Hybrid Turboexpander and Fuel Cell System for Power Recovery at Natural Gas Pressure Reduction Stations
MetadataShow full item record
This study investigates the performance of a hybrid turboexpander and fuel cell (HTEFC) system for power recovery at natural gas pressure reduction stations. Simulations were created to predict the performance of various system configurations. Natural gas is transported at high pressure across large distances. The pressure of the natural gas must be reduced before it is delivered to the consumer. Natural gas pressure reduction is typically achieved using pressure reduction throttling valves. In a limited number of cases pressure reduction is achieved using a turboexpander. This method has the added bonus of power generation. There is a considerable temperature drop associated with the turboexpander process. Preheating is required in many cases to avoid undesirable effects of a low outlet temperature. This preheating is typically done using gas fired boilers. The hybrid system developed by Enbridge and Fuel Cell Energy is a new approach to this problem. In this system a Molten Carbonate Fuel Cell (MCFC) running on natural gas is used in conjunction with the turbine to preheat the gas and provide additional low emission electrical power Various system configurations were simulated and factors affecting the overall performance of the systems were investigated. Power outputs, fuel requirements and efficiencies of various system configurations were found using typical gas flow variation data. The simulation was performed using input data from the current city gate pressure reduction station operated by Utilities Kingston. Using the data provided by Utilities Kingston the performance of various potential HTEFC system configurations were compared. This thesis illustrates the benefits of using this type of analysis in a feasibility study of future HTEFC systems for power recovery at natural gas pressure reduction stations. Improvements could be made to the accuracy of the simulation results by increasing the complexity of the individual component models.