The South African Solar challenge is a competition that challenges international teams to design and build a vehicle powered only by solar power. This international endurance challenge is held in South Africa within the month of September. To be successful in this challenge, teams need to acquire the most distance from their solar car in the eight days. This is achieved by completing the main route as well as optional additional loops. An important part in the challenge strategies would include a model of solar radiation for September in South Africa. The purpose of this paper is twofold. Firstly, to create a model of radiation power for the time and location of the race and secondly evaluate the model. This paper shows how to create a model using historical data from GeoSun, a South African based company with the goal of acquisition and analysis of solar radiation data and solar energy. GeoSun offers a variety of services and products for solar energy projects and research. The results of this paper consist of benchmarking the created model against measurements taken with the system deployed on the actual challenge. The results indicates error percentage and actual power loss between the race data and the model, thus proving the model's use in future challenges. However, a recommendation would be to evaluate the on day cloud cover and adjust the model accordingly.