Debangsu Bhattacharyya

Solid Oxide Fuel Cell Modeling

This paper discusses the modeling of a solid oxide fuel cell using both lumped and distributed modeling approaches. In particular, the focus of this paper is on the development of a computationally efficient lumped-parameter model for real-time emulation and control. The performance of this model is compared with a detailed distributed model and experimental results. The fundamental relations that govern a fuel cell operation are utilized in both approaches. However, the partial pressure of the species (fuel, air, and water) in the distributed model is assumed to vary through the length of the fuel cell. The lumped model approach uses the partial pressure of the species at the exit point of the fuel cell. The partial pressure of the species is represented by an equivalent RC circuit in the lumped model.

Plant-wide control system design: Primary controlled variable selection

Abstract This work is focused on the development of a rigorous, model-based approach for the selection of primary controlled variables as part of a plant-wide control system design methodology. Controlled variables should be selected for their self-optimizing control performance and controllability while ensuring satisfactory performance in terms of dead-time and closed loop interactions. This work has considered both self-optimizing and control performance as well as has addressed issues related to loop-interactions and superstructure constraints. The new three-stage approach developed in this work results in a large-scale, constrained, mixed-integer multi-objective optimization problem. For solving this problem, a parallelized, bi-directional branch and bound algorithm with dynamic search strategies has been developed to solve the problem on large computer clusters. The proposed approach is then applied to an acid gas removal unit as part of an integrated gasification combined cycle power plant with CO 2 capture.

Dimensional optimization of a tubular solid oxide fuel cell

Abstract Solid oxide fuel cells (SOFCs) are very promising for their potential applications as power generators. However, the cost of these cells needs to be significantly reduced to make them a commercial success. Cost of the materials is a significant component of the overall cost. An improvement of the power density with respect to the weight of the cell, termed as gravimetric power density in this study, can help to achieve a lower material cost. On the other hand, a compact design is required for both man-portable and stationary powerhouse applications. The power density with respect to the overall volume of the cell is termed as volumetric power density in this study. A nonlinear constrained multiobjective optimization study using a lexicographic approach is performed to maximize the gravimetric and the volumetric power density of a tubular SOFC. The decision variables are the radius of the anode channel, the cell length, and the annulus size. To be used for optimization studies, a detailed steady state model is developed that can capture changes in the concentration, activation, and ohmic losses due to changes in the decision variables. The model is extensively validated with experimental data collected from an industrial cell spanning a wide range of temperatures, H 2 flow rates, and DC polarizations. Although the model predictions are found to be satisfactory for most operating conditions, a significant mismatch between the simulation results and the experimental data is observed when the H 2 flow rate is low. The validation study helps to identify the feasible region for the optimization study. The optimization study shows that significant improvements in both the power densities are possible for all the operating conditions considered in this study. The electrical efficiency of the cell also gets improved due to the optimization. In one of the operating conditions, about 30% improvement in the gravimetric power density and about 65% improvement in the volumetric power density are obtained due to the optimization. The percentage changes of the decision variables compared to their base case values are found to be similar for all the voltages other than the voltages close to the open circuit potential (OCP).

Dynamic modeling and system identification of a tubular solid oxide fuel cell (TSOFC)

Solid Oxide Fuel Cells (SOFCs) are high temperature fuel cells with a strong potential for stationary power house applications. However, considerable challenges are to be overcome to connect these cells to the power grid. The cells have to satisfy the changing demand of the grid without sacrificing their efficiencies and without causing any structural or material damage. Such an operation, coupled with fast and highly nonlinear transients of the transport variables, leads to a very challenging control problem. This requires an efficient and robust controller. For synthesizing such a controller, a well-validated dynamic model is essential. In this work, a dynamic model is validated by using experimental data from an industrial cell. The data are generated over a broad range of cell temperatures, reactant flow rates, DC polarizations, and amplitudes of step. In the process of validation, it is identified that the Knudsen diffusion and an extended active area for the electrochemical reactions play key roles in determining the current transients of the cell. The dynamic model is used for identification of reduced order models that can be solved in real time for implementation in the MPC framework. Several linear and nonlinear models are considered and the best model is chosen according to the AIC values of the models. Both SISO and MIMO models are identified. For the MIMO model, voltage and H2 flow are considered as inputs. Power and utilization factors are considered as outputs. A linear model such as ARX model is found to be satisfactory for most SISO cases. However, a nonlinear model such as NAARX model with more cross terms is found to improve the model performance significantly for the MIMO case. All through this work, efforts have been made to synthesize the simplest, yet representative model that can be used for real-time applications.

