Slicing and Dicing the UCMR3 Data
Oral Presentation
Prepared by A. Eaton
Eurofins Eaton Analytical, 750 Royal Oaks Drive, Monrovia, CA, 91016, United States
Contact Information: [email protected]; 626-386-1125
ABSTRACT
Monitoring began under EPA's Unregulated Contaminant Monitoring Rule in 2013 and the second public release in January, representing nearly 1500 water systems (PWS) and up to 12,000 samples. Additional data will be released quarterly. Our lab is doing UCMR3 monitoring for upwards of 400 PWS across the country, providing more current data.
Most metals, as predicted by us in 2012, are frequent in PWS, ranging from ~40% for molybdenum to >99% for strontium. Detection at levels exceeding available Health Reference Levels (HRLs) is minimal, although some systems have very high levels. Of greater interest is confirmation that groundwater sources are much higher than surface water systems, as was observed for some metals in old WaterRF studies. There is minimal evidence for significant increases in the distribution system, largely because the levels that might be added are small compared to point of entry (POE) levels.
Chromium is very prevalent, and most POE samples are predominantly in the hexavalent form, leaving little potential for increases in the distribution system due to further oxidation. If POE values are relatively low there is some distribution system increase but the amount is minimal compared to reference levels.
Chlorate, as published by Stanford et al, occurs in both systems disinfected with bulk hypochlorite and also on-site generation, although values from OSG systems in general are lower than those from bulk hypochlorite systems. This will in many ways impact compliance with any MCL, because controlling chlorate formation from bulk hypochlorite is easier than with OSG.
For organics, not only is 1,4-dioxane, a known anthropogenic contaminant, occurring much more frequently than originally expected (19% of PWS), but more than 5% of systems exceed the HRL. It occurs in both ground water and surface water. Contrary to expectations of 1,4-dioxane being associated with former 1,1,1-TCA impacted sites, it may be largely occurring as a result of other uses such as personal care products and industries such as wetting of textiles.
1,2,3-TCP, a potential driver of a regulation for carcinogenic VOCs, is not a widespread contaminant, except in CA and HI. Otherwise occurrence is very localized. The UCMR3 VOC of most concern from a toxicological perspective, 1,3-butadiene, has so far not been detected at all in UCMR3, consistent with prior predictions.
Perfluorinated compounds (PFCs) are generally found only in localized areas, again as predicted. The UCMR3 patterns are somewhat different from other studies (e.g. Oregon State Univ and USGS) with much lower reporting limits (ng/L vs 10-100 ng/L for UCMR3), where lower MRLs demonstrate much broader occurrence. For the PFCs having a good assessment of HRLs is critical for regulatory decisions.
Hormones occur very infrequently, as predicted, and the most commonly occurring hormones are not on the contaminant candidate list (CCL) making interpretation of the results from a health perspective challenging. So far all of the hormone hits have been in surface water systems, as would be expected for a contaminant that is mainly a wastewater effluent derived compound.
Both the frequency of occurrence of the UCMR3 contaminants and the patterns should be used for determining how to approach UCMR4 monitoring.
Oral Presentation
Prepared by A. Eaton
Eurofins Eaton Analytical, 750 Royal Oaks Drive, Monrovia, CA, 91016, United States
Contact Information: [email protected]; 626-386-1125
ABSTRACT
Monitoring began under EPA's Unregulated Contaminant Monitoring Rule in 2013 and the second public release in January, representing nearly 1500 water systems (PWS) and up to 12,000 samples. Additional data will be released quarterly. Our lab is doing UCMR3 monitoring for upwards of 400 PWS across the country, providing more current data.
Most metals, as predicted by us in 2012, are frequent in PWS, ranging from ~40% for molybdenum to >99% for strontium. Detection at levels exceeding available Health Reference Levels (HRLs) is minimal, although some systems have very high levels. Of greater interest is confirmation that groundwater sources are much higher than surface water systems, as was observed for some metals in old WaterRF studies. There is minimal evidence for significant increases in the distribution system, largely because the levels that might be added are small compared to point of entry (POE) levels.
Chromium is very prevalent, and most POE samples are predominantly in the hexavalent form, leaving little potential for increases in the distribution system due to further oxidation. If POE values are relatively low there is some distribution system increase but the amount is minimal compared to reference levels.
Chlorate, as published by Stanford et al, occurs in both systems disinfected with bulk hypochlorite and also on-site generation, although values from OSG systems in general are lower than those from bulk hypochlorite systems. This will in many ways impact compliance with any MCL, because controlling chlorate formation from bulk hypochlorite is easier than with OSG.
For organics, not only is 1,4-dioxane, a known anthropogenic contaminant, occurring much more frequently than originally expected (19% of PWS), but more than 5% of systems exceed the HRL. It occurs in both ground water and surface water. Contrary to expectations of 1,4-dioxane being associated with former 1,1,1-TCA impacted sites, it may be largely occurring as a result of other uses such as personal care products and industries such as wetting of textiles.
1,2,3-TCP, a potential driver of a regulation for carcinogenic VOCs, is not a widespread contaminant, except in CA and HI. Otherwise occurrence is very localized. The UCMR3 VOC of most concern from a toxicological perspective, 1,3-butadiene, has so far not been detected at all in UCMR3, consistent with prior predictions.
Perfluorinated compounds (PFCs) are generally found only in localized areas, again as predicted. The UCMR3 patterns are somewhat different from other studies (e.g. Oregon State Univ and USGS) with much lower reporting limits (ng/L vs 10-100 ng/L for UCMR3), where lower MRLs demonstrate much broader occurrence. For the PFCs having a good assessment of HRLs is critical for regulatory decisions.
Hormones occur very infrequently, as predicted, and the most commonly occurring hormones are not on the contaminant candidate list (CCL) making interpretation of the results from a health perspective challenging. So far all of the hormone hits have been in surface water systems, as would be expected for a contaminant that is mainly a wastewater effluent derived compound.
Both the frequency of occurrence of the UCMR3 contaminants and the patterns should be used for determining how to approach UCMR4 monitoring.
