There are numerous reasons why modeling and simulation are important in the development of new systems. Often it is not economically feasible or safe to experiment with the actual system. Sometimes the actual system is not available or does notyet even exist. By constructing a simulation model, the feasibility and efficiency of the system can be evaluated for different configurations before system components are actually bought, assembled and connected.
Three simulation models of the oxygen production system are of particular interest. The first involves the static mass flow of CO2, CO, and 02 within the system, and is based on thermodynamic and conservation of mass principles. From this simulation we can estimate what the production rate of oxygen will be in steady state, and what various system pressures and temperatures can be expected.
The second simulation concerns the static energy flow within the system, and is based on thermodynamic and conservation of energy principles. This simulation allows conditions that were assumed in the first model to be verified, and enables an estimation to be made of the power requirements of the different system components.
The third simulation model will describe the dynamics of the system, and will be based on a modeling and design methodology involving the use of bond graphs. Bond graphs were originally introduced to model mechanical systems, and have since been adapted by Dr. Franqois Cellier to model chemical reaction systems. Chemical reaction bond graphs model dynamic chemical reactions through the use of six variables: chemical potential, molar flow, hydraulic/pneumatic pressure, volume flow, temperature, and entropy flow. This model will enable temperature and energy to be balanced for each separate subsystem of the entire system. These modular subsystems can then be connected together to form the model for the entire oxygen production plant, and valuable information can be learned about system start up and shut down. In addition, control strategies can be studied both for normal operation and for handling emergency situations.
The subsystem that performs the actual separation of the carbon dioxide into oxygen and carbon monoxide is a zirconia cell. A dynamic model for this cell will be developed based on the chemical reaction and power balance that takes place within the cell. Separate program modules will be developed for each of the system processes, and then the simulation software DYMOLAwill be used with DESIRE or ACSL to simulate a hierarchical coupling of these processes.
The static simulation models are written in Ada and are currently running on a MicroVax workstation at the University of Arizona. A remote controlling computer (another MicroVax) sends input data to the local simulation computer over an Ethernet connection. The simulation executes on the local computer and then sends its results back to be displayed.
As the simulations and actual construction of the oxygen production plant enter their mature stages, modelvalidation can be made and the simulation predictions used to give insight into which system configuration will result in the optimum input/output behavior.
Interested in reading the full paper? (2 pages, 233,956 bytes, pdf)