2015 Informs Annual Meeting

SB61

INFORMS Philadelphia – 2015

2 - Baseball- Discovering the Moneyball Effect Sean Barnes, University of Maryland, 4352 Van Munching Hall, University of Maryland, College Park, MD, 20742, United States of America, sbarnes@rhsmith.umd.edu, Margret Bjarnadottir This case uses player evaluation and personnel decision-making in the Major League Baseball to introduce many of the key steps of data mining projects. The data mining process is a unique combination of art and science, and teaching the art of data mining is challenging to do in a standard classroom setting with small data sets. The goal of this case is to move beyond the simple “cookie cutter” data sets, and introduce students to the challenges of dealing with real data to answer important questions, as well as introduce or reinforce multiple data mining methods. The case builds on a very rich data set collected by the authors, which allows for students or groups of students to arrive at very different answers to the same question. For example, what is the best predictive model? Which players should be pursued? 3 - Medication Waste Reduction in an In-hospital Pharmacy: A Case that Bridges Problem Solving between a Traditional Case and an Industry Project Gregory Dobson, University of Rochester, University of Rochester, Rochester, NY, United States of America, greg.dobson@simon.rochester.edu, Vera Tilson This Operations Management case describes a waste-reduction project in a compounding pharmacy in a hospital. Every day, pharmacy technicians prepare a large number of patient-specific medication doses and then deliver these doses to various hospital units. With the rising cost of medications, pharmacy managers become concerned that a significant number of compounded medication doses are not administered to patients and are subsequently wasted. The students are asked to quantitatively evaluate proposed changes to the compounding and delivery process and to estimate savings from process reconfiguration. Two large datasets are provided with the case to facilitate hypothesis generation regarding probable causes of waste and to analyze proposed changes. The analysis will deepen students’ spreadsheet skills as well as mathematical modeling of inventory problems. The case is presented in parts, and it is discussed over one and a half class meetings – simulating the steps of a field project: interviewing a client, framing the client’s problem, formulating a data request, and then analyzing the data and delivering a recommendation. This case has been used in a core MBA operations management class; it could also be used in a health-care operations or in a business modeling course. SB61 61-Room 111B, CC Dimensionality Reduction Techniques for Generation Capacity Expansion Problems with Intermittent Resources Sponsor: ENRE – Energy II – Other (e.g., Policy, Natural Gas, Climate Change) Sponsored Session Chair: Fernando de Sisternes, Argonne National Laboratory - MIT Energy Initiative, 77 Massachusetts Avenue, E19-341, Cambridge, MA, 02139, United States of America, ferds@mit.edu 1 - On the Temporal Resolution of Electric Sector Capacity Expansion Models James Merrick, Stanford University, Huang Engineering Center, Stanford, CA, 94305, United States of America, jmerrick@stanford.edu When hours are assumed to be independent, we can capture hourly information by removing similar hours and weighting a representative. Applying hierarchical clustering to a sample dataset, I show that while the variability of demand can be captured in the order of 10 hours, including wind and solar resources warrants in the order of 1000. A similar analysis is undertaken for days and weeks. I conclude with how the approach can reduce resolution further using information about problem structure. 2 - A Robust Method to Choose Representative Weeks in Renewable Generation Capacity Expansion Problems Fernando de Sisternes, Argonne National Laboratory - MIT Energy Initiative, 77 Massachusetts Avenue, E19-341, Cambridge, MA, 02139, United States of America, ferds@mit.edu, Ignacio Núñez We propose a new week selection method for capacity expansion formulations with unit commitment constraints that is robust to the renewable capacity in the system. This method enables the use of such formulations in determining the optimal amount of thermal and renewable capacity. The proposed method selects a week combination with hourly resolution that represents the energy below the net load duration curve and its inter-temporal variability across different renewable capacity scenarios.

3 - Optimization-based Method for Scenario Reduction in Generation Expansion Models Ignacio Núñez, Research Assistant Professor, University of the Andes, Av. Monseñor Alvaro del Portillo 12.455, Las Condes, Santiago, RM, 7620001, Chile, ijnunez@uandes.cl, Fernando de Sisternes Including intermittent generation in generation expansion models requires a detailed representation of the short-term operation of the system, increasing dramatically the dimensionality of the problem. We present a mixed integer linear model that selects a specific number of representative scenarios (e.g. days or weeks) and weights to minimize the maximum error in representing the net load duration curve for different capacities of solar photovoltaic and wind power. 4 - Modeling a Paradigm Shift: Distributed vs. Centralized Options in Electricity Capacity Planning Jesse Jenkins, PhD Student And Researcher, MIT Engineering Systems Division and MIT Energy Initiative, 77 Massachusetts Avenue, Cambridge, MA, 02139, United States of America, jessedj@mit.edu Distributed energy resources such as solar PV, combined heat and power, batteries, and demand response, may compete with conventional power plants in the provision of electricity services. This research employs state-of-the-art dimensionality reduction techniques to ensure the computational tractability of a generation capacity expansion model that considers tradeoffs between distributed and centralized energy options, including impacts on network costs, losses, congestion, and system services. SB62 62-Room 112A, CC Modeling, Analysis, and Management of Water- Centric Systems Sponsor: ENRE – Environment I – Environment and Sustainability Sponsored Session Chair: Nagesh Gavirneni, Associate Professor, Cornell University, 325 Sage Hall, Ithaca, United States of America, nagesh@cornell.edu 1 - Inland Waterway System Dynamics Furkan Oztanriseven, University of Arkansas, 1900 N. Garland Avenue, Apt. #26, Fayetteville, AR, 72703, United States of America, oztanriseven@gmail.com, Heather Nachtmann The growth of population and economic advancement led to a higher demand for transportation services. However, the expansion in the transportation sector comes with substantial costs, such as higher gas emissions and traffic congestion. Inland waterways offer an environmental friendly and economically sound transportation alternative. In this study, we utilized a system dynamics model to better understand the interconnected relationships between the economy and the inland waterways performance. 2 - A Novel Mathematical Model for Water Value Chain Management The amount of water in the world is limited and a large amount of water is wasted due to ineffective operation for water infrastructure. A decision model which can efficiently manage the water value chain is urgently needed to reduce water consumption. In this research, we develop a model for water management where the water flows at different stages of transitions (e.g., source, distribution) are modeled. The proposed model is demonstrated to be able to significantly reduce water consumption. 3 - Automated Analysis of Online Reviews to Improve Visitor Experience in New York State Parks Nagesh Gavirneni, Associate Professor, Cornell University, 325 Sage Hall, Ithaca, NY, United States of America, nagesh@cornell.edu, Hari Udayapuram The New York State Park system has a large volume of visitor feedback on online platforms such as Yelp, TripAdvisor, and Google. We design, develop, and implement software systems that can download, organize, and analyze the text from these online reviews and help the park managers identify strategies to improve the visitor experience at their facilities. Mengqi Hu, University of Illinois at Chicago, Chicao, IL, United States of America, mhu@uic.edu, Afshin Ghassemi

86