In 1988, EPA issued recommended national water quality criteria to protect aquatic life from aluminum. For a pH range of 6.5 to 9.0, EPA recommended that chronic levels of aluminum not exceed 87 μg/l and that acute concentrations not exceed 750 μg/l. The criteria applied across all hardness and dissolved organic carbon (“DOC”) ranges even though the toxicity of aluminum was understood generally to decrease with increasing hardness and DOC. EPA observed in 1988 that although its earlier recommendations had expressed the criteria in terms of “total recoverable” aluminum, the better approach was to measure the “acid soluble” portion of aluminum (defined as the aluminum that passes through a 0.45 micron filter after the sample has been acidified to a pH of between 1.5 and 2.0 with nitric acid). EPA’s expressed reason for choosing to measure the acid soluble portion of the aluminum was that “it will measure precipitates, such as carbonate and hydroxide precipitates of aluminum, that might exist in the effluent and dissolve when the effluent is diluted with receiving water.”
Few states initially adopted the EPA proposal. One of them that did was West Virginia, which did so in 1993. It quickly came to regret its decision. WVDEP soon identified many waters as impaired based solely on aluminum concentrations, but also discovered that many were not actually impaired. It has worked since to rescind or alter its criteria. Most recently, in 2013 WVDEP adopted hardness-based criteria, but to-date that proposal is still awaiting EPA approval. EPA has consistently cited its on-going efforts to revise the 1988 criteria as grounds for delaying any decision on WVDEP’s latest proposal.
In July 2017, EPA proposed new aquatic life aluminum criteria and put them out to public comment. On their face, without critical review, they appear easier to meet than did the 1988 criteria. Below is a table comparing the criteria:
The 2017 proposed criteria, though, rely on “multilinear regression” to compute aluminum toxicity across a range of pH, dissolved organic carbon (“DOC”) and hardness levels. As a consequence, the criteria varies within certain ranges of these variables.
We have reviewed the public comments on EPA’s proposal. Most commenters support the use of statistical tools to adjust the criteria with changes in pH, DOC, and hardness because aluminum toxicity generally decreases with increasing DOC and hardness. However, the criteria are based on toxicity tests in which the ranges of those variables were bounded. As a consequence, EPA’s proposal does not authorize adjustment of the aluminum criteria for hardness and DOC values outside the ranges evaluated in the toxicity tests even though many waters have much higher levels of DOC and hardness.
The inability to extrapolate aluminum toxicity into higher ranges of DOC and hardness prevents regulators from recognizing the likelihood that aluminum toxicity continues to decrease above the ranges used in the laboratory tests—which would allow regulators to increase the allowable concentrations of aluminum. Many commenters suggested EPA should account for a wider range of DOC and hardness. Many of the commenters believed that without those adjustments the criteria would cause many healthy waters to be needlessly declared as “impaired”—resulting in “false” 303(d) listings and inappropriately stringent NPDES permit limits.
More importantly, though, the proposed criteria are expressed as “total” rather than as “acid soluble” or “dissolved” aluminum. That proposal has been met with widespread opposition. Commenters observed that laboratory waters used in toxicity tests relied on the use of soluble aluminum salts that are designed to dissolve in water and thereby convert to a bioavailable form. By expressing the criteria in total form, EPA has falsely assumed that the aluminum particles found in natural waters are likely to become as toxic as the soluble aluminum salts used in the toxicity tests.
EPA claims that unless “total” aluminum is measured, the criteria will fail to account for aluminum that can be sorbed to clay particles or complexed to DOC and later be converted to a toxic form in the water column. Thus, claims EPA, use of “total” aluminum measures will account for colloidal forms and hydroxide precipitates of aluminum that can dissolve under natural conditions and thereby become bioavailable. In response, several commenters observe that use of the standard “acid-digestion” method to measure total aluminum uses nitric acid to lower the pH of the sample to a range which does not exist in natural waters and in doing so breaks molecular bonds that would not otherwise be broken. Thus, they claim, however valid EPA’s claims about colloidal or complexed aluminum, the use of the “total” form will significantly overestimate the fraction of aluminum that can become bioavailable under natural conditions. The same commenters oppose EPA’s efforts to justify the use of the “total” form as “conservative,” saying that EPA simply should not express the criteria in total form unless it is able to advance a test method that appropriately accounts for aluminum solubility.
We have prepared a summary of most of the substantive comments on the proposed criteria. Please contact Bob McLusky if you would like a copy.
This article was authored by Robert G. McLusky, Jackson Kelly PLLC.