The United States military builds forward operating bases (FOBs) in remote regions. While most bases use spot generation, in which a diesel generator is attached directly to a load, microgrids that combine these individual loads and augment generators with photovoltaic (PV) cells and batteries as a distributed energy system are emerging as a safe, efficient alternative to provide energy needs to the FOB. In initial work, an optimization model was developed that produced the minimum-cost microgrid design (e.g., the combination of batteries, generators, and solar panels) and ideal dispatched power to support a FOB for one year with hourly resolution. However, this model used deterministic inputs for the load and PV resources. We develop an approach to simulate the occupancy of the FOB, which is then used to generate realizations of environmental control unit loads. These realizations share observed weather inputs with the PV model used to generate realizations of PV power output. A case study illustrates that solutions to models that account for variability in FOB occupancy in addition to weather are more likely to meet load than solutions to a deterministic model.