Project 18A: Fuel Cell Balance of Plant
The objective of this project is to determine and optimize a fuel cellís efficiency by modeling the interconnections of fuel cell components. A fuel cell power plant consists of many different subsystems and components that work together to achieve the power produced by the fuel cell. These components supply the fuel cell stack with air, humidity, and fuel, as well as maintain the temperature of the fuel cell stack. Supplying these fluids can require the use of compressors or pumps, thus using some of the power the fuel cell produces. Through modeled analysis and simulation, this parasitic power requirement can be minimized while optimizing the power produced by the fuel cell.
The balance of plant study includes the balance of the fluid properties throughout the system. Humidification is a large part of the study because the inlet air and fuel must be humidified to keep the conductivity of the fuel cell membrane constant and to prevent the membrane from drying out. Because the exhaust of the fuel cell stack contains a high concentration of water, it is used to humidify the incoming air and fuel. This recirculation requires a study of the fluid properties, the electrochemistry in the stack, and pump performance.
Modeling of components includes thermodynamics equations, fluid flow equations, vendor performance data, experimental data, and empirical data from tested systems. The focus is on both general and specific fuel cell systems. By identifying and separating parameters that apply to most fuel cell plants and those that are specific to individual plants, a generic model can be generated with adjustable parameters to be applied to any specific system. The specific system targeted in this project is an off-road vehicle platform from John Deere ePower Technologies.
The modeling of these systems is done in MatLab Simulink, and then integrated into National Instruments LabView to present a user interface where input conditions can be changed to instantly simulate the system and return results. Below are examples of the steady state model in both of these programs. A dynamic model is also being developed with special attention in areas that require a controlled response to increasing and decreasing power loads.
Figure 1: Plant Diagram
Figure 2: SimuLink Model-Steady State, showing the front page with system parameters and output results. The left side shows the different component blocks inside the system block.
Figure 3: LabView Virtual Instruments dashboard presents a familiar graphic user interface.