Saving $100 million is no small feat but a team of Vanderbilt engineers pulled it off for Metro Nashville.

Faced with new federal water quality regulations and a state order to update its circa-1800s combined sewer system-at a cost of $825 million-and substantially curtail sewage overflows into the Cumberland River, the city contracted with an engineering firm. It in turn contacted Vanderbilt for help in determining what, exactly, the river's quality was and how badly it was damaged by the overflows.

The outcome saved the city millions and thrust the Vanderbilt engineers into the forefront of using mathematical models in determining and predicting the impact pollutants have on water quality.

Previously, no other large city in the country had successfully used such a model.

"Everybody thought the river was a lot worse than it was," said Edward Thackston, professor and chair of the Department of Civil and Environmental Engineering. "Just about every time it rained, the sewers overflowed and this untreated waste was going into the river."

To limit the impact on water quality, a yearly limit of four overflows was set by the Environmental Protection Agency, which recognized that doing away with all overflows "meant rebuilding all the sewers and storm drainage systems from each house, which is totally impossible," Thackston said. Limiting overflows is still a daunting, expensive task but one which many cities are turning to by constructing huge underground storage tanks to hold the waste until it can be minimally treated.

That's what Nashville planned to do, but the city also explored another option the EPA offered: demonstrating "that if more than four overflows a year are occurring and not harming the river or violating water quality standards, then you could have the overflows as necessary," Thackston said.

Proving that the overflows weren't doing an appreciable amount of damage is a considerable challenge, one most cities haven't pursued because it requires a lot of research and the use of an EPA-recommended mathematical model that, Thackston said, proved to be "so complex it would have taken several years of gathering data to establish a baseline to use it." Instead, Thackston and his team looked for another model, and found one devised by the Army Corps of Engineers.

In simple terms, the computer model is a series of equations "where you input data of quantities of various chemicals" and the model "predicts how they will react and affect water quality."

To predict the impact on the Cumberland from sewage overflows, the researchers intensively sampled the river and the overflows for three years to get the input data and then fed the computer program information about the river's flow rate, its chemical composition and temperature, and the characteristics and quantities of the estimated sewage overflows. The computer then took all the information, factored the various chemical reactions and generated a prediction of the river's water quality.

To Thackston's amazement, the model worked. "This model had never been used before in a combined water flow and temperature and water quality mode in an impounded river," he explained. "So nobody knew if it would be accurate enough. We tried it and the changes the model predicted would occur were almost identical to what did occur. We were so shocked I thought we'd made mistakes or someone cheated.

"Water is so complex. There are so many things going on with thousands and thousands of different chemical compounds. It's a phenomenally complex system and to boil it down to 20 or 30 processes and be able to predict what's going to happen at, say, Ashland City, is just amazing. But it worked."

The five-year effort, for which Vanderbilt was paid $350,000, included the Corps of Engineers, the U.S. Geological Survey, Tennessee water quality officials, Metro Nashville officials and other state and federal agencies.

Thackston and his team were also able to show that the heavy metals which state officials worried were flowing into the river from the sewers were coming from other sources and that high levels of bacteria were also coming from upstream of the city's outflow pipes.

"We found out," Thackston said, "that if we eliminated all the overflows you'd still violate the state's water quality standards by a factor of five or 10" because of the storm-water runoff from farms, industrial sites and homes with septic tanks.

"Why should we spend several hundred million dollars if it's not going to make the water safer to drink or swim in?" Thackston asked.

Further research is underway to find out where the additional pollutants are coming from and what can be done about it, Thackston said.

The Vanderbilt engineers' findings allowed Nashville to revise its waste treatment upgrade plan and, with state approval, delete plans to build three waste storage basins, saving the city $106 million. What is more important, the engineers found a computer model that works in predicting water quality. It's a method that stands to save hundreds of millions of dollars across the country if other cities use it to determine the extent of their water quality woes. It can also help find unidentified pollutant sources by determining more precisely how much damage is generated by known sources.

"What we did for the country was, number one, show that this new model would work and predict water quality. We also showed that the EPA's demonstration approach is a feasible and economical approach."

For more information about the Department of Civil and Environmental Engineering at Vanderbilt, visit its website at www.vuse.vanderbilt.edu/~ceeinfo/cee.htm

- Elisabeth Kauffman