Manipulated data in Shell's Benzene Historical Exposure Study.

In 1983, in the face of mounting evidence of excess leukemia among workers at Shell Oil's Wood River (IL) and Deer Park (TX) petroleum refineries, Shell initiated the Benzene Historical Exposure Study (BHES). Shell's prior research had implicated occupational exposure to benzene as the source of the excess leukemia. The BHES report submission, which ultimately found no link between exposure and the excess morbidity, coincided with OSHA's planned hearings over a new regulatory standard for benzene. Over the next two decades, Shell published several papers based on or expanding the BHES data, all of which concluded that the excess of leukemia was unrelated to benzene. A review of the raw data on which Shell and its consultants relied reveals that Shell manipulated and omitted data in order to reach conclusions that exculpated it from liability and helped delay stricter benzene regulation.