We use models to simulate the real world mainly for prediction purposes. However, since any model is a simplification of reality, there remains a great deal of uncertainty even after the calibration of model parameters. The model’s identifiability of realistic model parameters becomes questionable when the watershed of interest is small, and its time of concentration is shorter than the computational time step of the model. To improve the discovery of more reliable and more realistic sets of model parameters instead of mathematical solutions, a new algorithm is needed. This algorithm should be able to identify mathematically inferior but more robust solutions as well as to take samples uniformly from high-dimensional search spaces for the purpose of uncertainty analysis. Various watershed configurations were considered to test the Soil and Water Assessment Tool (SWAT) model’s identifiability of the realistic spatial distribution of land use, soil type, and precipitation data. The spatial variability in small watersheds did not significantly affect the hydrographs at the watershed outlet, and the SWAT model was not able to identify more realistic sets of spatial data. A new populationbased heuristic called the Isolated Speciation-based Particle Swarm Optimization (ISPSO) was developed to enhance the explorability and the uniformity of samples in high-dimensional problems. The algorithm was tested on seven mathematical functions and outperformed other similar algorithms in terms of computational cost, consistency, and scalability. One of the test functions was the Griewank function, whose number of minima is not well defined although the function serves as the basis for evaluating multi-modal optimization algorithms. Numerical and analytical methods were proposed to count the exact number of minima of the Griewank function within a hyperrectangle. The ISPSO algorithm was applied to the SWAT model to evaluate the performance consistency of optimal solutions and perform uncertainty analysis in the Generalized Likelihood Uncertainty Estimation (GLUE) framework without assuming a statistical structure of modeling errors. The algorithm successfully found hundreds of acceptable sets of model parameters, which were used to estimate their prediction limits. The uncertainty bounds of this approach were comparable to those of the typical GLUE approach.

The Tsallis entropy is applied to derive both 1-D and 2-D velocity distributions in an open channel cross section. These distributions contain a parameter m through which the Tsallis entropy becomes a generalization of the Shannon entropy. Different m parameter values are examined to determine the best value for describing the velocity distribution.Two Lagrangian parameters that are involved in the final form of 1-D velocity distribution equation are determined from observations of mean velocity and the maximum velocity at the water surface. For channels which are not wide and where the maximum velocity does not occur at the water surface, a 2-D velocity distribution is more appropriate. The Tsallis entropy is applied to derive 2-D velocity distributions. A new parameter M is introduced which represents the hydraulic characteristics of the channel. The derived velocity distributions are verified using both field data and experimental data. The advantages are found by comparing with Parandtl-von Karman, power law and Chiu’s velocity distributions.

The advent of smart well technology, which is the use of down hole sensors to adjust well controls (i.e. injection rate, bottomhole pressure, etc.), has allowed the possibility to control a field in all stages of the production. This possibility holds great promise in better managing enhanced oil recovery (EOR) processes, especially in terms of applying optimization techniques. However, some procedures for optimizing EOR processes are not based on the physics of the process, which may lead to erroneous results. In addition, optimization of EOR processes can be difficult, and limited, if there is no access to the simulator code for computation of the adjoints used for optimization. This research describes the development of a general procedure for designing an initial starting point for a surfactant flood optimization. The method does not rely on a simulator's adjoint computation or on external computing of adjoints for optimization. The reservoir simulator used for this research was Schlumberger's Eclipse 100, and optimization was accomplished through use of a program written in Matlab. Utility of the approach is demonstrated by using it to optimize the process net present value (NPV) of a 5-spot surfactant flood (320-acres) and incorporating the optimization solution into a probabilistic geological and economic setting. This thesis includes a general procedure for optimizing a surfactant flood and provides groundwork for optimizing other EOR techniques. This research is useful because it takes the optimal solution and calculates a probability of success for possible NPVs. This is very important when accessing risk in a business scenario, because projects that have unknown probability of success are most likely to be abandoned as uneconomic. This thesis also illustrates possible NPVs if the optimal solution was used.

