Quantifying the Value of Learning for Flexible Water Infrastructure Planning

Uncertainty in future climate change challenges water infrastructure development decisions. Flexible infrastructure development, in which infrastructure is proactively designed to be changed in the future, can reduce the risk of overbuilding unnecessary infrastructure while maintaining reliable water supply. Flexible strategies assume that water planners will learn over time, updating future climate projections and using that new information to change plans. Previous work has developed methods to incorporate learning using climate observations into flexible planning but has not quantified the impact of different amounts of learning on the effectiveness of flexible planning. In this work, we develop a framework to assess how differences in the amount of learning about climate uncertainty affect the value of flexible water infrastructure planning. In the first part of our framework, we design climate scenarios with different amounts of learning using an exploratory Bayesian modeling approach. Then, we quantify the impacts of learning on flexibility using simulated costs and infrastructure decisions. We demonstrate this framework on a stylized case study of the Mwache Dam near Mombasa, Kenya. Flexible planning is more effective in avoiding over‐ or underbuilding under high‐learning scenarios, especially in avoiding overbuilding in wet climates. This framework provides insight on the climate conditions and learning scenarios that make flexible infrastructure most valuable. Key Points We present a framework to assess how the rate of learning about climate uncertainty affects the value of flexible water infrastructure Exploratory Bayesian modeling generates climate learning scenarios with high and low potential to reduce future precipitation uncertainty In a case study in Mombasa, Kenya, flexible water infrastructure is valuable in high‐learning scenarios and wet precipitation conditions