SITE CURRENTLY UNDER CONSTRUCTION - WATCH YOUR STEP! 555-555-5555

New Bottom Line Volume 5.13 – Life Cycle Analysis: Only as Good as the Data

June 18, 1996

As this space has observed before, “life cycle analysis (LCA) is a key toolfor environmentally sensible design and decisions, and information is a key challenge to LCA. The concept of identifying and evaluating the “cradle to grave” environmental and resource impacts of any product or process is sound, insightful, and worthwhile–so much so that it is the subject of growing attention around th world in a wide range of industries, and a component of the emerging ISO 14000 standards. But concept is not execution, and execution is hampered not only by a lack of uniform standards, bit even more importantly by significant shortfalls in the data need to conduct meaningfully accurate LCAs. This is a big problem, because in the words of Diana Bendz of IBM “if you take even something as simple as a coffee cup–paper or plastic or ceramic– and do the analysis three times with current data, you get three different answers.” And it’s a big problem because Design for Environment ultimately depends on accurate LCAs.

Consequently, the reliability of LCAs is often questionable. Brad Allenby, LLNL Director of Energy and Environment, was characteristically direct: “Show me who sponsored any LCA study,” he asserted, “and I’ll predict the conclusions.”

Lawrence Livermore National Laboratory (LLNL) recently brought together national labs, university researchers, corporate practitioners from automobiles and electronics industries, and developers of LCA software to address the challenge of “Materials Databases for Design for Environment.”

Of the many issues addressed, one was key: “How do you know?” A classic example of “garbage in garbage out,” LCAs are dependent on information that is as yes incomplete, imprecise and unstable. One person spoke of the challenge of choosing the most environmentally acceptable solders for electronic components. Lead is problematic because of its toxicity, which is well known and amply documented. Bismuth and indium represent technically feasible alternative, at least in some applications, but data on the biological impact is extremely limited by comparison.

Here’s the problem, as framed by Allenby in an imaginary dialog between a manager and an engineer.

Manager: “We want to design and manufacture an environmentally preferable
product.”
Engineer: “OK. Do you have a list of materials I should use?
M: “No. You figure that out.”
E: “OK. Can you at least tell me which materials are environmentally better?”
M: “Uh, no.”

“If we can’t answer simple questions like that,” says Allenby, “we’re a long way from being able to put LCA to work.”

Ron Williams of GM Research offered practical concerns that echoed Allenby’s. “What makes a difference to engineers,” he noted, “is to prioritize and focus on critical questions; ask simple logical questions; provide clear useful answers; encourage action. “The commitment is there,” he concluded, “but it still comes down to the question of ‘should we use A or B?'”

And therein lay the challenge of the meeting, and the challenge facing any company attempting to practice DFE. As imperfect as the tools may be, you still have no choice but to choose…and hopefully do better next time Allenby enumerated five key characteristics of acceptable materials databases. There would need to be transparency of data and assumptions, in other words that data sources and assumptions underlying them should be readily available to any user, who could then make their own judgment of the precision of the data. They would need to be robust to change and evolution, since they certainly will change and evolve as the field and technology mature. Since not all factors can be appropriately quantified, systems should allow qualitative as well as quantitative comparisons. Because we are long way from the possibility of universally accepted
datasets (and because there were questions raised about the integrity of datasets from commercial rather than not-for-profit providers), LCA systems should allow comparison between different data sets. And they should evoke and integrate values, which are always present though often unstated in design and management decisions.

It was not all as discouraging as I may make it sound, since the conference also featured a dozen examples of software tools that appear to do a creditable job of supporting a systematic evaluation of LCA–as long as users don’t fall prey to the Spreadsheet Syndrome of believing the results just because they appear structured and precise, whether the underlying logic supports that or not.

Also notable are a number of collaborative efforts, including an industrial guidance group coordinated by the Pacific National Laboratory, including industry associations form aluminum, petroleum, plastics, chemical and other industry associations; a program of automakers, and aluminum, plastic and steel industries to share life cycle databases, and commitment by the big three automakers to use these in decision making; and of course the continuing efforts of the International Organization for Standardization (ISO).

(c) 1996 Gil Friend. All rights reserved.

New Bottom Line is published periodically by Natural Logic, offering decision support software and strategic consulting that help companies and communities prosper by embedding the laws of nature at the heart of enterprise.

Gil Friend, systems ecologist and business strategist, is President and CEO of Natural Logic, Inc.

May be posted intact–including this notice–in any non-commercial forum.
Please inquire at “reprint_rights at natlogic dot com” before reproduction in any commercial forum.