Balance of plant calculations of an integrated system using the waste heat from a solid oxide fuel cell (SOFC) powered data center to cool its servers.
To skim on the electricity cost of traditional mechanical vapor compression air conditioner, data centers have come up with many solutions. In a cool climate, outside air can be used to cool the servers. In a hot and dry environment, evaporative cooling provides cool air by spraying outdoor air with water. Cooling tower is used to chill water for some water-cooling on top of air-cooling. However, these solutions can't be applied universally and can put a constraint in water usage.
SOFC-powered data center is poised for waste heat cooling with its large amount of medium to low grade thermal energy in the exhaust. heat2cool explores the integration of waste heat cooling and SOFC power generation in the basic level of energy & material balances, to better address the cooling and water consumption problem.
The picture below explains the crux of the cooling loop. The indoor air is evporatively cooled, cools the server rack, then the moisture is removed with desiccant to repeate the cycle of evaporative cooling and drying. The desiccant is regenerated using the high temperature exhaust gas of the solid oxide fuel cell.
solve_BoP.m calculates all streams denoted in the P&ID below, and determine the cooling you can derive from SOFC exhaust heat. matching_operation.m calculates the fuel cell operation output that matches the required cooling of servers with the waste heat cooling generated by SOFC.
Design highlights:
- Recirculation of indoor air without introducing outdoor air.
- Water recaptured by flue gas condensation to compensate for fuel cell and evaporative cooler consumption.
An example of using the MATLAB code is inside the jupyter notebook example.ipynb
- base_case.m to generate the base case condition or declare your own system inputs in a similar manner.
- solve_BoP.m will solve the balance of plant.
- matching_operation.m calculate the FC operation that matches the servers cooling need with the waste-heat cooling capacity.
The following information is needed (all temperature in Kelvin, all humidity is in fraction):
- Outdoor environment temperature: T_env
- Outdoor environment humidity: H_env
- Cold aisle temperature: T_cold_aisle
- Hot aisle temperature: T_hot_aisle
- Indoor humidity (based on cold aisle temperature): H_room