2016 INFORMS Annual Meeting Program

SB04

INFORMS Nashville – 2016

SB04 101D-MCC Energy Storage and Virtual Trading in the Smart Grid Sponsored: Energy, Natural Res & the Environment, Energy I Electricity Sponsored Session Chair: Miguel F Anjos, Polytechnique Montreal, C.P. 6079, Succ. Centre-ville, Montreal, QC, H3C 3A7, Canada, anjos@stanfordalumni.org 1 - Optimizing Energy Flows For A Grid Connected Smart House Producing Renewable Energy Luce Brotcorne, INRIA, luce.brotcorne@inria.fr Ekaterina Alekseeva, Michel Gendreau, Mohammed Skiredj We focus on optimizing energy flows for demand management of a grid connected smart house equiped with a system combining photovoltaic electricity and battery . The smartly scheduled way of using, storing, generating, buying and selling energy allows customers to reduce electricity payments, to be less dependent on the grid and avoid creating peak power demand in the grid. We propose a stochastic mathematical linear program to make an optimal decision with the lack of perfect information related to purchasing electricity prices and energy produced by PV generator. 2 - Capacity Expansion Modeling For Storage Technologies Elaine Thompson Hale, Senior Engineer, National Renewable Energy Laboratory, Golden, CO, United States, elaine.hale@nrel.gov, Brady Stoll, Trieu Mai The Resource Planning Model (RPM) is a capacity expansion model designed for regional power systems and high levels of renewable generation. Recent extensions capture value-stacking for storage technologies, including batteries and concentrating solar power with storage. After estimating per-unit capacity value and curtailment reduction potential, RPM co-optimizes investment decisions and reduced-form dispatch, accounting for planning reserves; energy value, including arbitrage and curtailment reduction; and three types of operating reserves. Multiple technology cost scenarios are analyzed to determine level of deployment in the Western Interconnection under various conditions. 3 - Optimizing Storage Operations In Transmission-constrained Networks For Medium And Long-term Operation Diego Alejandro Tejada Arango, Universidad Pontificia Comillas, IIT, Madrid, Spain, Diego.Tejada@iit.comillas.edu Sonja Wogrin, Efraim Centeno The main objective is to present a new approach to model the storage operation in the context of Medium- and Long-Term Operational Planning (MLTOP). This approach is based on the system-state framework but including transmission constraints. A DC power flow approach is used to represent the transmission network. The methodology is related to clustering techniques using information such as demand and wind generation per node. Case studies are presented in order to compare the newly proposed methodology and the hourly approach. The results illustrate the computational time reduction without loss of accuracy in the solution. 4 - A Model Of Virtual Trading And The Forward Day Ahead Market Gauthier De Maere D’Aertrycke, GDF Suez, Boulevard Simon Bolivar 34, Brussels, Belgium, gauthier.demaeredaertrycke@gdfsuez.com, Yves Smeers, Andreas Ehrenmann The day ahead market plays an ambiguous role in restructured electricity markets. It is meant to help physical transactions such as the starting of machines in the unit commitment but is also intended to be a forward market capable of transferring the vagaries of real time prices into forward prices. Virtual trading was introduced for that purpose. We provide a model of virtual trading and give conditions for achieving the objective. We discuss what those conditions would imply in case of important penetration of decentralised energy. We also show some numerical experiment.

2 - A Multi-response Multilevel Model With Application In Nurse Care Coordination Bing Si, Arizona State University, Tempe, AZ, United States, bingsi@asu.edu, Jing Li Nurse care coordination plays a vital role in promoting patient outcomes. The recently developed Nurse Care Coordination Instrument (NCCI) enables quantitative data to be collected on nurses’ coordination activities, demographics, workload and practice environment. Driven by this, we propose a novel multi- response multilevel model with joint fixed/random effect selection across multiple responses and apply it to a dataset collected across four U.S. hospitals using the NCCI. Our study conducts the first quantitative analysis linking multiple care coordination metrics with multilevel predictors and thus provides important insight into how care coordination might be impacted or improved. 3 - Optimal Expert Knowledge Elicitation For Bayesian Network Structure Identification Yan Jin, University of Washington, Seattle, WA, United States, yanjin@uw.edu, Cao Xiao This talk is about a systematic approach that combines observational data and expert knowledge to better learn the influential relationships between variables for networked systems, as well as automates the expert elicitation process and collect the most informative expert knowledge, optimally matched to the observational data, to improve the learning of the BN structure. Applications include event cascade modeling of Alzheimer’s disease and human We consider the problem of change detection in dynamic attributed networks. First, networks are modeled through a generalized linear model (GLM). Then, a state-space model is built by considering a linear state model over the parameters of the GLM. Extended Kalman filter is used for estimating and predicting the parameters of the state-space model. For each upcoming network, a Pearson residual based on the actual network and its prediction is calculated. The Pearson residuals are monitored through an EWMA control chart. Comparison of this method with its static counterparts shows significant improvement in detecting changes. resource management key performance indicator measurement. 4 - Temporal Monitoring Of Dynamic Attributed Networks Mostafa Reisi, Georgia Tech, mostafa.reisi@gmail.com Chair: Maria Esther Mayorga, North Carolina State University, 400 Daniels Hall, Dept. of Industrial & Systems Engineering, Raleigh, NC, 27695, United States, memayorg@ncsu.edu 1 - Nicholson Student Paper Prize Maria Esther Mayorga, North Carolina State University, Dept. of Industrial & Systems Engineering, Raleigh, NC, 27695, United States, memayorg@ncsu.edu This session highlights the finalists for the 2016 George Nicholson Student Paper Competition. 2 - Robust Monotone Submodular Function Maximization Rajan Harish Udwani, James B. Orlin, Andreas S. Schulz, Massachusetts Institute of Technology, Cambridge, MA, rudwani@mit.edu We consider a robust formulation, introduced by Krause et al. (2008), of the classical cardinality constrained monotone submodular function maximization problem, and give the first constant factor approximation results. The robustness considered is w.r.t. adversarial removal of up to \tau elements from the chosen set. We give both, fast and practical approximation algorithms with sub-optimal guarantees as well as more theoretical ones achieving the best possible guarantee. Finally, we also give a black box result for the more general setting of robust maximization of monotone submodular functions subject to an independence system. 3 - A Constant-Factor Approximation For Dynamic Assortment Planning Under The Multinomial Logit Model Ali Aouad, Massachusetts Institute of Technology, Cambridge, MA, aaouad@mit.edu Abstract to come 4 - Delay, Memory, and Messaging Tradeoffs in Distributed Service Systems Martin Zubeldia, Massachusetts Institute of Technology, Cambridge, MA, Contact: zubeldia@mit.edu Abstract to come SB03 101C-MCC Nicholson Student Paper Prize II Invited: Nicholson Student Paper Prize Invited Session

43