Impact of Nitrogen Removal in Wastewater Treatment on NDMA Formation at Downstream Drinking-Water Treatment Plants
Publication: Journal of Environmental Engineering
Volume 147, Issue 12
Abstract
De facto reuse (DFR) occurs when surface waters receiving wastewater effluent are used as source water by a downstream drinking-water treatment plant (DWTP). Wastewater effluent may contain high concentrations of N-nitrosodimethylamine (NDMA) precursors. NDMA is a toxic disinfection byproduct (DBP). Nitrification at wastewater treatment plants (WWTPs) reduces NDMA precursors in this effluent. The present work examines NDMA concentrations in the distribution systems of 31 DWTPs subject to high percentages of DFR. WWTPs contributing to this DFR were categorized by extent of nitrogen treatment: no intentional nitrogen treatment, nitrification, or nitrification followed by denitrification. The fraction of upstream effluent discharges receiving nitrogen treatment was calculated, and the relationship between this fraction and observed NDMA concentrations was assessed. Results indicate that high-DFR DWTPs with large fractions of wastewater undergoing nitrification upstream had NDMA concentrations comparable to a representative national sample of plants. High-DFR plants with little upstream nitrification had significantly higher NDMA detection rates and concentrations, with dependence on the distance between the upstream WWTPs and the downstream DWTP. High DFR low nitrification, high DFR high nitrification, and a national data set of plants had detect rates of 79%, 54%, and 50% and median detects of , , and , respectively, when considering a 150-km cutoff. This suggests that nitrification at WWTPs may have an added benefit of reducing NDMA formation at chloraminating DWTPs located downstream in high-DFR areas.
Introduction
De facto reuse (DFR) is the unplanned reuse of discharged wastewater effluent that occurs when the receiving body of water is used as a drinking-water source downstream. Rice and Westerhoff (2015) completed a nationwide study of 1,210 drinking-water treatment plants (DWTPs), each serving more than 10,000 people, and concluded that roughly 50% of the plants were affected to some degree by DFR. Further, Rice et al. (2013) observed that, in 2008, higher volumes of wastewater affected the 25 most DFR-impacted DWTPs identified by the USEPA nearly 30 years earlier (Swayne et al. 1980). The number of drinking-water plants affected by DFR and the volume of wastewater affecting them are likely to increase further in the future with increasing urbanization and population growth in the United States. Thus, chemicals that persist through wastewater treatment (e.g., nutrients or pharmaceuticals) may reach more drinking-water intakes at higher concentrations.
A primary concern surrounding DFR has been the potential for wastewater effluent to affect disinfection byproduct (DBP) formation at downstream DWTPs (Krasner et al. 2013, 2009a, 2008; Mitch et al. 2003) or to directly contribute DBPs from the disinfection of treated wastewater (Schreiber and Mitch 2006). DBPs are formed when organic matter reacts with a disinfectant being used at a treatment plant. These by-products are toxic (Plewa et al. 2004, 2002; Richardson et al. 2007), and consumption of treated water containing DBPs has been associated with cancer (Villanueva et al. 2006, 2004) and negative reproductive outcomes (Cedergren et al. 2002; Magnus et al. 1999) in epidemiological studies. Two classes of organic DBPs [trihalomethanes (THMs) and haloacetic acids (HAAs)] are currently regulated (USEPA 2006) as surrogates for unmeasured and unregulated DBPs that are potentially the source of observed health risks. The suite of DBPs that are formed during disinfection are highly dependent upon source water characteristics (Kristiana et al. 2017; Roccaro et al. 2011). The presence of nitrogen can result in the formation of nitrogenated disinfection byproducts (N-DBPs) (Kristiana et al. 2017; Singer 1994; Sgroi et al. 2018), which are more toxic than carbonaceous DBPs (Plewa et al. 2008; Richardson et al. 2007).
Municipal wastewater discharges are a significant source of organic and inorganic nitrogen, which raises concerns for N-DBPs in chlorinated wastewater effluent as well as the potential for increased N-DBP formation at downstream DWTPs. A key group of N-DBPs associated with wastewater and disinfected wastewater effluent are the highly toxic N-nitrosamines, including N-nitrosodimethylamine (NDMA) (Gerrity et al. 2015; Lee et al. 2015; Russell et al. 2012; Schreiber and Mitch 2006; Zeng et al. 2016). Domestic wastewater serves as a significant source of precursors to NDMA formation (Mitch et al. 2003); common secondary-treated wastewater effluent contains sufficient NDMA precursors to form concentrations of over NDMA under formation potential conditions (Hanigan et al. 2012; Mitch and Sedlak 2004). In addition, NDMA forms during wastewater treatment disinfection and thus is frequently observed in wastewater treatment plant (WWTP) effluent (Chuang and Mitch 2017; Dai et al. 2015; Krasner et al. 2009b; Lee et al. 2015; Mitch and Sedlak 2004). In some watersheds, wastewater effluent is the dominant source of NDMA and its precursors compared with other sources: agricultural runoff, stormwater runoff, and algal blooms (Sgroi et al. 2018; Zeng et al. 2016).
