Implications for Columbia River Basin


This study was designed as a reconnaissance to inform researchers, scientists, plant operators, policy-makers, and regulators about what compounds are being delivered to the Columbia River through two pathways, WWTP effluent and stormwater runoff. The target analytes included a broad array of compounds, including some of Oregon’s SB 737 priority pollutant list as well as emerging contaminants such as endocrine disruptors, personal care products, flame retardants, and other contaminants for which there is little available data for the Columbia River Basin. These data can provide a useful framework for directing future work to identify and reduce contaminant concentrations in the Columbia River Basin. 


Four Case Studies


Two pharmaceuticals—one over-the-counter and one prescription—and two AOCs—one musk and one detergent metabolite, both endocrine disruptors—were selected for further study. All four of these compounds were consistently detected in WWTP effluent (fig. 6) and in the Columbia River in previous studies. Diphenhydramine is an over-the-counter antihistamine used to treat the symptoms of hay fever, allergies, and the common cold; to prevent and treat motion sickness; to treat insomnia; and to control abnormal muscle movements in patients with early Parkinson’s syndrome (PubMed Health, 2010). Diphenhydramine was detected in samples collected at all nine WWTPs (table 10) with a median concentration of 0.062 µg/L. During a study in the lower Columbia River estuary, a trace-level amount of diphenhydramine was detected in filtered water from the Columbia River at Point Adams near the mouth of the river in April 2005 (Lower Columbia River Estuary Partnership, 2007). At Point Adams, a large amount of water is available for dilution, but diphenhydramine was still detected. Little is known about the effects of diphenhydramine on aquatic biota, but earthworms living in soils treated with biosolids accumulated diphenhydramine (Kinney and others, 2008). 


Trimethoprim is an antibiotic prescribed for urinary tract infections and also can be used to treat pneumonia and “traveler’s diarrhea” (PubMed Health, 2008). When combined with sulfamethoxazole, trimethoprim is often used to treat ear infections and chronic bronchitis. Besides its use for humans, it also is registered for use in dogs, horses, cattle, and swine. The median concentration of trimethoprim measured in WWTP effluents in this study was 0.089 µg/L, and it was detected in eight of the nine WWTPs sampled (table 10). In August 2004, trimethoprim was detected at low levels in the Columbia River at Warrendale (just downstream of Bonneville Dam), the Willamette River at Portland, and the Columbia River at Beaver Army Terminal near Longview. The direct environmental effects of trimethoprim are not known, but the presence of antibiotics in the aquatic environment could lead to microbial resistance (Kümmerer, 2004). 


Galaxolide^®, the short name for hexahydrohexamethylcyclopentabenzopyran (HHCB), is a synthetic fragrance used in cosmetics, cleaning agents, detergents, air fresheners, and perfumes (International Flavors and Fragrances, Inc., 2007). Galaxolide has been shown to be anti-estrogenic (Schreurs and others, 2005) and bioaccumulate in the food web (Hu and others, 2011). Galaxolide was detected in samples collected at all nine WWTPs (table 9) with a median concentration of 1.2 µg/L. During the lower Columbia River estuary study, trace levels of galaxolide were detected in the Columbia and Willamette Rivers during both high- and low-flow sampling events (Lower Columbia River Estuary Partnership, 2007).


Nonylphenol compounds are a group of nonionic detergent metabolites that are present in cleaners and sealers, and are produced in large quantities in the United States (table A2). They are known endocrine disruptors and are toxic to aquatic life, causing reproductive effects in aquatic organisms (U.S. Environmental Protection Agency, 2006). Nonylphenol compounds are resistant to natural degradation in water and their presence in WWTP effluent is often as a breakdown product from surfactants and detergents. For this discussion, the results for the five nonylphenol variants (para‑nonylphenol, NP1EO, NP2EO, OP1EO, OP2EO) analyzed are summed to provide a total measure of nonylphenol compounds. These compounds were detected at 8 of the 9 WWTPs (table 9) with a median concentration in Portland of 22 µg/L and a median for the other 7 plants of 3.6 µg/L. From a study of emerging contaminants in the surficial bed sediment of the lower Columbia River and its tributaries in 2005–06, nonylphenol compounds were detected in small tributaries (Fanno Creek), larger tributaries (the Tualatin River and Columbia Slough), and the Columbia River (upstream of the Willamette River and near the Cowlitz River) (Nilsen and others, 2007). Therefore, nonylphenol compounds may have been discharged from WWTPs consistently enough and at high enough concentrations to be measurable in the receiving-water environments. 


The freshwater aquatic-life criteria for nonylphenols are 28 µg/L for the acute (1-hour average) criterion and 6.6 µg/L for the chronic criterion (4-day average) (U.S. Environmental Protection Agency, 2006). Therefore, the concentrations measured at the Portland WWTP exceed the chronic criterion, but they do not represent a 4-day average. The European Union has banned nonylphenol and nonylphenol ethoxylates because they have been determined to be a hazard to human and environmental safety (Sierra Club, 2005).


