Civil & Environmental Engineering
Permanent URI for this communityhttp://hdl.handle.net/1903/2221
Browse
11 results
Search Results
Item DATA-DRIVEN RISK MODELING FOR INFRASTRUCTURE PROJECTS USING ARTIFICIAL INTELLIGENCE TECHNIQUES(2023) Erfani, Abdolmajid; Cui, Qingbin; Civil Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Managing project risk is a key part of the successful implementation of any large project and is widely recognized as a best practice for public agencies to deliver infrastructures. The conventional method of identifying and evaluating project risks involves getting input from subject matter experts at risk workshops in the early phases of a project. As a project moves through its life cycle, these identified risks and their assessments evolve. Some risks are realized to become issues, some are mitigated, and some are retired as no longer important. Despite the value provided by conventional expert-based approaches, several challenges remain due to the time-consuming and expensive processes involved. Moreover, limited is known about how risks evolve from ex-ante to ex-post over time. How well does the project team identify and evaluate risks in the initial phase compared to what happens during project execution? Using historical data and artificial intelligence techniques, this study addressed these limitations by introducing a data-driven framework to identify risks automatically and to examine the quality of early risk registers and risk assessments. Risk registers from more than 70 U.S. major transportation projects form the input dataset. Firstly, the study reports a high degree of similarity between risk registers for different projects in the entire document of the risk register, and the probability and consequence of each risk item, suggesting that it is feasible to develop a common risk register. Secondly, the developed data-driven model for identifying common risks has a recall of over 66% and an F1 score of 0.59 for new projects, i.e., knowledge and experience of similar previous projects can help identify more than 66% of risks at the start. Thirdly, approximately 65% of ex-ante identified risks actually occur in projects and are mitigated, while more than 35% do not occur and are retired. The categorization of risk management styles illustrates that identifying risks early on is important, but it is not sufficient to achieve successful project delivery. During project execution, a project team demonstrating positive doer behavior (by actively monitoring and identifying risks) performed better. Finally, this study proposes using a data-driven approach to unify and summarize existing risk documents to create a comprehensive risk breakdown structure (RBS). Study results suggest that acquired knowledge from previous projects helps project teams and public agencies identify risks more effectively than starting from scratch using solely expert judgments.Item DEVELOPING A STATISTICAL VEHICLE DRIVER BEHAVIOR MODEL FOR ECO-ROUTING DEPLOYMENT(2022) Zhou, Weiyi; Zhang, Lei; Civil Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Predicting energy consumption accurately and reliably is critical for route optimization in eco-routing. State-of-the-practice methods for calculating energy consumption utilize second-by-second speed, acceleration, and power demand. Such models can achieve high accuracy but are not suitable for forecasting usages due to strict requirement of inputs and computing resources. Other methods used to predict energy consumption rely on average speed data to reduce data collection and computation efforts. However, they ignore the individuality of driving behavior, which is particularly important in near-term predictions of energy consumption, as shown in this paper. This study develops an input-output hidden Markov model (IOHMM) to cope with the influence of external environment and driving behaviors on individual driving features. The model is built and trained using passively collected geospatial location data. The approach furthermore improves the prediction of vehicle specific power (VSP) distribution, a critical parameter for energy predication, through predicted driving features. The model is tested in the Washington D.C. metropolitan area, and the performance is evaluated by comparing various indicators with the real-world values obtained from in-vehicle fuel recording devices. In general, the IOHMM behavior model demonstrates an overall cruising speed accuracy of 86.85% and acceleration rate accuracy of 82.73%. The behavior-integrated energy prediction model outperforms the traditional approaches by increasing the energy prediction accuracy to 86.81%. Results obtained from this study corroborate the importance of behavioral richness, environmental dynamics, and computation efficiency.Item WATER-ENERGY-CLIMATE NEXUS: INTERDEPENDANCIES AND TRADEOFFS, AND IMPLICATIONS FOR STRATEGIC RESOURCE PLANNING(2017) Liu, Lu; Forman, Barton; Civil Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)The water-energy nexus has been an active area of research in recent decades and has been explored in many different directions pertaining to its core. It is imperative to manage water and energy in a holistic approach as there are critical interconnections between the two systems. Climate change is an intrinsic environmental variable that has vital implications for the study of water-energy nexus, and hence, the term water-energy-climate nexus is used