Probabilistic assessment of cost, time, and revenue in a portfolio of projects using stochastic agent-based simulation

Unforeseen conditions and uncertainties typically surround construction projects during the execution phase. These uncertainties are the main reason for schedule delays and cost overruns. Stochastic simulation models have been developed to consider projects' uncertainties and various probabilities. Although numerous studies have been conducted on causes and effects of uncertainties in individual projects, the uncertainties in a portfolio of projects have not been assessed. This study presents an agent-based simulation model developed to consider time and cost uncertainties in an owner's portfolio of projects. The model can simulate different conditions and scenarios, taking into account costs inflation during projects' execution periods. It contributes to the portfolio management body of knowledge by simulating the progress of a portfolio of projects against different scenarios, considering stochastic time, cost, and inflation rates. The simulation results assist owners to do probabilistic assessment and reduce uncertainties and risks related to time, cost, and revenue of their portfolio of projects and calculate the confidence level of their decisions and strategies. To demonstrate the capabilities of the model, an example portfolio containing 50 projects is modeled and different scenarios are evaluated. Results of the simulation indicate that the model works well and provides detailed information at the portfolio level and projects' level. Fitting probability distributions to the results, it is seen that using the most likely numbers as deterministic parameters may lead to overestimating portfolio revenue and cost. The results indicate that using the portfolio stochastic agent-based simulation model can have significant effects on organizations' predictions regarding their costs and revenues.