The Method of Manufactured Universes is presented as a validation framework for uncertainty quantification (UQ) methodologies and as a tool for exploring the effects of statistical and modeling assumptions embedded in these methods. The framework calls for a manufactured reality from which "experimental" data are created (possibly with experimental error), an imperfect model (with uncertain inputs) from which simulation results are created (possibly with numerical error), the application of a system for quantifying uncertainties in model predictions, and an assessment of how accurately those uncertainties are quantified. The application presented for this research manufactures a particle-transport "universe," models it using diffusion theory with uncertain material parameters, and applies both Gaussian process and Bayesian MARS algorithms to make quantitative predictions about new "experiments" within the manufactured reality. To test further the responses of these UQ methods, we conduct exercises with "experimental" replicates, "measurement" error, and choices of physical inputs that reduce the accuracy of the diffusion model's approximation of our manufactured laws. Our first application of MMU was rich in areas for exploration and highly informative. In the case of the Gaussian process code, we found that the fundamental statistical formulation was not appropriate for our functional data, but that the code allows a knowledgable user to vary parameters within this formulation to tailor its behavior for a specific problem. The Bayesian MARS formulation was a more natural emulator given our manufactured laws, and we used the MMU framework to develop further a calibration method and to characterize the diffusion model discrepancy. Overall, we conclude that an MMU exercise with a properly designed universe (that is, one that is an adequate representation of some real-world problem) will provide the modeler with an added understanding of the interaction between a given UQ method and his/her more complex problem of interest. The modeler can then apply this added understanding and make more informed predictive statements.

In oil/gas and chemical industries, dynamics is one of the most essential characteristics of any process. Time-dependent response is involved in most steps of both the physical/engineering processes and the equipment performance. The conventional Quantitative Risk Assessment (QRA) is unable to address the time dependent effect in such dynamic processes. In this dissertation, a methodology of Dynamic Operational Risk Assessment (DORA) is developed for operational risk analysis in oil/gas and chemical industries. Given the assumption that the component performance state determines the value of parameters in process dynamics equations, the DORA probabilistic modeling integrates stochastic modeling and process dynamics modeling to evaluate operational risk. The stochastic system-state trajectory is modeled based on the abnormal behavior or failure of the components. For each of the possible system-state trajectories, a process dynamics evaluation is carried out to check whether process variables, e.g., level, flow rate, temperature, pressure, or chemical concentration, remain in their desirable regions. Monte Carlo simulations are performed to calculate the probability of process variable exceeding the safety boundaries. Component testing/inspection intervals and repair time are critical parameters to define the system-state configuration; and play an important role for evaluating the probability of operational failure. Sensitivity analysis is suggested to assist selecting the DORA probabilistic modeling inputs. In this study, probabilistic approach to characterize uncertainty associated with QRA is proposed to analyze data and experiment results in order to enhance the understanding of uncertainty and improve the accuracy of the risk estimation. Different scenarios on an oil/gas separation system were used to demonstrate the application of DORA method, and approaches are proposed for sensitivity and uncertainty analysis. Case study on a knockout drum in the distillation unit of a refinery process shows that the epistemic uncertainty associated with the risk estimation is reduced through Bayesian updating of the generic reliability information using plant specific real time testing or reliability data. Case study on an oil/gas separator component inspection interval optimization illustrates the cost benefit analysis in DORA framework and how DORA probabilistic modeling can be used as a tool for decision making. DORA not only provides a framework to evaluate the dynamic operational risk in oil/gas and chemical industries, but also guides the process design and optimization of the critical parameters such as component inspection intervals.

none

The transaction of agricultural products, which take place in the local farmers¡¦ market, has a variety of modes. According to Oliver. E Williamson (1991), the characteristics of transactions differ from incentives intensity, administrative control, and performance attributes. Based upon the foregoing, agricultural transaction modes in the local farmers¡¦ market distinctions into three kinds: (1) market mode (2) hybrid mode, and (3) hierarchy mode. The purpose of this study is attempting to abstract the crucial dimension of agricultural transactions for organizing some transactions this way and other transactions another. The overall objectives of this study can be said to be threefold: (1) To investigate and collect literatures about the agricultural transaction modes which are presently most popular in Taiwan. (2) To identify and distinguish these transactions into market mode, hybrid mode, and hierarchy mode as suggested by Williamson. (3) As an exploratory study, to suggest uncertainty as the crucial dimension of agricultural transactions alternatives, to demonstrate why and how uncertainty rule on the selection decision. The results from this study suggest that: (1) When the uncertainty of agricultural transactions is high, hierarchy mode appears to be feasible relatively. (2) When the uncertainty is low, market mode appears to be feasible relatively. (3) When the uncertainty is in an intermediate situation, hybrid mode appears to be feasible relatively. A case study of the broiler contracts of the Charoen Pokphand Enterprise (Taiwan) Co. shows that the relationship between feasible contract mode and uncertainty of transaction is consistent with theoretical expectation.