Optimal scheduling and its Lyapunov stability for advanced load-following energy plants with CO2 capture

Abstract In this study, a two-level control methodology consisting of an upper-level scheduler and a lower-level supervisory controller is proposed for an advanced load-following energy plant with CO2 capture. With the use of an economic objective function that considers fluctuation in electricity demand and price at the upper level, optimal scheduling of energy plant electricity production and carbon capture with respect to several carbon tax scenarios is implemented. The optimal operational profiles are then passed down to corresponding lower-level supervisory controllers designed using a methodological approach that balances control complexity with performance. Finally, it is shown how optimal carbon capture and electricity production rate profiles for an energy plant such as the integrated gasification combined cycle (IGCC) plant are affected by electricity demand and price fluctuations under different carbon tax scenarios. The paper also presents a Lyapunov stability analysis of the proposed scheme.

Innovative computational tools and models for the design, optimization and control of carbon capture processes

Abstract The development and scale up of cost effective carbon capture processes is of paramount importance to enable the widespread deployment of these technologies to significantly reduce greenhouse gas emissions. The U.S. Department of Energy initiated the Carbon Capture Simulation Initiative (CCSI) in 2011 with the goal of developing a computational toolset that would enable industry to more effectively identify, design, scale up, operate, and optimize promising concepts (Miller et al., 2014). The CCSI Toolset consists of both multi-scale models as well as new computational tools. This paper focuses specifically on the PSE-related computational tools and models that provide new capabilities for integrating multi-scale models with advanced optimization, uncertainty quantification (UQ), and surrogate modeling techniques.

Adaptive Kalman filter for estimation of environmental performance variables in an acid gas removal process

In this paper, adaptive Kalman filter (KF) algorithms are implemented in an acid gas removal (AGR) process for estimating the key environmental performance variables. It was found that by using a KF where the measurement noise covariance matrix (R) is adopted based on the residual sequence, the composition of the top and bottom streams from the H2S absorber in the AGR process could be estimated accurately even in the presence of large noise-to-signal ratio and poor initial guesses for R. Estimation accuracy of a KF, where the process noise covariance matrix (Q) is adopted, is found to be superior in comparison to the traditional KF, even in the presence of large mismatches between the linear and nonlinear models and a poor initial guess for Q.

Dynamic simulation and analysis of a solid oxide fuel cell (SOFC)

Abstract Dynamic simulation of an anode-supported tubular SOFC is performed to study the transients of current and the transport fields in the cell. The time constants of the system are studied and the possible explanations for their variation within the cell and at various overpotentials are presented. It was observed that the gain of the system varies depending upon the operating conditions of the system and the directionality of the step.

Acid gas removal from syngas in IGCC plants

Abstract In this chapter, a number of traditional as well as novel technologies for acid gas removal from synthesis gas are discussed. For the traditional technologies, physical-, chemical-, and hybrid-solvent-based technologies are discussed. A number of novel technologies are under development, mainly with the goal of reducing the cost of CO 2 capture. Some of these novel technologies are still at the lab-scale, while others are being evaluated at pilot-scale or higher. These technologies include various warm gas cleanup technologies such as those based on solid sorbents as well as those based on membrane technologies where the synergistic sorption-enhanced water-gas shift reaction is a valuable option for CO 2 capture cases. Other novel approaches under investigation are cryogenic, chemical looping, and gas hydrate technologies. Further research will help to realize the full potential of these novel technologies.

Algorithmic modelling of structural connectivity for process plants

In this paper, we propose a new systematic methodology to identify dynamic changes in the connectivity among various equipment items in a process plant. Drawing analogy to neurological systems, the paper develops a framework to obtain dynamic causal model to capture the intrinsic and stimulus-driven connectivity among equipment items. An expectation maximization algorithm following a Bayesian framework is employed to obtain the elements of the connectivity matrices using a maximum likelihood approach. The algorithm is tested on a reactor-separator system. The proposed approach can help to decompose large-scale process plants into strongly-connected and weakly-connected systems. If only strongly-connected equipment items are considered together for control structure selection, this approach can result in a computationally less intensive problem for dynamic controlled variable selection of large-scale process plants.