Informs Annual Meeting Phoenix 2018

INFORMS Phoenix – 2018

SD43

2 - Constructing a Value Model for an Organization Ralph L. Keeney, Duke University, CA, United States

n SD43 North Bldg 227B Energy and Climate I Emerging Topic: Energy and Climate Emerging Topic Session Chair: Peter Larsen, Lawrence Berkeley National Laboratory,Berkeley, CA, 94720, United States 1 - Integrated Modelling Tool for Electricity Rate Design in Systems with High Penetration of DERs Miguel Heleno, Lawrence Berkeley National Laboratory, Berkeley, CA, United States This talk presents a modeling tool to support decision-making for electricity system planning and retail rate design in the context of distribution grids with high penetration of distributed energy resources (DERs). This tool is based on an interaction between two layers of optimization: the first, representing end-users, aims at finding the optimal behind-the-meter DER adoption; the second, representing the system operation, determines the optimal portfolio of utility assets that mitigates the impact of DERs deployment and ensures the steady state operation of the distribution grid. 2 - Drivers of the Reliability Contribution of Solar+Storage for Utilities in Florida Andrew Mills, Lawrence Berkeley National Lab., 1 Cyclotron Red, Uc Berkeley, CA, 94720, United States Abstract not available. 3 - Evaluating the Economics of Power Interruptions and Investments in Resilience Peter Larsen, Lawrence Berkeley National Laboratory, 1 Cyclotron Road, MS 90-4000, Berkeley, CA, 94720, United States This presentation will introduce and elaborate on a number of U.S. government- sponsored projects that explore the economics of power interruptions and investments that mitigate cyber security threats and other risks to U.S. power system reliability and resilience. 1 - The Regional Economic Accounting Tool: A Demonstration of Updates and Extensions Vanessa Vargas, Sandia National Laboratories, Albuquerque, NM, United States The Regional Economic Accounting tool (REAcct) was developed by economists at Sandia National Laboratories (SNL) for estimating order-of-magnitude economic impacts within scenario/real-time analysis. REAcct uses input-output modeling, geo-spatial data computational tools, and publicly available economic data to estimate the impacts of disruptive events and allows for detailed specification of sectors, regions, and disruption intervals. There are a number of improved capabilities for REAcct, including seasonal impact analyses, improved indirect impact estimation, and a proxy for economic system resilience. System resilience is related to a system’s ability to absorb, adapt, and recover from a disruptive event. Studies of past disruptive events indicate that factors which affect an economic system’s ability to recover can vary across regions, industries, or even firms in the same industry. We have established a preliminary method for incorporating resilience based on firm and industry characteristics in regions affected during an acute disruptive event. Incorporating quarterly data into the analysis helps tailor the results of analyses both to current economic conditions and economies with seasonal production. This paper details the changes to REAcct and provides a demonstration of the tool for specific geographies.

Constructing an organizational value model provides a foundation to build agreement on the organization’s objectives and to guide consistent decision- making among all important decisions. The concepts, process, and procedures to construct an organizational value model are discussed and illustrated using a value model constructed for the US Army Corps of Engineers. 3 - The Costs and Benefits of Homeland Security Research Deltof von Winterfeldt, CA, United States Since 2004 the U.S. Department of Homeland Security’s Office of University Programs has funded close to half a billion dollars on research, resulting in over two hundred research products, including software tools and other technologies. We evaluated the costs and benefits of selected research products, showing a high rate of return on this investment. 4 - Exploring Benefits of Value-focused Brainstorming Based on Associative Network Model Chen Wang, Tsinghua University, Beijing, China, Ying Xiang Generating alternatives is crucial for making good decisions. We propose an empirical method that helps examine the ability of decision makers to generate effective alternatives and the impact of objective-related stimuli. We apply machine learning methods to construct the cognitive map of a brainstorming exercise by projecting key words of the group discussion onto a latent cognition space. We then measure the degrees of exploitation and exploration for the alternative-generation process and demonstrate the benefits of value-focused brainstorming using recorded data. Joint Session ISim/Practice Curated: Simulation Modeling Applications Sponsored: Simulation Sponsored Session Chair: Mohammad Dehghanimohammadabadi, Northeastern University, Boston, MA 1 - Estimating the Real Demand in Bike-Sharing Systems Ashkan Negahban, Pennsylvania State University, 30 E. Swedesford Rd, Malvern, PA, 19355, United States The objective of the first part of this presentation is to: (1) provide a formal comparison between the distribution of bike-inter pickup times and the underlying (latent) customer inter-arrival time; and, (2) propose an alternative data filtering method to improve demand estimation by preserving the portion of the observed bike inter-pickup times that represent actual customer inter-arrival times. The second part introduces an iterative methodology combining simulation, nonparameteric bootstrapping, and indifference-zone analysis to estimate the true demand in bike-sharing systems when data analysis approaches fail. 2 - Introducing an Efficient Way to Optimize a Simulated Model with Data-table Inputs Mohammad Dehghanimohammadabadi, Northeastern University, 170 Brookline Avenue, Unit 1025, Boston, MA, 02115, United States The major goal in simulation-optimization studies is to obtain the most appropriate configuration settings of the model. This could be easily achieved via embedded optimizers in simulation packages such as OptQuest. However, these optimizers fail to optimize the simulation models designed based on data-table inputs. This study extends the boundaries of the existing simulation-optimization tools by incorporating data-tables as a control parameter of the simulation model. This new approach could significantly enhance the applicability of the simulation and reduce efforts to optimize data-table driven simulation models. 3 - An Agent-based Model of Individual Forgetting and Learning Behavior in Epidemics Kaiming Bi, Kansas State University, 1605 Roof Drive, Manhattan, KS, 66502, United States, Yuyang Chen, Chih-Hang Wu, David Ben-Arieh We presents two mathematical models, information forgetting curve (IFC) model and memory reception fadingcumulating (MRFC) model, to examine forgetting and learning behaviors of individuals during an epidemic. Both models consider behavior-change information may affect agent emotions and subsequently influence an individual’s behavior. IFC model utilizes a forgetting curve to process epidemic information, and MRFC model formulates disease information variations using the It process. An agent-based simulation model also developed to mimic the epidemic prevalence of the 2009 Chicago H1N1. n SD42 North Bldg 227A

105