throughout the dissertation in reference to the interdependencies and tradeoffs between these systems. This dissertation is composed of three research studies under the domain of the water-energy-climate nexus, and they are interconnected through the intrinsic linkages among the three systems. The first study deals with the vulnerability of U.S. thermoelectric power plants to climate change. Findings suggest that the impact of climate change is lower than in previous estimates due to the inclusion of a spatially-disaggregated representation of environmental regulations and provisional variances that temporarily relieve power plants from permit requirements. This study highlights the significance of accounting for legal constructs and underscores the effects of provisional variances along with environmental requirements. The second study demonstrates the adaptation measures taken by the U.S. energy system in the face of constraints on water availability. Results show that water availability constraints may cause substantial capital stock turnover and result in non-negligible economic costs for the western U.S. This work emphasizes the need to integrate water availability constraints into electricity capacity planning and highlights the state-level challenges to facilitate regional strategic resource planning. The last study assesses the potential of surface reservoir expansion for major river basins around the world as an adaption measure to secure a reliable water supply. Results suggest that conservation zones and future human migration will have a substantial, heterogeneous impact on the maximum amount of reservoir storage that can be expanded worldwide. Findings from this study highlight the importance of incorporating human development, land-use activities, and climate change drivers when quantifying available surface water yields and reservoir expansion potential. This dissertation takes an integrated holistic approach to examine water and energy system interrelationships, and assesses the role of climate change in reshaping the interconnectivity. The three studies are tied in to each other by identifying some of the challenges the society is facing in the water-energy-climate nexus (first study) and providing a few possible solutions in both energy supply (second study) and water supply (third study) sector. Novelty of this dissertation includes but not limited to 1) explicit representation of state-level environmental regulations pertaining to power plant operations in the U.S. 2) integrated approach that captures the interactions of energy system with other sectors of the economy; and 3) global assessment of reservoir capacity expansion potential with consideration of multiple constraints. General conclusions, along with further details, provide insights for sustainable resource planning and future research directions.Item INTEGRATED DYNAMIC DEMAND MANAGEMENT AND MARKET DESIGN IN SMART GRID(2014) Asudegi, Mona; Haghani, Ali; Civil Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Smart Grid is a system that accommodates different energy sources, including solar, wind, tidal, electric vehicles, and also facilitates communication between users and suppliers. This study tries to picture the interaction among all new sources of energy and market, besides managing supplies and demands in the system while meeting network's limitations. First, an appropriate energy system mechanism is proposed to motivate use of green and renewable energies while addressing current system's deficiencies. Then concepts and techniques from game theory, network optimization, and market design are borrowed to model the system as a Stackelberg game. Existence of an equilibrium solution to the problem is proved mathematically, and an algorithm is developed to solve the proposed nonlinear bi-level optimization model in real time. Then the model is converted to a mathematical program with equilibrium constraints using lower level's optimality conditions. Results from different solution techniques including MIP, SOS, and nonlinear MPEC solvers are compared with the proposed algorithm. Examples illustrate the appropriateness and usefulness of the both proposed system mechanism and heuristic algorithm in modeling the market and solving the corresponding large scale bi-level model. To the best knowledge of the writer there is no efficient algorithm in solving large scale bi-level models and any solution approach in the literature is problem specific. This research could be implemented in the future Smart Grid meters to help users communicate with the system and enables the system to accommodate different sources of energy. It prevents waste of energy by optimizing users' schedule of trades in the grid. Also recommendations to energy policy makers are made based on results in this research. This research contributes to science by combining knowledge of market structure and demand management to design an optimal trade schedule for all agents in the energy network including users and suppliers. Current studies in this area mostly focus either in market design or in demand management side. However, by combining these two areas of knowledge in this study, not only will the whole system be more efficient, but it also will be more likely to make the system operational in real world.Item Offtake Strategy Design for Wind Energy Projects under Uncertainty(2014) Zhu, Xinyuan; Cui, Qingbin; Civil Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Energy