Loadings to the River


For these four example compounds, instantaneous loadings from each WWTP were calculated and then extrapolated to represent a daily load to the Columbia River (table 25). These loads were determined by multiplying the concentration of a given contaminant by the daily discharge for the WWTP (table 12) and a conversion factor. This calculation provides a measure of the instantaneous load of the contaminant entering the river at the point of discharge if it is assumed that the measured concentration is a good representation of the concentration throughout the day for the WWTP effluent. The data necessary to quantify the variability for each contaminant throughout the day at a given WWTP were not collected for this study, but for exploratory purposes, these instantaneous loads can be useful to indicate potentially important sources of contaminants to the Columbia River. The instantaneous loads for the Portland WWTP were consistently higher than for any other wastewater-treatment plant because the discharge from the Portland WWTP is higher than for any other plant. The discharge from Portland is five times larger than the next largest WWTP in Vancouver. Additionally, the concentrations measured from the Portland WWTP also tended to be higher, particularly for the nonylphenols, indicating that this maybe an area to focus future reduction efforts. 


The calculations were taken a step further by dividing the instantaneous load by the 7Q10 streamflow of the Columbia River at that point of discharge (table 2) and multiplying by a conversion factor to determine the resulting concentration in the Columbia River that could be attributed to this incoming load (table 25). These calculations illustrate that the Columbia River carries a large amount of water and is able to “absorb” a variety of inputs because of its ability to dilute.


The calculated concentrations were all small, mostly less than 1 ng/L, except near the Portland WWTP. These concentrations, besides Portland, would not be measurable in the Columbia River main stem using standard sampling techniques. Although these calculated concentrations are small in the context of the whole river, the localized effect of these continuous inputs to the mixing zone is understudied and underrepresented. The aquatic biota using these areas may be exposed to higher concentrations than in other areas. 


Comparison to SB 737 Plan Initiation Levels


As part of Oregon’s SB 737 process to identify persistent pollutants and reduce concentrations entering Oregon’s waterways (Oregon Department of Environmental Quality, 2010a), ODEQ was tasked with developing PILs for each of these pollutants. These PILs are used to determine if a city will be required to develop a reduction plan for each persistent pollutant. ODEQ selected these PILs from several existing values, including drinking-water maximum-contaminant levels for those that are established, peer-reviewed national and international government documents, and scientific literature that reflect current scientific information. 


The reconnaissance data from this study were compared to the SB 737 list and PILs to provide some preliminary information. Of the 118 persistent pollutants on the SB 737 list, 63 were not analyzed as a part of this study, and 13 were analyzed only in stormwater-runoff samples (table 26). Senate Bill 737 applies only to WWTP effluent and not to stormwater runoff. Of the remaining 42 pollutants analyzed in wastewater, 27 were detected. Only 4 of the 27 were measured at least once at a level greater than the assigned PIL (table 27). One of these, fluoranthene, was detected only in the sample from the Wenatchee WWTP (E 0.11 µg/L), which is in Washington and, therefore, not subject to the requirements of Oregon SB 737. Another PAH, anthracene, was detected in a sample collected at the St. Helens WWTP at greater than the PIL of 0.01 µg/L. It was measured at a level less than the LT-MDL (reported as “Present”), but the quantified result slightly exceeded the PIL of 0.01 µg/L. 


The other two compounds from the SB 737 list that were detected in this study at concentrations that were greater than the PILs are both sterols that are ubiquitous in wastewater. Cholesterol is a structural component of mammalian cell walls and is transported in the blood plasma of all animals. Its effects in the aquatic environment are not understood. Coprostanol is a metabolite of cholesterol excreted in fecal matter from mammals and is, therefore, detected in urban effluents. It can be accumulated by organisms living in municipal effluent outfalls and has been determined to have estrogenic effects in freshwater mussels (Gagné and others, 2001). According to ODEQ, “Research suggests that coprostanol has an affinity to estradiol-binding sites and … large amounts may pose a threat to aquatic invertebrates” (Oregon Department of Environmental Quality, 2010a). 


A large number of the AOCs and pharmaceuticals analyzed in WWTP-effluent samples in this study were detected, yet only 27 of the 42 compounds on the SB 737 persistent pollutants list that were analyzed in this study were detected in these WWTP-effluent samples. The reason for this dichotomy is that many of the contaminants on the persistent pollutant list are PAHs, metals, and currently used pesticides—all compounds that are likely to be detected in stormwater but not necessarily wastewater. To illustrate this point, 45 of the 50 compounds on the persistent pollutant list that were analyzed in stormwater-runoff samples in this study were detected; however, SB 737 does not apply to stormwater. One category that is absent from the persistent pollutant list is pharmaceuticals, the contaminant class that many people associate with WWTP effluent. Sufficient documentation about the persistence or bioaccumulative ability of many pharmaceuticals is not available to allow consideration for the persistent pollutant list. 


First posted April 25, 2012