Big oil, tiny barons: microbes can unleash trapped petroleum.

Supply shortages have pushed oil prices above $70 per barrel, but nearly 380 billion barrels of crude oil--in the United States alone--are stuck in the pores of rocks or on the surfaces of sand grains. A new study proves the feasibility of using specialized microbes to lift trapped oil that's inaccessible to current pumping technologies.

Several decades ago, researchers found that bacteria in the genus Bacillus produce detergent molecules as waste. Some preliminary lab and field studies suggested as adding these microbes to oil wells could release significant amounts of trapped oil in the same way as detergent lifts stains out of clothing. However, other work showed that the microbes had no effect, says microbiologist Michael McInerney of the University of Oklahoma in Norman.

McInerney and his team have now conducted a rigorous test of microbial activity both in the lab and in some small oil wells.

The researchers first injected mixtures of cultured Bacillus bacteria and nutrients into sandstone or sand-packed columns containing entrapped oil. They found that when the organisms produced a detergent concentration of at least 60 milligrams per liter (mg/1) of culture, the microbes unleash up to 40 percent of the trapped oil.

Last summer, the scientists moved their experiments into five small, nearly spent oil wells located near the town of Oil Center, Okla. The researchers shut off the oil pumps and injected test solutions into the wells. Two of the wells received hundreds of billions of Bacillus bacteria, along with carefully measured nutrients including sugars, nitrogen, and other minerals. Two other wells received only the nutrients, and a fifth well got an injection of only water.

After 4 days, McInerney's team turned the pumps back on and took samples of liquid coming from each well. The researchers reported this week at the annual meeting of the American Society for Microbiology in Orlando, Fla., that the two wells that had received the microbes yielded live microbes. Detergent there measured about 90 mg/1, well above the threshold for oil removal shown in the lab. The tests revealed no Bacillus bacteria or detergent in the other wells.

McInerney notes that oil flow increased slightly in the presence of the microbes in the early results, but because of mechanical problems with the restarted pumps, the team didn't collect data over a long period.

"This is a proof of principle," McInerney says.

The team's next step, says McInerney, is to measure longer-term oil production from small microbe-treated wells. Then, the researchers will turn to larger wells.

The new study impresses oil field microbiologist Gerrit Voordouw of the University of Calgary. "In this study, they've gone the additional distance in verifying [microbial activity] in the oil wells," he says. Previous studies of oil wells had been "quick and dirty," he adds.