use from wind, solar, and other renewable sources is a public policy at the federal and state levels to address environment, energy, and sustainability concerns. As the cost of renewable energy is still relatively high compared to fossil fuels, it remains a critical challenge to make renewable energy cost competitive, without relying on public subsidies. During recent years, much advance has been made in our understanding of technology innovations and cost structure optimization of renewable energy. A knowledge gap exists on the other side of the equation - revenue generation. Considering the complexity and stochastic nature of renewable energy projects, there is great potential to optimize the revenue generation mechanisms in a systematic fashion for improved profitability and growth. This dissertation examines two primary revenue generation mechanisms, or offtake strategies, used in wind energy development projects in the U.S. While a short-term offtake strategy allows project developers to benefit from price volatility in the wholesale spot market for profit maximization, a long-term offtake strategy minimizes the market risk exposure through a long-term Power Purchase Agreement (PPA). With Conditional Value-at-Risk (CVaR) introduced as a risk measure, this dissertation first develops two stochastic programming models for optimizing offtake designs under short and long-term strategies respectively. Furthermore, this study also proposes a hybrid offtake strategy that combines both short and long-term strategies. The two-level stochastic model demonstrates the merit of the hybrid strategy, i.e. obtaining the maximized profit while maintaining the flexibility of balancing and hedging against market and resource risks efficiently. The Cape Wind project in Massachusetts has been used as an example to demonstrate the model validity and potential applications in optimizing its revenue streams. The analysis shows valuable implications on the optimal design of renewable energy project development in regard to offtake arrangements.Item Methods for Employing Real Options Models to Mitigate Risk in R&D Funding Decisions(2011) Eckhause, Jeremy Michael; Gabriel, Steven A; Civil Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Government acquisitions requiring research and development (R&D) efforts are fraught with uncertainty. The risks are often mitigated by employing a multi-stage competition, with multiple projects funded initially until a single successful project is selected. While decision-makers recognize they are using a real options approach, analytical tools are often unavailable to evaluate optimal decisions. The use of these techniques for R&D project selection to reduce the uncertainties has been shown to increase overall project value. This dissertation first presents an efficient stochastic dynamic programming (SDP) approach that managers can use to determine optimal project selection strategies and apply the proposed approach on illustrative numerical examples. While the SDP approach produces optimal solutions for many applications, this approach does not easily accommodate the inclusion of a budget-optimal allocation or side constraints, since its formulation is scenario specific. Thus, we then formulate an integer program (IP), whose solution set is equivalent to the SDP model, but facilitates the incorporation of these features and can be solved using available commercial IP solvers. The one-level IP formulation can solve what is otherwise a nested two-level problem when solved as an SDP. We then compare the performance of both models on differently sized problems. For larger problems, where the IP approach appears to be untenable, we provide heuristics for the two-level SDP formulation to solve problems efficiently. Finally, we apply these methods to carbon capture and storage (CCS) projects in the European Union currently under development that may be subject to public funding. Taking the perspective of a funding agency, we employ the real options models presented in this dissertation for determining optimal funding strategies for CCS project selection. The models demonstrate the improved risk reduction by employing a multi-stage competition and explicitly consider the benefits of knowledge spillover generated by competing projects. We then extend the model to consider two sensitivities: 1) the flexibility to spend the budget among the time periods and 2) optimizing the budget, but specifying each time period's allocation a priori. State size, scenario reduction heuristics and run-times of the models are provided.Item ANALYSIS AND SIMULATION OF ENERGY USE AND COST AT A MUNICIPAL WASTEWATER TREATMENT PLANT(2011) Feng, Yilu; Brubaker, Kaye L; Civil Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)The cost of electricity, a major operating cost of municipal wastewater treatment plants, is related to influent flow rate, power price, and power load. With knowledge of inflow and price patterns, plant operators can manage processes to reduce electricity costs. Records of influent flow, power price, and load are evaluated for Blue Plains Advanced Wastewater Treatment Plant. Diurnal and seasonal trends are analyzed. Power usage is broken down among treatment processes. A simulation model of influent pumping, a large power user, is developed. It predicts pump discharge and power usage based on wet-well level. Individual pump characteristics are tested