My work so far has focused on the following specific areas of research:
- Solid oxide fuel cells (SOFCs) :
These high temperature fuel cells offer the highest energy conversion efficiencies and are fuel flexible. We are developing first principles models at the smallest scales (Å to μm) to improve the fundamental understanding of the highly coupled physical and chemical processes in these devices.
Current projects:- Detailed reaction models for SOFC electrochemistry : We are working on continuum-scale models which couple surface transport in the active region of SOFC electrodes with a detailed microkinetic description of the electrochemistry.
- Control relevant modeling of SOFCs : We are developing coarser-scale cell, stack and system models that can be used to simulate the dynamics of these devices in order to facilitate the design of model predictive controllers for solid oxide fuel cell systems.
- Sulfur poisoning of SOFCs : We are modeling the sulfur poisoning of fuel-side (electro)chemistry in an SOFC anode and the direct simulation of the impact of this poisoning on anode performance.
- First principles catalysis and ab–initio thermodynamics :
Most chemical transformations in nature as well as industry would not be possible without the various catalysts that allow these reactions to proceed at ‘reasonable’ rates. Density functional theory (DFT) is a standard computational chemistry tool that is used to compute reaction energetics for surface reactions on heterogeneous catalysts. However, the reaction energetics obtained directly from DFT are the ‘zero Kelvin electronic energetics’ which need to be corrected for real world reaction conditions (temperature and pressure). The combination of DFT calculations with methods from statistical mechanics to obtain corrected reaction energetics is a relatively young research area commonly known as ab-initio thermodynamics.
Current projects:- Ab–initio thermodynamics of H2S–H2–Ni : We are examining the sulfur poisoning of Ni using ab–initio thermodynamics, and through our work are extending the technique to incorporate competitive adsorption as well as the effect of surface coverage effects on the energetics.
- Real time optimization :
It is easy to imagine that at any point in time there is an optimal manner in which we should run an industrial process: the idea of optimality is usually seen as the operating point where one makes the most profit or uses the least resources, subject to constraints imposed by the physics of the process, the product demand, and the availability of resources. This definition can be mathematically framed as an optimization problem and solved, in most cases, using standard available techniques. Real Time Optimization (RTO) is the set of techniques that are designed to track any changes in the circumstances of the process e.g., a drop in ambient temperature which causes an increase in heat losses from the distillation columns in a refinery, or daily changes in the cost and prices of the feedstocks and products. While standard RTO algorithms are designed to tackle situations such as the ones above, they cannot yet directly incorporate the inherent variation in the many parameters that define a process. I am interested in methods that incorporate the uncertainty in a process directly into the process optimization problem formulation.
Current projects:- Blending optimization in petroleum refineries : In my MSc dissertation, guided by Prof. Fraser Forbes at the University of Alberta, I worked on using stochastic optimization methods to incorporate uncertainty in the quality of blend feedstocks into gasoline blending RTO (petrol is called gasoline in North America). Blending RTO calculates the most profitable mix of available feedstocks to satisfy the required demand for any blended product. Along with researchers at ABB's Corporate Research Center in Bangalore, we are now re-examining ways of including the various uncertainties inherent in a blending process.