Informing Environmental Decision Making by Combining Life Cycle Assessment and Risk Analysis
“It is easy to envision how life cycle thinking can inform the conduct of assessments of risk. Stepping through the life cycle of a material identifies the pertinent exposure pathways and forms of a substance and elicits the need for more detailed evaluation at particular life cycle stages to characterize impacts.”
Address correspondence to:
Dr. Jo Anne Shatkin
62 Summer St.
Boston, MA 02110
With the explosive growth of research and development of engineered nanoscale materials and their incorporation into a wide range of products, many people are calling for regulation and proactive assessments to ensure that these materials and products will not adversely affect health or the environment. Can the safety of these new substances and products be ensured? Some are more confident than others. Combining life cycle assessment (LCA) and risk analysis (RA) into a prospective and semi-quantitative analysis offers an analytical approach for early identification and evaluation of potential impacts.
International calls for integrating life cycle and risk analysis for substances have focused on the need to address the breadth of potential concerns about nanoscale materials. These calls in part reflect the limitations of past approaches to chemical and other emerging technology issues, such as biotechnology, for which it is generally recognized that early assessments to identify and address potential health and safety concerns could have led to more sustainable development. Several examples highlight the inadequacies of piecemeal responses to contaminants already detected in the environment. One such example is the prevalence of antibiotic-resistant bacteria now thriving outside of nosocomial (hospital) environments, from a combination of wide use of pharmaceutical antibiotics and consumer products containing them. Microbes have become more virulent in response, having learned to adapt and resist the antimicrobial substances.
The Challenge of Assessing Nanomaterials
One of the reasons for the extensive international discussions about health and environmental impacts of nanomaterials and nanotechnologies is the limited information currently available to conduct detailed life cycle assessments and risk analysis. The current inability to adequately measure the inputs to life cycle and risk models prevents data collection for quantitative assessments of nanomaterial impacts. So, how can these data-driven tools be useful? With many voices calling for an integration of life cycle and risk analysis to evaluate the potential impacts of nanomaterials and nanotechnologies, how can these approaches be combined to inform sound decision making for nanoscale and other emerging materials and products?
Risk analysis shares with life cycle assessment the goals of structured evaluations to understand and analyze the potential impacts of substances and products on health and the environment. Both life cycle and risk analysis are conceptual frameworks often translated into quantitative assessments, conducted by application of current knowledge and extrapolation to models. They predict the relative significance of one material, technology or product in comparison to another. Governmental bodies are incorporating risk assessments into decision making but largely do not use tools such as life cycle impact assessments (LCIA) for regulatory determinations. For nanomaterials, the current lack of data and vetted analytical tools suggests that detailed evaluations will require significant resources and time, both luxuries amid competing priorities. Alternatively, streamlined assessments, intended to prioritize concerns and inform further evaluation where warranted, can be conducted relatively simply, with the insertion of bounding assumptions where data are missing to test the sensitivity of input parameters.
Although detailed risk analysis and life cycle impact assessments require considerable data, for emerging contaminants generally the needed data are not available. Compounding this limitation is the indication that specific attributes of some nanoscale materials are important, which may require new types of data to be collected. Existing measurement methods (i.e., based on mass) may apply the wrong metrics for engineered nanoscale material characterization. For example, the large surface area relative to mass of nanoscale particles means that there is much greater and more changeable reactivity than for a similar mass of larger particles of a substance. Surface properties have been shown to affect both environmental and biological properties. A host of other characteristics, including particle size and size distribution, the level of particle agglomeration, and the purity of the material, affect the way nanoscale materials behave, whether they are taken up into biological organisms, and whether they cause toxicity (Oberdörster et al. 2005). The usual units of measurement for substances, such as mass and concentration, which do not allow the distinction of 5 nanometer particles from 25 nanometer particles of a similar chemical substance (e.g., titanium dioxide), are inadequate measures for engineered nanoparticles. The current lack of ability to predict the behavior of nanoscale particles and the products in which they are used is reflected in the limited data sets describing physical, chemical, and biological aspects of nanomaterials, including pertinent data for estimating fate, transport, persistence, and toxicity. The lack of data is compounded for nanoscale materials because of this limited understanding of the key metrics and the lack of agreed-upon and reproducible measurement methods, which are generally optical and require human quantification.
Some researchers argue that complete data sets are necessary before nanomaterials can be comfortably allowed into commercial products that will enter into the environment. An alternative approach is to begin testing life cycle and risk analysis frameworks in a semi-quantitative approach to identify where they do and do not apply and what data will be necessary to complete evaluations.
In these early phases of product development, the need to develop data requires new methods and adjustment of models. In development of any model, the question of whether particular data are worth the value of collection must be addressed. In the early stages, it is appropriate to make conservative assumptions and test them for criticality before expending major effort to gather the data and use them. A 2006 workshop jointly sponsored by the European Commission and the Project on Emerging Nanotechnologies concluded that existing ISO frameworks for carrying out life cycle assessment are appropriate for nanomaterials and nano-enabled products. The workshop report encourages interaction between risk experts and industrial ecologists and recommends that researchers proceed in the absence of data, identifying uncertainties and making data available (Woodrow Wilson Center for Scholars and the European Commission 2007).