NDMA has become a significant DBP of concern in drinking water as well, particularly at treatment plants using chloramine as a disinfectant (Krasner et al. 2013; Mitch et al. 2003; Richardson et al. 2007; Russell et al. 2012). NDMA is not currently regulated; however, the California Department of Health Services (now the California Department of Public Health) has set a notification level of (ppt) in drinking water (California Department of Health Services 2002). The USEPA has included NDMA on the Contaminant Candidate List (CCL) 3 and 4 (USEPA 2009, 2016); these lists include chemicals that are known or anticipated to occur in public water systems and may require future regulation under the Safe Drinking Water Act (SDWA). NDMA is considered a probable human carcinogen and has an estimated risk threshold of for consumption of drinking water according to the EPA Integrated Risk Information System (IRIS) (Peto et al. 1984; USEPA 1987). In the more recent Announcement of Preliminary Regulatory Determinations for Contaminants on the Third Drinking Water Contaminant Candidate List, EPA described NDMA as likely to be carcinogenic to humans by a mutagenic mode of action and listed a drinking-water health reference level of at the risk level ( at the risk level) (USEPA 2014). Occurrence data indicate NDMA is the most frequently detected nitrosamine in drinking water, with 27% of plants included in the Second Unregulated Contaminant Monitoring Rule (UCMR2) reporting detection above the minimum reporting limit (MRL, ) at least once from 2008 to 2010. This represented 8% of chlorinating plants and 78% of chloraminating plants included in the UCMR2 survey that were assessed in this study; see section “Methodology” (USEPA 2012b).
Because wastewater effluent is known to be a large source of NDMA precursors and is also frequently discharged upstream of drinking-water intakes, the role of wastewater treatment processes in NDMA precursor removal has been explored. Nitrification has been observed to reduce NDMA formation potential (NDMAFP) (Gerrity et al. 2015; Krasner et al. 2009a, b, 2008; Mitch and Sedlak 2004) through reductions in hydrophilic natural organic matter (NOM) and dissolved organic nitrogen (DON). Krasner et al. (2009a) observed a halving of median NDMAFP after “good nitrification” ( as N) relative to no nitrification, whereas denitrification was observed to increase NDMAFP relative to nitrification alone but decrease it relative to conventional treatment without nitrification. In-river decay of NDMA precursors has also been observed (Schreiber and Mitch 2006), likely due to microbial nitrogen transformations: decay rates in the Quinnipiac River were for the spring and in the summer, and the spring decay rate corresponded to a 3.5-day half-life. Chen et al. (2009) observed a reduction in NDMAFP of 50%–60% across the length of the effluent-dominated Santa Cruz River, with NDMAFP declining more rapidly during the summer than winter. It is thus expected that in many cases, precursors would still be present in significant concentrations when water reaches downstream DWTPs.
Field observations linking wastewater treatment-based effects on finished water NDMA at downstream drinking-water treatment plants have not been reported previously. The work presented here examines NDMA formation at high-DFR DWTPs, with a focus on nitrogen treatment levels at upstream WWTPs. The goal was to determine whether treatment decisions at WWTPs are associated with NDMA formation in downstream DWTPs and whether this effect was mediated by the distance between the WWTP and DWTP. To achieve this goal, DWTPs affected by DFR were selected and the contributing upstream WWTPs were identified to assess treatment levels. Then, a comparative statistical analysis was performed to evaluate trends for NDMA occurrence in relation to the presence of DFR, level of WWTP treatment, and effluent travel distance.
Although the focus of this work is on the relationship between DFR and wastewater treatment plant nitrogen treatment and NDMA formation, other potential sources of NDMA precursors are likely present in source waters, including those from agricultural and urban runoff (Bei et al. 2016; Chen and Young 2009; Leavey-Roback et al. 2016). Runoff effects are complicated by variable flow within river systems as well as effects of land use and agricultural practices. These potential NDMA precursor effects are not considered in the present study. Further, in-plant treatment choices are well known to affect NDMA formation, particularly the use of polymer coagulants like polyDADMAC. Krasner et al. (2020) observed increasing NDMA formation with increasing use of polyDADMAC. This work assumes polymer usage has an equally distributed impact across all subgroupings of systems.
