Environ. Sci. Technol. 2006, 40, 6969-6975
Modeling Global-Scale Fate and Transport of Perfluorooctanoate Emitted from Direct Sources J A M E S A R M I T A G E , † I A N T . C O U S I N S , * ,† ROBERT C. BUCK,‡ KONSTANTINOS PREVEDOUROS,† MARK H. RUSSELL,‡ MATTHEW MACLEOD,§ AND STEPHEN H. KORZENIOWSKI‡ Department of Applied Environmental Science (ITM), Stockholm University, SE-10691 Stockholm, Sweden, E. I. du Pont de Nemours & Co., Inc., P.O. Box 80023, Wilmington, Delaware, 19805, and Institute for Chemical and Bioengineering, Swiss Federal Institute of Technology, ETH Zurich, CH-8093 Zu ¨ rich, Switzerland
The long-term (1950-2050) global fate of perfluorooctanoate (PFO) is investigated using the global distribution model, GloboPOP. The model is used to test the hypotheses that direct PFO emissions can account for levels observed in the global oceans and that ocean water transport to the Arctic is an important global distribution pathway. The model emission scenarios are derived from historical and projected PFO emissions solely from direct sources. Modeled ocean water concentrations compare favorably to observed PFO concentrations in the world’s oceans and thus ocean inventories can be accounted for by direct sources. The model results support the hypothesis that longrange ocean transport of PFO to the Arctic is important and estimate a net PFO influx of approximately 8-23 tons per year flowing into the model’s Northern Polar zone in 2005, an amount at least 1 order of magnitude greater than estimated PFO flux to the Arctic from potential indirect sources such as atmospheric transport and degradation of fluorotelomer alcohols. Modeled doubling times of ocean water concentrations in the Arctic between 1975 and 2005 of approximately 7.5-10 years are in good agreement with doubling times of PFO in Arctic biota estimated from monitoring data. The model is further applied to predict future trends in PFO contamination levels using forecasted (2005-2050) direct emissions, including substantial reductions committed to by industry. Modeled ocean water concentrations in zones near to sources decline markedly after 2005, whereas modeled concentrations in the Arctic are predicted to continue to increase until approximately 2030 and show no significant decrease for the remaining 20 years of the model simulation. Since water is the primary exposure medium for Arctic biota, these model results suggest that concentrations in Arctic biota may continue to rise long after direct emissions have been substantially reduced or eliminated.
* Corresponding author phone: (+46)(0) 8 164012; fax: (+46)(0) 8 6747638; e-mail:
[email protected]. † Stockholm University. ‡ E. I. du Pont de Nemours & Co., Inc. § Swiss Federal Institute of Technology. 10.1021/es0614870 CCC: $33.50 Published on Web 10/20/2006
2006 American Chemical Society
Introduction Perfluorocarboxylates, F(CF2)nCO2H, n g 7 (PFCAs), and their potential precursors are of increasing scientific and regulatory interest because they are highly persistent in the environment (1) and have been found globally in wildlife and in humans (2-9). PFCAs have been manufactured since the early 1950s and used in numerous industrial and consumer applications, largely in the northern hemisphere (10). Despite the long period of use, efforts to characterize the various global transport pathways for this group of chemicals have only recently been initiated. For example, atmospheric modeling estimated 0.4 metric tons per year of PFO may be deposited to the Arctic (65-90° N) from the global emission, distribution and degradation of 8-2 fluorotelomer alcohol assuming annual emissions of 1000 tons (11). However, that study modeled the potential contribution from only a single indirect PFO source, assumed ten times greater emissions than estimated from industry data (10), and did not attempt to estimate the PFO contribution from other sources or via other transport mechanisms. This study builds upon a recent review (10) that presented a detailed accounting of the estimated total global historic emissions of PFCAs from all known sources. The study differentiated between “direct sources” of PFO released to the environment during manufacturing and use of PFCAs and “indirect sources” occurring when PFO is present as a chemical reaction impurity or where substances degrade to form PFO in the environment. A key conclusion was that the majority (∼80%) of PFO sources to the environment are direct sources from global fluoropolymer manufacture and use. Indirect sources, including the degradation of fluorotelomer precursors (12-18), were estimated to make a minor contribution to the total global emissions of PFO. We hypothesized (10) that direct emission followed by longrange ocean water transport of PFO to the Arctic could play an important role in its global-scale distribution and fate. Here, we undertake a mechanistic investigation of global fate, ocean water transport to the Arctic, and long-term trends in fluxes and concentrations of PFO using the global distribution model GloboPOP (19). The model is evaluated against current monitoring data and applied to forecast the effects of substantial reductions in the direct environmental emissions of PFO to the environment that are currently being implemented by industry. Much has been written about the difficulties of modeling the environmental fate of perfluorinated compounds because of their unusual amphiphilic properties (e.g., 10, 20). However, we believe that preoccupation with these difficulties has hindered needed efforts to develop global-scale mass balances that evaluate the direct sources of PFO with a specific focus on oceanic environmental transport pathways. We have therefore modeled the fate of the PFO anion in the global environment using a commonly applied mass balance approach. This approach is applicable since PFO exists almost exclusively in the anionic form at environmentally relevant pH and the partitioning and degradation behavior of this anionic form can be reasonably estimated. Our goal is to compose an initial global-scale mass balance model to evaluate whether identified direct emissions of PFO from manufacture and use can account for observed concentrations of PFO in the environment. This initial mass balance provides the means to identify key aspects of the model and the emission estimates that should be refined to improve the mass balance and allow more systematic comparisons with monitoring data. VOL. 40, NO. 22, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
6969
TABLE 1. Estimated Historical (1950-2004) and Future (2005-2050) PFO Emissionsa
PFO emission source
1950-2004 min-max (metric tons)
% of total PFO emission (average)
2005-2050 min-max (metric tons)
% of total PFO emissions (average)
FP manufacturing (APFO) APFO manufacturing FP dispersion (APFO) AFFF-ECF FP manufacturing (APFN) Consumer & Industrial Products APFN manufacturing PVDF (APFN)
Direct Sources 2060-4090 370-590 215-340 50-100 3-10 2-10 1-2
