Cornell University and @RISK Help the LCRA Manage Water Supplies

Effective storage and distribution of water is a challenge in most regions, but in hot and arid Central Texas water supply affects every person, farm, and business. The Lower Colorado River Authority (LCRA) manages water supplies for municipalities, farms and industries along a 600-mile stretch of the Texas Colorado River which includes the state capital Austin. In a unique partnership, Cornell University Operations Research students collaborated with the LRCA to develop a water supply model using Palisade’s @RISK software.

In addition to managing the water supply and environment of the lower Colorado River basin, the LCRA plays a variety of roles in Central Texas: generating electricity, developing water and wastewater utilities, providing public recreation areas, and supporting community and economic development. Of note, the LCRA operates six hydroelectric dams and owns about 16,000 acres of recreational lands, with more than 40 parks, natural science centers and nature preserves.

The Water Management Challenge
The LCRA manages the water it has available for sale to municipal, industrial, agricultural, and other users through a series of “water rights” permits or contracts. Water supply is guaranteed to the LCRA’s one million residential users, even in times of drought, through “firm” contracts with the local municipalities. Lower priority agricultural water supply is contracted annually at the beginning of the year and is considered “interruptible.” That is, in times of low supply or drought, water use for agricultural irrigation can be interrupted. Rapid population growth combined with limited supply has made water resource management vital to the region’s well-being.

The LCRA’s challenge is to predict the amount of water available for low-priority agricultural use contracts while ensuring that municipal, industrial, and other firm water needs are met. Inflow and rainfall are uncertain, as is high priority municipal and industrial consumption. Additional uncertainties—like loss due to evaporation—must also be factored into the equation.

The Cornell-@RISK Solution
Cornell University graduate students Geeta Bhadauria, Raoul Costa de Beauregard, and Andrew Pollock, working under the supervision of Professor Huseyin Topaloglu, developed a water supply model using @RISK for Excel. According to Dr. Topaloglu, “The challenge is to ensure that demand for drinking water and other necessities are met while also providing for industrial, agricultural and recreational use.”

Ample historical data exists for uncertain factors such as water inflow, rainfall, and consumption. @RISK was able to represent these uncertainties using probability distribution functions. The Cornell team used @RISK’s distribution fitting capabilities to create highly accurate distribution functions based on the actual historical data for these uncertainties.

Through Monte Carlo simulation, @RISK predicts the amount of water available and the probabilities of different amounts occurring. This tells the LCRA how much water is available for contract each year. The model allows for optimal amounts of water to be available for interruptible contracts while ensuring sufficient water is retained in reservoirs to meet demand during low rain years.

Palisade’s @RISK was at the heart of the model. @RISK handled all the uncertainties, dealt with the variations, and delivered clear results. According to Mark Jordan, Manager of River Management at the LCRA, "As demands on our river increase, LCRA is always looking for more efficient ways to manage the water supply. This research shows the potential of new methods."

» @RISK
» Lower Colorado River Authority
» Cornell University Operations Research and Industrial Engineering