Stochastic Optimization Approaches for Performance-based Contract Evaluation

Performance-based contracting (PBC) is a relatively new type of contracting promising to deliver higher capital equipment performance and better align the interests of customers and suppliers. In PBC, unlike traditional contracting, the customer pays the supplier for the performance of the asset, not for the resources used in the maintenance services. This dissertation applies stochastic optimization approaches for PBC evaluation to investigate the contractual elements quantitatively and help the contractual parties in three distinct ways. Firstly, we use stochastic programming with simple recourse on the side of the customer. We consider scenario-based stochastic programming models for capital equipment consisting of a single component to evaluate PBC alternatives. The model measures the deviation of implied operational availability and realized operational availability within a limited budget. The readiness risk of contracts is also computed by determining conditional value-at-risk to provide insight for the customer. Secondly, we evaluate PBC from the supplier's standpoint using real options theory. We use approximation scheme to decide the best time to replace the asset. The supplier's objective is to maximize his revenue by determining the optimal asset replacement time in the face of uncertainty in asset performance. Asset performance is assumed to follow a stochastic process, and we use stochastic dynamic programming to calculate the option price. Finally, we address the problem of PBC evaluation by considering a principal-agent model for structuring PBC along with two other common types of contracts: fixed-cost and cost-plus contracts. We apply stochastic mathematical programming with equilibrium constraints to evaluate the combination of contracts. The risk of cost uncertainty is incorporated into the model to relieve the impact of extreme and unexpected events that may happen in practice. With these three models, the dissertation demonstrates how PBC evaluation can be conducted by considering uncertainty in asset performance and contract parameters; it shows how each party can achieve the best outcome.Ph.D