Banking On Safer Drinking Water

Bustling rivers such as the Mississippi and the Ohio are thesource of drinking water for millions of Americans. But theserivers are likely to contain hazardous chemicals and pathogensthat must be removed or neutralized before the water is safe fordrinking.

New research suggests that the soil alongside these channelscould help with the cleanup.

Drawing river water through the adjacent earth may strip awaysome unwanted pollutants, including harmful viruses, protozoa andbacteria, researchers at The Johns Hopkins University believe.

The idea is called river-bank filtration. Instead of pullingwater directly from the river for subsequent treatment anddistribution, utilities drill wells nearby. River water drawnfrom these wells must first pass through hundreds of feet ofsoil, which serves as a natural filter.

How effective are river banks as a cleansing agent? Scientists atHopkins have launched a three-year study to find out. Theresearchers want to determine how well river-bank filtrationremoves pathogens, such as viruses, protozoa and bacteria. Theyalso want to find out if this process removes other organicmatter that reacts during disinfection in a treatment plant toform potentially harmful by-products.

"You can surmise that going through the ground should be a bettermethod; that's your intuition," says Edward J. Bouwer, leadinvestigator in the study. "But we want to quantify how muchbetter it is."

The work is important because dangerous microbes and cancer-causing chemicals sometimes slip past traditional watertreatment. "We're worried about resistant pathogens created bythe overuse of antibiotics," says Bouwer, a professor in Hopkins'Department of Geography and Environmental Engineering. "Somebacteria are becoming more resistant to disinfection. We're alsoworried about the by-products created during disinfection. Someof them may be carcinogenic."

The Hopkins study, supported by a $300,000 grant from the U.S.Environmental Protection Agency, will focus on water drawn fromwells alongside the Wabash, Ohio and Missouri rivers near TerraHaute, Ind.; Louisville, Ky.; and Kansas City, Mo. The wells areall owned by the American Waterworks Service Co., which isproviding $256,000 in labor, equipment and other in-kind supportfor the study.

The company is testing a natural filtration method developed inEurope two decades ago to produce cleaner drinking water frompolluted rivers such as the Rhine. "The focus in the 1970s wasmainly to get rid of taste and odor problems and to removehazardous materials like pesticides and hydrocarbons," Bouwersays. "But there has been limited recent work to look at how itaffects the pathogens and by-products of disinfection. We will bedoing work that is more relevant to today's concerns."

The study will pay particular attention to disinfection by-products, a source of growing alarm. After vegetation decays,tiny bits of surviving plant material make their way into a watersource. In its natural state, this organic matter may affect howwater tastes or smells, but it poses no health risks. However,when treatment plants add a disinfectant to kill pathogens, thechemical can also react with the harmless organic matter,creating not-so-harmless by-products. For example, chlorine mixedwith the plant material can produce chloroform, a suspectedcarcinogen.

For water utilities, that poses a dilemma. "Viruses and otherpathogens in drinking water are a real threat," says Bouwer. "Butthe trade-off is, if you use disinfectants to kill them, you'regoing to create these by-products that are harmful themselves."

One solution is to remove more of the harmless plant matterbefore a disinfectant is added to the water. One way is to "feed"this material to bacteria in the soil before the water reachesthe treatment plant. Bouwer's team--which includes William P.Ball and Charles R. O'Melia, both faculty members in Hopkins'Department of Geography and Environmental Engineering--wants tofind out if river-bank filtration will accomplish this. Theresearchers will compare water samples taken directly from therivers with those drawn from nearby wells. They will alsoreplicate the treatment process in labs at Hopkins to determinethe quality of drinking water produced from both sources.

First results from the experiments are expected in about 12 to 18months. When the project is completed, the researchers hope topublish design guidelines to show other utilities how and whereto dig the most effective river-bank filtration wells.

Related Web Sites:

Johns Hopkins Department of Geography and EnvironmentalEngineering: http://www.jhu.edu/~dogee/

Edward Bouwer's Home Page: http://www.jhu.edu/~dogee/bouwer.html

Color slide available; contact Phil Sneiderman