Research

My research focuses on the application of systems engineering tools to problems in engineering and biology. My research group has made theoretical contributions to real-time optimization of finite-time processes, nonlinear robust control and modeling of nonlinear processes. Applications include implementation of systems engineering tools to a pilot-scale fluidized bed hydrochlorination reactor (Mitsubishi Polysilicon), polymerization reactors (Honeywell), lab-in-a-chip (Applied Liquid Logic), protection against UV-induced skin carcinogenesis (Mitchell Cancer Institute), improvement in baffle-making process (Metalforms), and optimization of methanol production (Natgasoline). We have received funding from the National Science Foundation, Army Research Labs, Air Force Office of Scientific Research, National Institutes of Standards and Technology, Department of Education, Abe Mitchell Cancer Research Fund, Alabama Space Grant Consortium, as well the chemical industry for my research.

Recent projects from my research group are described below:

Process Design for 1,3 Butadiene Production

As cracker feedstock around the globe is trending towards the lighter feedstocks, butadiene production facilities worldwide are now run at turndown capacities. In this research, models are developed to study how operation of a conventional plant that produces butadiene from naphtha at turndown ratios of feed rates affects the purity of 1,3-butadiene. Dynamic simulations indicate that the effect of fluctuations in feed flow rate on product purity can be minimized via the use of ratio controller to change the solvent flow rate and a composition controller to change the side-draw flow rate. Next, it is shown that 1,3-butadiene can be produced from natural gas via a novel integrated plant. The manufacturing process consists of the following steps: (1) conversion of natural gas to methanol, (2) conversion of methanol to ethylene and (3) conversion of ethylene to 1,3-butadiene. The ASPEN Plus environment is utilized to simulate the overall plant and the predictive capabilities of this model is tested by comparing the results from experimental data from individual plants.

Implementation of Advanced Control

While there are several examples of industrial implementation of advanced process control technologies in the refining industry, there are very few realistic examples that demonstrate the advantages of modern control technology in natural gas production and in power generation. In this research, a hierarchical control system comprised of Dynamic Matrix Control (DMC) and basic regulatory control loops is constructed to optimize the plant operation. This methodology is applied to a natural gas dehydration plant where the objective is to optimize the plant operation in terms of reducing MEG losses and minimize the operating costs with ensured product qualities under various process upsets. Next, this methodology is tested on a dynamic model of the Allam power cycle. It is shown that this advanced control strategy is superior to PID controllers in terms of performance and profitability due to its ability to achieve constrained optimization of the economic dependent variables.

Analysis of Hydrogen Reformer for Fuel Cells

Fuel cells that utilize hydrogen are promising energy conversion units that have a high intrinsic efficiency. However, there are operational difficulties in storing hydrogen. One way to alleviate this problem is to generate hydrogen in situ from a liquid fuel such as ethanol. Previous work from our group has focused on steady state behavior of steam reformers for producing power. However, when power demand fluctuates with time as in the case of a power generator for emergency back-up power for domestic use or in a fuel-cell powered vehicle, the ethanol feed rate into the reformer also fluctuates. For this reason, it is important to study the dynamics characteristic of the reformer in order to ensure that the fuel-cell systems generates sufficient power to meet the fluctuating power demand. In this research, a mathematical model for an ethanol reformer is developed that captures the temporal and spatial variation of the species involved in the reforming reactions. The reformer is modeled as a tubular non-isothermal, non-isobaric packed-bed reactor operating at unsteady state. The partial differential equations resulting from this model are solved numerically after estimating the model parameters from the literature. Simulation results show how the flow rate of hydrogen changes with time at the reactor exit. The reformer model is coupled with a fuel-cell stack model and the effect of fluctuating power demand from an experimentally determined power profile from a 3-bedroom house on the hydrogen and ethanol flow rates are studied. Based on these dynamic studies, a battery backup system is developed that stores excess power when the ethanol feed rate exceeds the power demand and utilizes power from the battery when the reformer is unable to provide the necessary hydrogen during increasing power requirement.

Impact of the Affordable Care Act

This research examines the causal effect of expanding health insurance on diabetes incidence. Comprehensive county-level data from the United States is utilized to study the effect of Medicaid expansion on diabetes rates. The analysis is based on cross-county variation according to Affordable Care Act healthcare reforms, along with county share-eligibility variation. Difference-in-difference and triple-difference statistical regression specifications are employed to control for confounding variables. A slight negative causal effect is observed between expanding health insurance and diabetes diagnoses.

Pilot-Scale Hydrochlorination Reactor Analysis

In the hydrochlorination reaction, silicon tetrachloride (STC), metallurgical silicon, and hydrogen are converted to trichlorosilane (TCS) at about 540 oC. A pilot-scale reactor was constructed to study the yield of TCS produced by the hydrochlorination reaction. The yield observed by experimentation compared favorably with a scalable mathematical model developed to predict the rate of TCS conversion.