Alternatives Assessment in a Regulatory Environment
Alternative assessment (AA) frameworks can be useful for making recommendations about chemical substitutions, but how they interact with regulatory frameworks causes some product stewards to shy away. “When we’re working on compliance, we like to have a clear goal that we met this criteria and that we’re complying with a particular expectation,” said Robert Skoglund, director of product stewardship and regulatory affairs at Covestro.
Working with AAs can seem like comparing apples to oranges, but the goals is to integrate disparate data sets and make a recommendation, either “these are acceptable alternatives” or “there are no acceptable alternatives,” said Skoglund, adding that the results have to be repeatable and defensible.
To help familiarize product stewards, Skoglund outlined the key components of effective AAs during a presentation at Product Stewardship 2017. He referred to the 2014 National Academy of Science report “A Framework to Guide Selection of Chemical Alternatives” as an additional resource.
Early in the assessment, there should be a screen for health and environmental impacts. If the proposed substitution doesn’t cross the acceptable threshold, it gets flagged as not an acceptable alternative.
The first step is to formulate the scope of the assessment and the problem it’s trying to solve or the question it’s trying to answer. At this stage, it’s important to identify stakeholders. “You’re basically saying, how will decisions be made and who will influence decisions,” said Skoglund.
From a product formulation perspective, you should identify the purpose and performance of the chemical being considered for replacement. If you can’t identify why the chemical is there and what it does, you can’t assess a substitute effectively, Skoglund warned.
Another key consideration is how you will handle gaps in your data. “You need to have a strategy for handling uncertainty,” said Skoglund. Likewise, you’ll need a strategy for handling tradeoffs. For instance, is one endpoint more important than another, such as human health versus the environmental impact? Skoglund posed a hypothetical: “Are you willing to reduce water use significantly if the product is a little more hazardous, if we can control the hazards through normal industrial controls?” He also noted that economic performance is always a factor in whether a substation is viable.
As practitioners of AAs, Skoglund offered some prime directives. The first is to articulate how the relevant attributes were identified and whether they’ll significantly impact the outcomes. Focusing on hazard as the only attribute, for instance, will lead to one outcome. Focusing on multiple attributes – hazard, price, greenhouse gas emissions, etc. – will lead to different outcomes.
Along the same lines, AA practitioners should be able to explain how the attributes were prioritized: Are they considered equal, additive, etc., and how do you get from multiple attributes to a final assessment or comparison?
Finally, the data collected in an assessment must be convertible to a value that’s comparable to other values. It could be binary (yes or no); in levels (high, medium, low); or numerical (1-2-3-4). Whatever the scheme, “you have to get from a body of information to a comparable value,” said Skoglund.
Elizabeth Miesner, principal at Ramboll Environ, followed up with additional thoughts on how AAs fit into regulatory frameworks. She cited the Lowell Center Alternative Assessment Framework and the California Guide to Safer Consumer Products Regulation as reference points.
As product stewards, some components of AAs are particularly important. For instance, lack of data to properly assess alternatives raises an important question: How much do you look at a chemical before you know you’ve made a better choice? For most chemicals, data is not available on all possible hazard endpoints, so organizations have to decide upfront how many hazardous traits to examine before deciding to change a product and moving forward with a substitute, said Miesner. Even if all this data existed, organizations would still be faced with tradeoffs. If one chemical is carcinogenic and the other is hepatotoxic, how will you choose one over the other?
AAs present the most difficulty when they take a hazard-only approach. In those cases, AAs can mischaracterize risk and not even address questions of feasibility. Similarly, AAs don’t always focus on the final product and its intended use. Although there’s a certain benefit to dealing only with hazard, it’s not easy in the long run to simply take out one chemical and substitute for another if it’s not addressing risk, said Miesner.