Integrating Life Cycle Assessment and Risk Analysis
How does life cycle impact assessment benefit from risk analysis? Most current life cycle impact assessment models consider the potential for risk, calculating health and environmental risks from all associated materials. When the data are not available to make these calculations, alternative endpoints can be considered. For example, Robichaud and colleagues (2005) compared nanomaterial manufacturing to other types of manufacturing but left out the contribution of nanomaterials to overall risk, due to lack of data. Impacts can still be assessed without quantitative characterization of them. Conducting screening-level analyses can inform decision making, including which parameters are driving impacts and identifying additional data necessary for more detailed comparisons. Further, stepping through the analysis provides insights about which impacts may be the most important drivers that require more detailed analysis.
As with other emerging substances, the integration of nanomaterials into other technologies means that health, environmental, and other impacts may vary at different life cycle stages. For example, the occupational exposures associated with manufacturing nanoparticles will be vastly different from those of a downstream product manufacturer inserting a part coated with nanoparticle-infused polymer. Supply chain and life cycle exposures and risks can vary greatly. Some sources of the current uncertainty are inherent to any emerging technology. Some sources of uncertainty are unique to nanoscale materials.
Nanomaterials currently in production are being incorporated into products. Thus, detailed evaluation of the toxicity of the raw material is of limited utility for characterizing risk when exposure may be to a product that is no longer at the nanoscale. Further, the ability to identify when in a product life cycle there is potential for exposure is a key variable that is not often addressed by traditional risk analysis. Understanding the potential for environmental exposure to these materials, however, is critical for good risk analysis. The process of modeling material migration through the life cycle is thus necessary to evaluating potential impacts.
It is easy to envision how life cycle thinking can inform the conduct of assessments of risk. Stepping through the life cycle of a material identifies the pertinent exposure pathways and forms of a substance and elicits the need for more detailed evaluation at particular life cycle stages to characterize impacts. Even in the absence of dose–response data, researchers can characterize the relative contribution to risk at each life cycle stage, by focusing on exposure potential. It seems that the easiest place for a life cycle assessment–risk analysis intersection is the exposure assessment phase of a risk analysis—that is, understanding where in a product life cycle there may be exposure to nanomaterials that could result in human or environmental exposure.
Many are calling for new approaches to incorporate life cycle approaches into risk analysis (e.g., COM 2004; Davis 2007; Shatkin 2008; US EPA 2008). The power of broad assessments of technology impacts is in the comprehensiveness of potential concerns that are investigated. The limitation of this breadth, however, is the simplification of some steps. Combining life cycle assessment and risk analysis brings the best of both worlds in applying more sophisticated analysis—in breadth and depth.
Early contributions to combining life cycle analysis and risk analysis stem from the Society for Risk Analysis (SRA), among others (Evans et al. 2002; Shatkin 2006). A symposium held at SRA in 2000 made inroads into key conceptual and methodological aspects. A 2006 SRA symposium explored alternative approaches for analyzing nanotechnology risks across the life cycle. One lesson learned from the 2006 SRA symposium is that units of analysis affect the results, meaning that the endpoints differ depending on whether life cycle analysis or risk analysis is the starting point, as well as the units of the denominator (i.e., risk in terms of what?). Whether outcomes are measured per mass, per unit volume, or in broader impacts, such as impact on climate, does not answer the issue of trade-offs. Which parameters are to be compared affects the outcome and so may require some experimentation.
One key concern is the limited data to inform a science-based, data-driven analysis. In the absence of quantitative measures, experts can inform parameterization, through one of many structured workshop approaches. One of the approaches for doing so, multicriteria decision analysis (MCDA), is described in this volume as one of many ways to conduct comparative assessments. Other researchers favor structured methods, such as expert elicitation.1 These methods cannot address the inherent value judgments in comparing a breadth of impacts, such as comparing a technology's carbon footprint to its impacts on aquatic species, but can identify and compare products on a set of clearly identified criteria.
Qualitative and semi-quantitative approaches can allow transparent evaluations to identify potential impacts early, locate data gaps, and prioritize materials and technologies with the best (least) impacts on a broad set of criteria. Eventually, as the key metrics are identified, measurement methods are developed, and the data are gathered, combined life cycle assessment and risk methods will become powerful tools for comparing materials, technologies, and alternative management approaches for emerging nanomaterials and nanotechnologies. The value of a combined approach is that broad impacts on health and environment are elucidated in a structured and consistent way that allows identification of the net environmental benefits as well as risks. There is a need for international discussion about the combined adoption of these analyses, to ensure that the relevant metrics are understood and incorporated into models that not only get the analysis right but achieve the needed analysis. Further, we need to ensure that new methodology does not result in paralysis by analysis, instead incrementally developing tools and analyses that inform sound and scientifically based environmental decision making.
Editor's note: For an example of the use of expert elicitation applied to nanotechnology, see the article by Wardak and colleagues (2008) in this issue.
About the Author
Jo Anne Shatkin, Ph.D., is an expert in risk analysis and is Managing Director of the CLF Ventures, a nonprofit consulting firm affiliated with the Conservation Law Foundation, an environmental advocacy organization based in Boston, Massachusetts. CLF Ventures works with clients to implement projects that have demonstrable environmental gain and economic advantage. She recently authored Nanotechnology: Health and Environmental Risks with CRC Press.