The use of biased grids as energy filters for charged particles has been a common practice in satellite-borne instruments such as the retarding potential analyzer (RPA). RPAs are currently flown on missions such as the Communications/Navigation Outage Forecast System (C/NOFS) and the Defense Meteorological Satellites Program (DMSP) to obtain estimates of geophysical parameters including ion velocity and temperature. It has been shown previously that the use of biased grids in such instruments creates a non-uniform potential in the grid plane, which leads to inherent errors in the inferred parameters. A simulation of ion interactions with a number of configurations of biased grids has been developed using a commercial finite-element analysis software package. Using a statistical approach, the simulation calculates the transmission characteristics for positively charged ions as a function of energy and angle of attack. The deviations in the transmission characteristics from the ideal assumptions are discussed with the purpose of modifying current grid designs to better approximate the idealization. Both single-grid and double-grid systems are modeled to investigate design considerations. The transmission characteristics are used to calculate the collected flux from Maxwellian ion distributions with 3D drift relative to the instrument. Perturbations in the performance of flight instrumentation relative to expectations from the idealized RPA flux equation are discussed. Relative errors in the inferred parameters are characterized as functions of ion temperature and drift velocity, improving the current understanding of RPA data.