Methodology
NDMA Occurrence Data
DWTP NDMA data were collected (in addition to other nitrosamines) between 2008 and 2010 as part of the UCMR2 screening survey that included all public water systems that served more than 100,000 people (398 systems), 320 public water systems serving 10,001 to 100,000 people, and 480 systems serving fewer than 10,000 people (USEPA 2012b). In the present analysis, only large (L, 10,001–50,000), very large (VL, 50,001–100,000), and extra large () systems as defined in the UCMR2 were considered (a total of 718 systems).
The UCMR2 data request specified that samples were to be collected at 3-month intervals for a 12-month period between January 2008 and December 2010. Thus, most drinking-water systems had at least four collection dates per sampling location (with multiple locations in the distribution system). NDMA samples analyzed in the present work were limited to those taken at the maximum residence time in the distribution system for plants listing surface water as the water source and using either chlorine or chloramine as the secondary disinfectant. Samples were collected by water utilities following a standard protocol and analyzed using method EPA 521, with a minimum reporting limit of NDMA (Munch and Bassett 2004; USEPA 2007).
For the present work, data from the UCMR2 (USEPA 2012b) were imported into Excel. Extracted data included every sample taken during the UCMR2; prior to analysis, data were filtered by five criteria: the “size” field (only “L,” “VL,” and “XL” were used), “contaminant” (only “NDMA”), “FacilityWaterType” (only surface water, “SW”), “SamplePointType” (only maximum residence, “MR”), and “DisinfectantType” (only those using chlorine “CL” or chloramine “CA”). The “AnalyticalResultValue” field provided NDMA measurement concentrations. This field was left empty when the measure was below the MRL, which was . For the present statistical analysis, samples below the reporting limit were replaced with (USEPA 2012a). Thus, the lowest nonzero value for individual measurements in the analysis presented here was .
High DFR DWTP Data Set
Previous work by Rice et al. (2015) identified 145 DWTPs with de facto reuse wastewater flow contribution by volume (under mean streamflow conditions) at a surface water intake that served a population of more than 10,000 people and participated in the UCMR2 data collection effort (USEPA 2007). The DFR percentages retrieved from Rice et al. (2015) represent the fraction of streamflow at the drinking-water intake that was sourced from upstream wastewater discharges under mean annual streamflow conditions as estimated by the National Hydrography Dataset Plus (NHDPlus version 2). Additional information on the nature of the DFR estimate can be found in work by Rice and Westerhoff (2015) and Rice et al. (2013, 2015). In the present analysis, we classified DFR as “high” when the DFR was in the upper 50% of DFR-impacted intakes described in Rice and Westerhoff (2015) ().
Of the DFR plants identified and located by Rice et al. (2015) with UCMR2 samples that met requirements (maximum residence time, surface water, secondary disinfectant type), those that were identified as having a single surface water intake were included in the present analysis. Those with multiple intakes were excluded because they could confound the assessment of relative contribution from upstream wastewater discharges. If a plant had a sample matching all criteria but was missing disinfectant information for certain samples, disinfectant information was deduced through disinfectant listed for other samples from the plant and confirmed through consumer confidence reports, when possible. If the disinfectant remained unclear, the sample was excluded. This additional evaluation was not done with the full UCMR2 data set because of the large number of treatment plants. Rather, samples without disinfectant information were excluded.
Plants with intake locations that could not be confirmed or potentially significant contributions from WWTPs in Mexico or Canada (for which treatment data were not available) were not considered. The resulting high-DFR subset was a group of 31 DWTPs; details are provided in Table S1 , and plant locations are identified in Fig. S1 . Previously Rice et al. (2015) identified 835 drinking-water plants serving populations greater than 10,000 with surface water intakes located downstream of wastewater treatment plants, and these were geographically distributed throughout the country. The data limitations in the present study resulted in a subset of plants located predominately in the east and Midwest, and thus, the study conclusions are not generalizable to the entire US.
High-NDMA Plants
DWTPs in the high-DFR set with high NDMA values (averages above the California notification level of , shown in bold in Table S1 ) were examined separately. USGS gauges (USGS 2018) were used to examine the source water stream flow in the time leading up to the sample collection date (provided in the UCMR2). Because NDMA samples were taken at the point of maximum residence time in the distribution system, and residence times in US drinking waters can range from days to weeks, it was assumed that there would be a lag time between source water extraction and sample date (USEPA 2002). The flow data in the time leading up to the sampling event were qualitatively compared with NDMA concentrations to assess whether high NDMA measures corresponded with source water conditions where flow was below the mean annual average; low flows could cause higher concentrations of NDMA precursors to be present in the source water through larger fractions of DFR contribution, as has been observed in Prescott et al. (2017).
