizing residential demand response, by offering electricity as a service with various levels of reliability. The proper pricing guarantees consumers self-select a level of reliability that corresponds to the reliability that the system can offer. However, traditional theory for menu design is based on numerous stringent assumptions. In addition, the objective of the menu design is to maximize social welfare, while the profit requirement is not accounted for. We design a priority service menu as the solution to a Stackelberg game modelled as a bi-level optimization problem. The approach is illustrated on the Belgian power market. 3 - Increasing Electric Vehicle Adoption via Strategic Siting of Charging Station Martim Joyce-Moniz, Polytechnique Montreal, Montreal, QC, Canada, Miguel Anjos, Bernard Gendron Governments everywhere have started setting ambitious goals for electric vehicle (EV) adoption for the next few decades. One important obstacle to massive EV adoption, however, is the lack of a reliable, wide-reaching network of publicly- available charging stations that allows drivers to charge their EV both along their daily commutes and longer travels. We present a multi-period optimization framework for the siting and sizing of fast charging stations over large landscapes, which incorporates evidence-based demand dynamics reflecting how new infrastructure impacts EV adoption growth. 4 - A Bilevel Framework for Optimal Price-Setting of Time-and-level-of-use Tariff Miguel F. Anjos, Polytechnique Montreal, Montreal, QC, Canada, Mathieu Besan on, Luce Brotcorne, Juan A. Gomez-Herrera The Time and Level of Use is an energy tariff for Demand Response, designed for residential users and providing suppliers with guarantee on the consumption. We formulate the supplier decision as a bilevel, biobjective problem optimizing for both financial loss and guarantee. A decomposition method is proposed, related to the epsilon-constraint and optimal value transformation. It allows for the computation of an exact solution by finding optimality candidates and then eliminating dominated ones. The method is effectively applied to experimental consumption data .

n MC46 North Bldg 228B Energy Markets II Sponsored: Energy, Natural Res & the Environment/Energy Sponsored Session Chair: Seyedamirabbas Mousavian, Clarkson University, NY, 13699- 5790, United States Co-Chair: Felipe A. Feijoo, l 1 - Deep Neural Network Ensemble Structures for Multi-step Forecasting of Ocean Wave Elevation and WEC Power Output Mohammad Pirhooshyaran, Lehigh University, Bethlemhem, PA, 18015, United States, Lawrence V. Snyder There exists an undeniable interest toward utilizing ocean energy as a renewable energy source. This research focuses on the Multiple step ahead prediction of ocean wave heights and power via innovative structure of Deep Neural Networks. An ensemble of Bidirectional Long short Term Memory (BLSTM) Networks is proposed to capture both long and short dependencies of the large historical data. Several optimization algorithms such as ADAM and RMSProb, SGD, etc., are considered to optimize the network. The results indicate that the proposed model is more accurate in compare with conventional forecasting methods such as Support Vector Regression (SVR) or SARIMA. 2 - Green Investment under Policy Uncertainty and Bayesian Learning Verena Hagspiel, Norwegian University of Science and Technology, Trondheim, Norway, Jacco J.J. Thijssen, Peder Dalby, Gisle Gillerhaugen, Tord Leth-Olsen This paper examines how investment behavior is affected by updating a subjective belief on the timing of a subsidy revision. We analyze a scenario where a retroactive downward adjustment of fixed feed-in tariffs (FIT) is expected through a regime-switching model. We find that investors are less likely to invest when the arrival rate of a policy change increases, and prefer a lower FIT with a long expected lifespan. If electricity is sold in a free market after retraction we find that if policy uncertainty is high, an increase in the FIT will be less effective at accelerating investment. However, if policy risk is low, FIT schemes can significantly accelerate investment, even in highly volatile markets. 3 - Quantifying the Effect of Natural Gas Price Uncertainty on Economic Dispatch Cost Uncertainty with Estimated Correlated Uncertainties Dan Hu, Iowa State University, Ames, IA, 50010, United States, Sarah M. Ryan In competitive electricity markets, vulnerability in gas supply to electricity generators creates a risk of high electricity prices. We propose a daily economic dispatch model that accounts for natural gas availability and cost from both contracts and the spot market. With probabilistic correlated inputs estimated from historical data for electricity demand and gas spot prices, we use Monte Carlo simulation to generate the resulting distributions of electricity dispatch cost both with and without gas price uncertainty. The effect of gas price uncertainty is assessed in terms of distance between the distributions and the effect of gas availability is addressed through sensitivity analysis. 4 - Toward a Synthetic Model for Distribution System Restoration and Crew Dispatch Bo Chen, Argonne National Laboratory, Lemont, IL, United States Distribution service restoration (DSR) is critical for improving the resilience and reliability of modern distribution systems. Restoring electricity service to affected customers requires multiple crews with different skill sets to perform multiple tasks that are procedurally interdependent. We introduce a synthetic model that integrates both a service restoration model and crew dispatch model based on a universal routing model considering the switching sequence for safely operating remotely/manually operated switches, and dispatch solutions for crews with different skill sets.

201