Preserving the Delaware

Analyzing the river as an inventory control system yields a solution to an ecological problem.

October 13, 2006
Print this page

Update, April 2010: In October 2007, the Delaware River Basin Commission put the Flexible Flow Management Policy (FFMP) into place. Based on operations research by Peter Kolesar, the policy is a result of years of work and negotiations among the river’s many and often embattled constituents: New York City, four state governments, conservationists, and fishermen and residents of upstate New York.

In its first two years of operation, the FFMP has already improved fishing and reduced potentially dangerous and wasteful water spills at no risk to New York City’s water supply. Over the long run, the FFMP is estimated to improve trout habitat by more than 200 percent and add $163 million in annual economic benefits through increased fishing and boating activity and potential flood mitigation. The policy resolved a decades-long dispute over the river’s resources, and offers a model for other regions facing similar challenges with their water supply. It is being studied by concerned parties in Atlant’s ongoing water wars. Read more about the proposal’s implementation at Public Offering.

The proposal was nominated for the 2010 Franz Edelman Prize, which recognizes transformative operations research.



Two summers ago, at an end-of-season picnic near his summer home in the Catskills, Professor Peter Kolesar stopped to chat with members of a conservation group. The group was trying to protect trout in the upper Delaware River by changing the rules that govern its reservoir system, which supplies about half of New York City’s water.



Kolesar, a dedicated fly fisherman who has spent much of his career figuring out ways to implement technology more efficiently, gave the group his card. “I thought I would put in an hour or two giving them some quick advice,” he says. Instead, he has devoted thousands of hours to the project. Earlier this year, he formed a coalition that united several environmental groups behind the cause, and using advanced statistical analysis and computer simulation models, he devised a proposal to change the way the Delaware River is managed.

For decades, fishermen have complained about the policies governing the Delaware. The rules derive from a 1931 Supreme Court decree written by Oliver Wendell Holmes, which stated that New York City could divert up to an average of 440 million gallons of water a day from the Delaware River Basin for the city’s use. To protect the downriver states of Pennsylvania, New Jersey and Delaware, the decree also required New York City to release enough water from its reservoirs to maintain a minimum flow at a gauge at Montague, N.J. In the 1950s, the agreement was modified to allow New York City to divert as much as 800 million gallons a day to meet the anticipated needs of its growing population.

At the time of the first decree, there wasn’t much of a movement to advocate for the river’s environmental interests. The plan left trout — a cold-water fish — particularly vulnerable during the summertime, when the river often has low flows. “The river heats up to an extent in July and August that can very easily kill fish,” Kolesar says. On some days, parts of the upper river look like little more than a tiny stream. The low flows expose the riverbanks and strand the insect population, the main food source for trout.

The current system allows the river flows to be highly erratic — posing an even greater threat to the trout and their insect prey, both of which require a steady supply of fresh water to survive. “The decree protected New York City and the down-basin states, but the river itself, particularly the upper river, got shortchanged,” Kolesar says. “The upper river has ecological and environmental needs that are not being met.”

Working in collaboration with Columbia’s Earth Institute and Department of Earth and Environmental Engineering, Kolesar came up with a plan that addresses all of these problems. He calls his proposal an adaptive release policy because, unlike the current operating rules, it takes variable factors into consideration: the seasons, the water needs of trout and the amount of water in the reservoirs. “The basic strategy is, if you have more water, release more water,” he says. “If you don’t have a lot, you have to be more conservative. It’s essentially a feedback mechanism, and that’s what makes it work.”

His proposal draws upon production principles that have been developed over the last 40 years for application in industry. Kolesar analyzed the Delaware as if it were a Toyota or Procter & Gamble plant. Like a manufacturer, the river is a system that faces a lot of variability, such as rainfall, temperature and the amount of water that has already accumulated. “If you were manufacturing cars or detergent, your issue would be to control how much inventory you have,” he says. “In this situation on the river, the water behind the dam is the inventory.”

Kolesar and his conservation coalition — which includes the Nature Conservancy, the Delaware River Foundation and Trout Unlimited — are now at a critical stage of talks with New York State, New Jersey, Pennsylvania, Delaware and New York City. The current operating rules expire in May, and the parties to the Supreme Court decree must unanimously decide by then how to move forward. While a number of the coalition’s ideas have been adopted in principle, an agreement hasn’t been finalized, and the recent floods on the Delaware have added new complexities to the negotiations.

Until that deadline, Kolesar will be in contact with representatives of New York City and the four state governments on an almost daily basis. “We’ve already had considerable success,” he says. The mathematical models he developed helped convince authorities that his plan would preserve New York City’s water supply even in the hypothetical drought of the century. “When we first presented our idea to New York City and New York State, they said it was an interesting theory, but it would never work in practice,” he says. “Showing that our plan really would maintain the water supply even in an extreme drought really turned their thinking around.”

Peter Kolesar is special lecturer in decision, risk and operations and professor emeritus at Columbia Business School.

Peter Kolesar

Professor Kolesar studies quality management and statistical quality control as well as applications of operations research and statistics, particularly in relation to the management of production and service systems. His recent research includes building models for the analysis and design of service systems with random cyclic customer demand patterns, accelerating the implementation and effectiveness of total quality management systems and optimizing credit-screening procedures. He...

View full profilePersonal Website