Schrödinger’s Cat, physicist Erwin Schrödinger’s famous thought experiment from the 1930s, explored the gray area between measurements and observation in quantum mechanics. The experiment considers that if a cat and a vial of poison are together in a closed box, and the vial can spontaneously break and release the poison, do we know whether the cat is currently alive or dead? Quantum mechanics states that only when we look inside the box and make the observation (i.e., take the measurement) we truly know for certain. The crux of the thought experiment is that without observation, at any one time the cat is both alive and dead. In the real world, examples of Schrödinger’s paradox are less sinister but still apply, particularly in the world of per- and polyfluoroalkyl substances (PFAS) and drinking water.
PFAS are today’s priority target chemical group, much like organochlorine pesticides were in the 1960s and PCBs in the 1970s, with useful functionality but also with some environmental consequences. A small handful of the most well-studied PFAS have received significant regulatory attention in the United States and abroad. One of the more notable PFAS-related regulatory developments to date is the U.S. Environmental Protection Agency’s (EPA’s) drinking water lifetime health advisories (HAs) for four PFAS, perfluorooctanoic acid (PFOA), perfluorooctane sulfonic acid (PFOS), perfluorobutane sulfonic acid (PFBS), and hexafluoropropylene oxide dimer acid and its ammonium salt (collectively known as GenX chemicals). Another notable rule is the proposed drinking water maximum contaminant levels (MCLs) for PFOA and PFOS as individual chemicals and PFBS, GenX chemicals, perfluorononanoic acid (PFNA), and perfluorohexane sulfonic acid (PFHxS) as a mixture.
In June 2022, EPA announced updated interim HAs for PFOA and PFOS of 0.004 parts-per-trillion (ppt) and 0.02 ppt, respectively. Surely, EPA’s interim HA, the “national non-enforceable, non-regulatory drinking water concentration levels of a contaminant at or below which exposure for a specific duration is not anticipated to lead to adverse human health effects” are concentrations that we can accurately and routinely measure in a water sample, right? As it turns out, not really. Typical minimum reporting limits for commercial laboratories running EPA standard methods for drinking water are around the low single digit ppt level. At lower levels below a part per trillion, commercial laboratories can detect these compounds but there is more uncertainty around the actual concentration.
Laboratory method detection limits are defined as the statistically calculated minimum concentration of a compound that can be measured using a specific method with 99 percent confidence that the value is above zero. Labs determine their own method detection limits for chemicals by various analytical methods. Factors such as instrument sensitivity and sample preparation add some variability between labs, but many commercial labs are within the same ballpark. In 2009, EPA published Method 537, which validated detection limits for PFOA down to 1.7 ppt. Considering that just a few months earlier EPA issued short-term provisional HAs for PFOA and PFOS of 400 ppt and 200 ppt, respectively, Method 537 was suitable to take an accurate measurement and then compare results to those HAs. Fast forward over a decade and our measurement capabilities haven’t yet caught up to today’s new HA.
The obvious disconnect between establishing a drinking water HA below what can be reliably measured in a water sample is in how we interpret the results. If a sample came back from the lab as non-detect for PFOA and the laboratory detection limit was 0.2 ppt, does the water sample contain PFOA above the HA of 0.004 ppt? Enter Schrödinger’s PFAS. Without the capability to make the observation at that low level, one might conclude that the sample could still have PFOA above the HA as there is no observed measurement that it’s not (i.e., the cat is alive). On the other hand, because the analytical method could not measure PFOA at the HA, there is also no measurement to show that the sample concentration is below the HA (i.e., the cat is dead). In Schrödinger’s world, we just don’t know for sure since we can’t make the relevant observation.
In March 2023, EPA saved the cat, so to speak. The agency proposed a MCL goal (MCLG), “the level of a contaminant in drinking water below which there is no known or expected risk to health,” for PFOA and PFOS of zero ppt. However, since we can’t measure down to zero ppt, EPA recognized that an enforceable MCL set at a concentration we can measure is appropriate. Considering current analytical methods, best available treatment technologies, and costs, EPA proposed a MCL for PFOA and PFOS of 4 ppt each.
Notably, analytical methods are the current bottleneck for setting the MCL any lower. EPA’s Fifth Unregulated Contaminant Monitoring Rule (UCMR5), published in 2021 and running from 2023 through 2025, is a large national drinking water sampling campaign for several PFAS, including PFOA and PFOS. In preparation for UCMR5, commercial laboratories submitted qualifications of their ability to achieve the lowest possible reporting limits. Then, with some statistical processing, EPA concluded that 4 ppt for PFOA and PFOS was predicted to be attainable by 75 percent of laboratories. This combined with EPA’s assessment that current drinking water treatment technologies can achieve reliably achieve concentrations less than 4 ppt (and is economically feasible) resulted in the proposed enforceable MCL. In Schrödinger terms, EPA took the PFAS out of the box and put it in a cage where we can keep a watchful eye on it. They also left the door open for more stringent drinking water levels in the future as measurement capabilities progress.
MCLs at the federal level must be adopted by the states, unless they choose to follow with even lower state-specific criteria. Aside from the analytical challenges, the proposed MCLs for PFOA and PFOS, and the enforceable hazard index for a combination of PFNA, PFBS, PFHxS, and GenX chemicals introduce other uncertainties, which could have implications for site investigations and remediation. For example, they may be considered in site-based decisions, such as applicable or relevant and appropriate requirements in Superfund cleanups, potentially resulting in more extensive investigations and remediation leading to increases in the time frames (and costs) required to achieve cleanup criteria. To reiterate from earlier, 4 ppt is a low concentration, which is even below some literature-reported anthropogenic background levels for these compounds. As such, the importance of conducting background studies to define site boundaries and conduct source evaluations to a public water supply may become even more important.