A contingency model describing the moderating effects of perceived environmental uncertainty on the relationship between human resource flexibility (HR flexibility) and firm performance was proposed and tested. This study aimed to examine the relationship between different dimensions of HR flexibility and firm performance and further investigated the moderating effect of environmental uncertainty on this relationship. A survey research was conducted using a sample of publicly traded firms listed in Taiwan Economic Journal data bank. Data was collected from different sources, including the opinions of CEO and HR managers in each company and the public disclosure of corporate information. Hierarchical regression analysis was used to test the hypotheses. After collecting empirical data and performing the factor analysis, five dimensions of HR flexibility, including behavior flexibility, skill flexibility, financial flexibility, functional flexibility, and market-oriented flexibility, were identified in this study. By testing Hypothesis1, results showed that skill flexibility, functional flexibility and market-oriented could predict some of the performance measures. However, behavior flexibility and financial flexibility had no significant influence on any firm performance measures. By testing Hypothesis2, three dimensions of environmental uncertainty were identified first. They were response uncertainty, effect uncertainty, and state uncertainty. Then, the results of the hierarchical regression models supported the argument that effect uncertainty positively moderated the influence of behavior, skill and functional flexibility on firm performance. But, the moderating effects of response and state uncertainty were not supported. Implications and future research directions were suggested in the final part of this study.

Original equipment manufacture(OEM) is typically the role played by the high-tech or traditional industries in Taiwan, and is marked by highly efficient vertical supply chains where division of manufacturing function is done very well. Supply chain management could help companies to integrate their business process and information systems. Moreover, what kind of strategies needed to be adopted, and what kind of information needed to be exchanged transaction. The research issues regarding supply chain strategy and information sharing become apparently important. In the study, supply chain type and supply chain uncertainty are the two major constructs that guide our research. Convergent assembly supply network is type of supply chain selected for case study, and the automotive industry is the actual case for research. The research results indicated that information sharing is most intensive between suppliers and manufacturers. The lack of coordination between production and sale in manufacturer was resulted from supply chain uncertainty which can be addresses by the aspects of procurement, manufacturing, and demand respectively. The current supply chain strategy of manufacturer has push more inventory to its suppliers. This is due to the made-to-stock(MTS) method applied by the manufacturer. In addition, continuous flow process has occupied part of the manufacturing process for automobile production, and high uncertainty resulted from this stage often led to problems for delivery of product parts in the subsequent stages. Furthermore, demand uncertainty and mass customization requirement from the manufacturer have push suppliers to get timely sale information in order to have quick response to orders. Based on the case findings, this research suggested propositions regarding supply chain strategies and information sharing for the convergent assembly supply network. Because of relatively high procurement and demand uncertainty, the MTS production strategy should be switched to assembly-to-order(ATO). In addition, it is necessary for the manufacturer to establish stable relationship with suppliers, and incorporate them into product development process. This will help the manufacturer to achieve ATO and shorten product development time due to the notion of design for assembly and concurrent engineering. For information sharing in supply chains, the focus is upon shop floor information transparency between suppliers and the manufacturer, exchange of sale information between suppliers and distributors, and exchange of information regarding quality, management capability, product development between suppliers and the manufacturer.

The cost of construction projects depends on their size, complexity, and duration. Construction management applies effective management techniques to the planning, design, and construction of a project from conception to completion for the purpose of controlling time, cost and quality. A real options approach in construction projects, improves strategic thinking by helping planners recognize, design and use flexible alternatives to manage dynamic uncertainty. In order to manage uncertainty using this approach, it is necessary to value the real options. Real option models assume independence of option holder and the impacts of underlying uncertainties on performance and value. The current work proposes and initially tests whether project management reduces the value of real options. The example of resource allocation is used to test this hypothesis. Based on the results, it is concluded that project management reduces the value of real options by reducing variance of the exercise signal and the difference between exercise conditions and the mean exercise signal.