in the plant. The model accurately simulates plant inflow and power use for two pumping stations [R2 = 0.68, 0.93 (inflow), R2 =0.94, 0.91(power)]. Wet-well stage-storage relationship is estimated from data. Time-varying wet-well level is added to the model. A synthetic example demonstrates application in managing pumps to reduce electricity cost.Item GUARANTEE DESIGN ON ENERGY PERFORMANCE CONTRACTS UNDER UNCERTAINTY(2011) Deng, Qianli; Cui, Qingbin; Jiang, Xianglin; Civil Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Due to the growing concerns with climate change and energy supply, Energy Performance Contracting (EPC), which uses the guaranteed future utility savings to repay the initial renovation investments, becomes increasingly popular. However, most Energy Service Companies (ESCOs) set the savings guarantee roughly based on their previous experience, which leads to inaccurate estimates in practice. This paper has built the stochastic models for the savings risks both from the energy price volatility and the facility performance instability, which follow the Geometric Brownian Motions (GBM) and Ito's lemma. Then, a flexible guarantee designing method for ESCOs is developed to minimize the financial risks and a case study has been conducted to show the application. Finally, suggestions have been made for how ESCOs set the guarantee and the extra profit sharing proportion in contracts based on the existing information. This method will help them appropriately allocate risks with successful contract negotiation.Item Multi-Period Natural Gas Market Modeling - Applications, Stochastic Extensions and Solution Approaches(2010) Egging, Rudolf Gerardus; Gabriel, Steven A; Civil Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)This dissertation develops deterministic and stochastic multi-period mixed complementarity problems (MCP) for the global natural gas market, as well as solution approaches for large-scale stochastic MCP. The deterministic model is unique in the combination of the level of detail of the actors in the natural gas markets and the transport options, the detailed regional and global coverage, the multi-period approach with endogenous capacity expansions for transportation and storage infrastructure, the seasonal variation in demand and the representation of market power according to Nash-Cournot theory. The model is applied to several scenarios for the natural gas market that cover the formation of a cartel by the members of the Gas Exporting Countries Forum, a low availability of unconventional gas in the United States, and cost reductions in long-distance gas transportation. The results provide insights in how different regions are affected by various developments, in terms of production, consumption, traded volumes, prices and profits of market participants. The stochastic MCP is developed and applied to a global natural gas market problem with four scenarios for a time horizon until 2050 with nineteen regions and containing 78,768 variables. The scenarios vary in the possibility of a gas market cartel formation and varying depletion rates of gas reserves in the major gas importing regions. Outcomes for hedging decisions of market participants show some significant shifts in the timing and location of infrastructure investments, thereby affecting local market situations. A first application of Benders decomposition (BD) is presented to solve a large-scale stochastic MCP for the global gas market with many hundreds of first-stage capacity expansion variables and market players exerting various levels of market power. The largest problem solved successfully using BD contained 47,373 variables of which 763 first-stage variables, however using BD did not result in shorter solution times relative to solving the extensive-forms. Larger problems, up to 117,481 variables, were solved in extensive-form, but not when applying BD due to numerical issues. It is discussed how BD could significantly reduce the solution time of large-scale stochastic models, but various challenges remain and more research is needed to assess the potential of Benders decomposition for solving large-scale stochastic MCP.Item Design and testing of a microbial fuel cell for the conversion of lignocellulosic biomass into electricity(2010) Gregoire, Kyla Patricia; Becker, Jennifer; Civil Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Previous research has demonstrated that microbial fuel cells (MFCs) have the ability to degrade soluble substrates such as wastewater; however, very few studies have attempted the conversion particulate biomass to electricity in an MFC. A single-chamber, air cathode MFC was developed using a solid, lignocellulosic substrate (corncob pellets) as the electron donor. The first trial, using a prototype reactor with a graphite rod anode, ran for 415 hours, and generated a maximum open circuit voltage and current of 0.67 V and 0.25 mA, respectively. The second trial employed graphite brush anodes and multiple microbial inocula. A pasteurized soil inoculum resulted in negligible power (P = 0.144 mW/m3). The addition of rumen fluid, which naturally contains cellulose-degrading microorganisms, and Geobacter metallireducens, resulted in Pmax values of 77 mW/m3 and 159 mW/m3, respectively. Analysis of hydrogen, methane, organic acids, and the mass of substrate consumed provided insight into the relationship between cellulose oxidation, methanogenesis, and power production.