BEES response from EPA

I am writing in response to your article in the November 1999 issue of

Environmental Building News (Vol. 8, No. 11) regarding BEES 2.0. As you may know, the U.S. Environmental Protection Agency (EPA) has played a major role in the development of BEES from its conception in the early ’90s to the present. From the beginning, our ambitious goal for the BEES project was to somehow make the complicated, expensive process of life-cycle assessment more accessible and understandable to the environmentally concerned purchaser.

Thanks to hard work and dedication of many people over the past 10 years or so, we know a considerable amount about how to assess and improve certain environmental features of products—energy efficiency or recycled content, for example. Unfortunately, we still know little about how to assess the overall environmental preferability of products—that is, how to deal with tradeoffs between the different environmental impacts of products. EPA, NIST, and others are trying to crack this very tough nut through developing evaluation tools like BEES. While I agree that there is still room for improving the methodology behind BEES, I would like to correct some erroneous impressions about this software that your article may have left for your readers.

Presenting complicated data in a straightforward way is certainly one of the most difficult tasks of anyone hoping to contribute to public understanding of the causes of environmental damage. While this is true for any single environmental impact, like global warming, combining data on multiple impacts (for example, for both global warming and acute human health effects), and communicating to the public about how to weigh the relative importance of those impacts is a much more difficult task. Yet, combining data and assigning weights to impacts are critical steps in providing useful life-cycle analyses.

In your article, you suggest that the normalization step of life-cycle assessment, in which data on different impacts is combined, would benefit from a nationwide database of environmental impact significance. We agree. Unfortunately, no such database exists. While there is now reason to be optimistic that a nationwide database of environmental impacts will be available in the next year, the BEES project team had to rely on another method of normalization—one based on an ASTM standard methodology—when it developed the structure of BEES years ago. Of course, NIST and EPA will continue to work together to refine this software as more options for life-cycle assessment methodologies become available. However, we do not see the lack of a perfect methodology as a reason for halting development of tools like BEES. Less than two years have passed since the release of the first version of this software, yet BEES has already succeeded both in engaging product manufacturers in examining the life-cycle impacts of their products and in forwarding the debate over accessible life-cycle assessment.

I disagree with your assertion that the environmental results in BEES are “highly questionable.” While there are inevitable limitations inherent in BEES, as there are in any comparative environmental assessment of building products, NIST has taken pains to document these limitations both in the software and the accompanying documentation. We are pleased with the improvements that have been made to BEES, thanks largely to lively and productive debate. Once final revisions are made to the software, it will be released late this Spring. I’d encourage your readers to order a copy of BEES 2.0 (either by calling EPA’s Pollution Prevention Information Clearinghouse at 202/260-1023 or by contacting the U.S. Green Building Council) at that time and examine it for themselves.

Ruth Heikkinen, Project Manager

Environmentally Preferable Purchasing Program

U.S. EPA

Washington, D.C.

Editors’ response:

Thank you for your comments. We certainly agree with you about the value of an initiative such as BEES and that the lack of an ideal methodology is no reason not to proceed. We stand by our concerns about the program, however, and feel that the limitations in terms of data quality and normalization methods are not made obvious to users of Version 1.0. We look forward to reviewing the next version.