Regulation of “toxicity” in the State of California has generally been following 3 State Water Resources Control Board (State Water Board) precedential decisions that require dischargers under an NPDES permit, which have a demonstrated reasonable potential to cause or contribute to an instream exceedance for toxicity, to receive a narrative effluent limitation for chronic toxicity, along with a numeric trigger that requires accelerated monitoring and a special study to attempt to determine the cause of any toxicity. This practical process has achieved the goal of discovering, investigating, and in many cases resolving potentially toxic discharges for the better part of 15 years, and importantly, the methods used were approved under the federal Clean Water Act and U.S. EPA’s implementing regulations.
However, more recently, some regional water boards veered from the mandates of the State Water Board’s precedential orders, and instead imposed a U.S. EPA approach found in guidance, but not adopted regulations, which utilizes a Pass/Fail assessment using a bioequivalence approach called the “Test of Significant Toxicity” or “TST.” The State Water Board now intends to adopt this informal approach, itself an arguably underground “rule,” as new statewide policy. This action presents a number of legal and policy issues, as the draft policy differs from federally promulgated test methods for toxicity set forth in 40 C.F.R. Part 136 in the following substantive ways:
Different null hypothesis. Instead of following U.S. EPA rule requirements that utilize a hypothesis presuming that a water sample (discharge or receiving water) is not toxic until proven to be toxic, the State Water Board’s new policy flips that presumption and initially presumes all water to be toxic until proven to be not toxic.
Different test endpoint. Instead of using one of the prescribed test endpoints in the U.S. EPA rules (e.g., the NOEC/LOEC or EC/IC25), a new endpoint of Pass/Fail is being proposed in the new policy despite U.S. EPA previously discouraging the use of Pass/Fail because it lacks sufficient information on the dose-response relationship. U.S. EPA’s 2002 promulgated methods (EPA-821-R-02-013) and guidance (EPA/600/R-95/136) both state: “Use of pass/fail tests consisting of a single effluent concentration (e.g., the receiving water concentration or RWC) and a control is not recommended.”
Different statistical approach. Instead of using one of the four specified hypothesis testing statistics set forth or incorporated by reference in U.S. EPA’s Part 136 promulgated methods, the new policy proposes the TST statistical approach. Relying upon one sentence in U.S. EPA test manuals, and ignoring the other context in the same paragraph, the draft policy promotes an unpromulgated statistical approach.
Different use of dose concentration-response curves. Instead of requiring the quality assurance steps previously praised by a federal judge, the draft policy removes a safeguard to reduce the likelihood that random “noise” in a biological test on live organisms will result in a false positive result. The new policy still requires that multi-concentration tests be conducted, but prevents use of the important information that might be gleaned. The policy instead relies on just two concentrations (the test and the control), which is not authorized under U.S. EPA rules without an approved Alternative Test Procedure (ATP). ATP approval for the two-concentration TST approach was withdrawn by U.S. EPA Region IX in 2017 after litigation challenging that ATP was filed. USEPA has determined that failing to incorporate concentration-response evaluation can more than double the single test false positive error rate so that it can no longer be presumed to be as low as 5%.
Different compliance approach. Instead of relying on multiple tests to prove persistent toxicity that could translate into instream impacts, the proposed Maximum Daily Effluent Limitation (MDEL) can be deemed a violation from a single test, which is explicitly discouraged by U.S. EPA. Such a policy also increases the likelihood of violations for a false “Fail” result, which is anticipated to occur statistically at least 5% of the time, and with certain test species, may be much higher. In addition, an MDEL makes no logical sense when the test itself takes up to 9 days to conduct.
The draft policy also omits a substantive implementation plan in favor of increasing monitoring if toxicity is observed (the outcome or toxic presumption can change merely by doing additional tests (replicates)), which calls into question the validity of the initial results.
In sum, the draft policy appears to be a solution without an identified problem. As noted by the draft policy, many regions of the State have very little toxicity detected under the existing toxicity assessment approach, yet the draft policy proposes a new approach that could promote toxicity being identified more frequently, or improperly. Even more concerning is that the policy appears focused only on toxicity from large municipal wastewater dischargers, many of which have substantially upgraded their treatment systems to recycle water, and exempts the policy from applying to smaller dischargers and non-NPDES discharges. Thus, the policy does not achieve the goal of statewide consistency, and could result is disparate identification and enforcement of toxicity-related issues.
Comments on the draft policy are currently due December 7, 2018, although many have requested an extension due to the hundreds of pages of new information. A State Water Board workshop is currently scheduled for 11:00 a.m. on November 28th in Sacramento at the Cal/EPA building, and an adoption hearing is anticipated in the Spring of 2019. Commenters should consider encouraging the State Water Board to affirm use of the existing, mandated and proven Part 136 methods used in nearly all states in this country, and to suspend implementation of the TST until adopted by U.S. EPA as a federally approved method and/or a new ATP is approved authorizing use of the divergent approach proposed.