Clean Water Needs Survey Data
WWTP flow and treatment data were taken from the 2008 Clean Watersheds Needs Survey (CWNS) (USEPA 2008). The CWNS is conducted every 4 years and is designed to assess the financial needs of the nation’s wastewater treatment and collection systems. The 2008 CWNS was selected because it was the survey taken nearest in time to the NDMA collection in the UCMR2. In this survey, wastewater treatment plant operational data are collected, including design capacity flow, actual flow, and treatment processes (ammonia and nitrogen removal). For the present analysis, actual (“existing”) plant flows reported in the CWNS were used to calculate the flow-based fraction of upstream wastewater discharges that had been treated with some degree of advanced nitrogen treatment. Plants that indicated that they conducted ammonia removal but not nitrogen removal are referred to here as nitrifying plants. Plants that included nitrogen removal as well as ammonia removal are described here as denitrifying. As noted previously, Krasner et al. (2009a) observed the degree of nitrification (e.g., partial versus full) affected NDMAFP; however, CWNS data are not sufficient to determine the degree to which nitrification or denitrification processes were operating as intended. Thus, degree of nitrification was not a variable in the present study.
Although wastewater effluent discharges are a primary source of NDMA precursors to surface water (Sgroi et al. 2018; Zeng et al. 2016), they are not the only source of nitrogen present in the water. This work does not account for additional discharges, such as from nonpoint sources. These sources may influence NDMA formation; however, there are inadequate data to quantify this influence.
Identifying Upstream Treatment Type for High DFR DWTPs
In order to compare NDMA formation at DWTPs with upstream wastewater treatment, contributing upstream WWTPs were identified using the map component of the De Facto Reuse Incidence in our Nations Consumable Supply (DRINCS) ArcGIS model, as developed by Rice and Westerhoff (2015) and Rice et al. (2013, 2015). This model spatially links wastewater treatment plant discharge points and drinking-water intake locations with USGS hydrologic flowpaths. Identification numbers for both DWTPs [public water system identification number (PWSID)] and WWTPs (CWNS ID) were presented in the ArcGIS map. Using the map, the USGS flowpaths were used to trace upstream of drinking-water intakes (identified by PWSID) and identify all discharging wastewater treatment plants by their CWNS identification numbers. Flow and treatment data for these WWTPs were available in the CWNS. Boundaries were limited to the United States.
Wastewater treatment plants with CWNS identification numbers that were not found in the 2008 CWNS were not included in the analysis; 337 wastewater plants were excluded from analysis, representing 9.7% of the all unique upstream plants observed. To examine whether the missing plants were likely to amount to significant unaccounted flows, 2004 CWNS flows were used to estimate the flow volumes for the plants missing in the 2008 CWNS. In all but one case, the total flow from excluded WWTPs was less than 5% of the total wastewater flow that was included in this analysis. The only instance for which this was exceeded was at a DWTP in Georgia (PWSID GA1210001). For this plant, upstream WWTPs where 2008 CWNS data were lacking had total estimated flows (from 2004 data) equivalent to nearly 10% of the flows from WWTPs that were included. GA1210001 was a chlorinating plant that did not yield any NDMA detects during the UCMR2, and thus these missing data did not affect the overall results of the analysis.
Wastewater Treatment Metrics
The fractions of wastewater effluent subjected to varying levels of nitrogen treatment were calculated as three metrics: nitrified, denitrified, and nitrogen treatment [Eqs. (1)–(3)]
(1)
(2)
(3)
Here, the terms Nitrified and Denitrified are used exclusively. The nitrified fraction does not include flows that are nitrified prior to being denitrified. N Treatment includes flows from both nitrification-only plants and nitrification-denitrification plants.
