A research collaboration led by Cornell University has found that collaborative partnerships that aim to spread the cost burden of water infrastructure projects among regional stakeholders often end up leaving local partners bearing the brunt of underlying supply and financial risks.

This imbalance is caused by a range of factors, from institutional complexity to hydroclimatic variability. But the researchers showed that AI-powered computing and modeling algorithms can help develop partnerships with much higher water supply benefits and a fraction of the financial risk.

To examine the trade-offs associated with these water infrastructure partnerships, researchers led by Patrick Reed, the Joseph C. Ford Professor of Engineering at Cornell Engineering University, used the Friant-Kern Canal in California’s San Joaquin Valley as a case study.

The team’s paper, titled “Resilient Water Infrastructure Partnerships in Institutionally Complex Systems Face Challenging Supply and Financial Risk Tradeoffs,” was published on August 27 in Nature communication. The lead author is Andrew Hamilton, a former postdoctoral fellow in Reed’s lab.

The impact went far beyond the Golden State, Reed said.

“This is not just a California story,” he said. “We have thousands of municipal water systems across the United States, all of which face some level of uncertainty and financial risk, and a tremendous amount of our infrastructure is aging and also facing conditions that it was not necessarily designed for.”

“How can we collaborate and unite on this? And who is going to face the risks?” Reed asked. “It’s a big issue. Our work is a step in the direction of highlighting some of the key concerns, but I think there’s a lot more to do, particularly given that most of the financial risk is going to be local.”

The combined impacts of climate change, economic growth, and regulatory changes have led to growing concerns about water scarcity in California and across the United States, increasing the need for investment in new infrastructure such as large canals and groundwater storage. A popular approach to financing large infrastructure projects is through cooperative partnerships—also called consolidation or regionalization—in which multiple water providers and users work together to finance, build, and operate shared facilities.

“When there are potentially dozens of independent regional entities managing water, the argument is that it makes sense from an efficiency and cost-effectiveness perspective to bring them together in cooperative partnerships, or to regionalize these distributed systems into a coordinated system, or to consolidate planning and investments,” Hamilton said.

But given the complexity of so many interconnected systems and the aggregated way in which benefits are calculated, it can be difficult to assess the vulnerabilities of individual partners. To better understand the conflicting goals and trade-offs, Reed, Hamilton and their co-authors — doctoral student Rohini Gupta and Greg Characklis and Harrison Zeff of the University of North Carolina at Chapel Hill — studied California’s San Joaquin Valley region, which is home to more than 4 million people and 5 million acres of irrigated farmland.

The researchers used AI-powered calculations and modeling, a “multi-objective intelligent search,” to represent system dynamics in 300,000 different candidate partnerships in more than 2,000 different scenario years with a variety of plausible climate conditions. The effort required enormous data collection and the development of the computational tools needed to get an accurate picture of the variables involved.

The team’s modeling showed that cumulative estimates of the benefits of cooperative partnerships can mask significant inequalities and negative impacts for partner water utilities that already have low financial margins.

As for the most profitable types of infrastructure projects, the researchers found that most partnerships can diversify their risks by investing in portfolios that include both canal expansion and groundwater storage, rather than relying on just one strategy to address water scarcity.

The researchers also found that, compared to the current Friant-Kern Canal partnership, their algorithm could help identify “Goldilocks” configurations that would deliver significantly more water while posing less financial risk to all participants.

The researchers are sharing the algorithm for free in the hope that water infrastructure partnerships can better understand and fund their own projects more effectively and equitably.

“The new intelligent search tools and simulation frameworks make it possible to test many different alternatives in many different scenarios and make the trade-offs clearer,” said Reed.

“We don’t want to give the impression that we know more about the system than the people who run it,” Hamilton said. “What we’re saying is that the system is complex and maybe we should step back and think about better ways to avoid unintended consequences when we make really large and in many cases irreversible investments.”