For almost 20 years, U.S. government officials relied on an approach called "risk assessment." However, in recent years a distinctly different approach - called "the precautionary principle" - has emerged.

If a federal judge approves the GEMS plan, slightly radioactive water from a sewage treatment plant will be released into the Delaware River, and slightly radioactive sludge will be incinerated during winter and spread onto local land during summer. The net effect will be a steady release of radioactivity from GEMS into the local environment for at least 30 years.

Using risk assessment, the U.S. Environmental Protection Agency argues that pumping radioactive water through 17 miles of sewer lines into a sewage treatment plant, then into local farmland, air and water, would be perfectly safe because the radioactivity would meet federal standards for drinking water.

To set the "allowable" level of radioactivity in drinking water, government scientists calculated the number of cancers likely to occur if a population of average people drank such water for a lifetime, and they judged that number of cancers "acceptable." In other words, they asked, "How much harm is allowable?" And they calculated that the GEMS radioactive discharge would fall within the "acceptable" limit for harm.

The "precautionary approach," instead, asks, "How much harm is avoidable?" The precautionary approach urges us to adopt the least harmful solution and can be summarized as better safe than sorry.

The precautionary approach introduces an element of humility. It recognizes that our knowledge is limited. It assumes that our risk assessments can't give us a complete picture of the dangers. For example, we don't know where the radioactivity would travel. In the river, and eventually the bay, would it enter fish and shellfish? Once in the air, would it enter our lungs? No one can say.

In the 30-year life of the operation, a substantial amount of radioactivity would be released. If we don't know where it would travel, or whether it would react with other contaminants to cause unanticipated harm, how can EPA scientists be sure it's a good idea? They can't. They hope it's a good idea. And based on that hope, they judge it acceptable to impose those risks on the public. If their judgment turns out to be wrong, there will be no turning back - no way to retrieve the radioactivity.

Not long ago, EPA scientists assured us that certain levels of toxic lead, mercury and PCBs were safe, and they now acknowledge that they were wrong. As time passes, they set stricter standards. In recent decades, government scientists have reduced the levels of radioactivity in drinking water that they consider safe.

We all want decisions based on the best available science. But we shouldn't fool ourselves about what science can and cannot tell us. Scientists still have a great deal to learn about pollutants released into ecosystems like the Delaware River, and even how they act inside the human body. Rather than claim that our risk assessments can tell us how much radioactivity is "safe" to discharge into the local environment, doesn't it make more sense to look for solutions that avoid these risks?

The Nuclear Regulatory Commission recommends this approach for radioactivity. It acknowledges that we can't ever be 100 percent sure how much radioactivity is safe. And so the commission recommends that we keep all radioactive exposures "as low as reasonably achievable." The EPA acknowledges that its goal for radioactivity in drinking water is zero because all exposure is assumed to cause some harm.

The precautionary approach urges us to look at all available alternatives, selecting the least risky. In the case of GEMS, one alternative stands out: The radioactive water could be treated on-site and the radioactivity contained at GEMS. Under this option, the GEMS radioactive discharge would be kept as low as "reasonably achievable."

On-site containment would avoid risks that might arise from a sewer- line break or other accidental spill. And it would avoid risks from radioactivity getting into fish or other parts of the Delaware's food chain.

If we allow GEMS to discharge radioactivity into the Delaware and it causes problems that we didn't anticipate, there will be nothing we can do about it. An ounce of prevention really is worth a pound of cure.