Planning for Surprise: Water Resources Development Under Demand and Supply Uncertainty I. The General Model

• Published in: Management science 1989-02, Vol.35 (2), p.149-163
• Main Authors: Erlenkotter, Donald, Sethi, Suresh, Okada, Norio
• Format: Article
• Language: English
• Subjects: Applied sciences; Buildings. Public works; Completion costs; Cost functions; Determinism; Economic benefits; equipment planning: capacity expansion; Exact sciences and technology; facilities; facilities/equipment planning: capacity expansion; Forecasting models; Groundwater; natural resources: water resources; probability: stochastic model applications; Supply and demand; Uncertainty; Water consumption; Water management; Water resources; Water supply; Water supply. Pipings. Water treatment
• ISSN: 0025-1909; 1526-5501
• DOI: 10.1287/mnsc.35.2.149

A study of water resources development in South Sweden revealed that water consumption unexpectedly stopped growing after construction had begun on a large project to expand water supplies. The project had a long lead time for completion, and planning had been based on traditional deterministic forecasts of future water use. To explore the implications of such an uncertain structural shift in future consumption patterns, we develop a model for determining the timing for initiating such a project. In this model, the original consumption forecast may be disrupted at some random future time by an event called a "surprise." As in the Swedish situation, we have modeled the surprise as an unexpected stagnation in future demand growth and benefit levels. We show that this form of uncertainty makes desirable delaying the initiation of the project beyond the time optimal under the original forecast. We then extend the model to include the lead time for the project as a variable, and demonstrate that the lead time should be reduced from that optimal under a deterministic future. Finally, we incorporate the option of abandoning the project when the surprise occurs, and show that higher salvage values for the project lead to earlier optimal commitment times.