Distance Dependence
Effluent organic matter and DBP precursors are known to degrade in surface waters (Chen et al. 2009; Schreiber and Mitch 2006). Organic nitrogen released in wastewater effluent may be mineralized to ammonia or ammonium and volatilize from surface water or may be oxidized to nitrite or nitrate by nitrifying organisms in the receiving water. Oxidized species may then be denitrified to or other gaseous species, such as . Schreiber and Mitch (2006) noted that NDMA precursors have half-lives of several days in surface waters. Thus, downstream impacts of WWTP effluent discharge may lessen with travel time or distance as nitrogen is transformed in the receiving water. To assess the influence of this effect, the present analysis comparing DWTP NDMA concentrations with upstream WWTP nitrogen treatment was replicated with distance-based cutoffs (as a proxy for travel time) for WWTPs included. For a given distance, both the wastewater flow treated with nitrification () and nitrification or denitrification () were recalculated with only flows from WWTPs within the selected radius upstream of the DWTP intake location. In addition to the initial analysis with no cutoff, radii of 50, 100, and 150 km were used; these were not intended to mimic flow distances. WWTPs located beyond the radius or that had flow paths outside the radius were removed from the list of contributing plants to the drinking-water treatment plant for each radius-based analysis. The counts of contributing WWTPs included in each analysis are shown in Table S2 .
Statistical Analysis and Data Set Representation
Left censoring of the data was significant because 80% of all extracted UCMR2 NDMA concentrations were below the MRL. Thus, a quantile-based comparison was selected to avoid imputation of values below the MRL biasing the analysis. The Mann–Whitney test (Mann and Whitney 1947) for differences in the median was used to test for statistically significant differences between sets of drinking-water treatment plants used when comparing high-DFR plants and the full UCMR2. An alternative hypothesis that the two medians were unequal was used. For exceedance and detection frequencies, the two-sample test of proportions was used with an alternative hypothesis that the two proportions were unequal. For comparisons of distributions, the Kolmogorov–Smirnov two-sample test for statistically significant differences in empirical distributions was used (Smirnov 1948). This provided insight on the full spread of the data and was not limited to the central tendency, which would be expected to exhibit bias due to the high proportion of censored data. An alpha of 0.05 was used to determine statistical significance.
The chi-square test was used to compare the categorical drinking-water plant size distribution of the high-DFR set with the larger UCMR2 set (Pearson 1900). This was done to examine whether the data sets had statistically significant differences in plant size composition. An alpha of 0.05 was used.
Correlation analysis was also used to examine the relationships between the nitrogen treatment metrics and individual sample NDMA measures. In this work, the Pearson correlation was selected (Pearson 1895). This was used to compare the linear relationships between the extent of treatment and NDMA formation.
To consider the representativeness of the DFR subset used for this work, which was selected to have high DFR and also to be limited to single intake plants, we considered whether the subset was similar to the UCMR2 set in terms of size of plants, disinfection type, and location. For size comparison, plants were grouped into L (10,001–50,000), VL (50,001–100,000), or XL () (USEPA 2012b).
In the UCMR2 set, the majority (218 of 318; 68.6%) fell in the XL category, whereas 58 (18.2%) and 42 (13.2%) fell in the VL and L categories, respectively. In the high-DFR set, the majority (19 of 31 plants; 61.3%) of DWTPs fell in the XL category, whereas 7 (22.6%) and 5 (16.1%) fell in the VL and L categories, respectively. Thus, the two sets had approximately the same size composition; the general shape of the plant size distribution was the same in the subset and the full UCMR. A chi-square goodness-of-fit test indicated no significant difference in plant sizes in the high-DFR set compared to the UCMR2 plants ().
In terms of disinfectant type, across the UCMR2 set, 63% of the DWTPs reported using chlorine, whereas 39% of plants reported using chloramine. Some plants used both disinfectants. Among the high-DFR plants, 58% of the DWTPs used chlorine, whereas 48% of the plants used chloramine.
Results and Discussion
The link between high DFR and detection of NDMA in finished water at drinking-water plants has been suggested previously (Rice et al. 2015). NDMA is well known to be associated with chloramination for high-DFR and low-DFR conditions. The UCMR2 data show few NDMA detections in chlorinating plants. For the UCMR2, 25 of 1,082 (2.3%) measurements at a chlorinating plant yielded a detection; within the high-DFR subset, 4 of 70 (5.7%) measurements at chlorinating plants exceeded the MRL. For chloraminating plants, NDMA was detected in 330 of 661 samples (∼ 50%) in the UCMR2, whereas high-DFR plants using chloramination reported detections of NDMA in 37 of 56 samples (∼ 66%); this difference in detection frequency was significant (). Fig. 1 shows the empirical cumulative distribution function (CDF) plots for the UCMR2 sample set (solid) and the high-DFR plant sample subset (dashed) at chloraminating plants.
EE.1943-7870.0001927/asset/dfafc412-b58f-426d-97f9-403ab68d5f2c/assets/images/large/figure1.jpg)
The CDF of the high-DFR plants is right-shifted relative to the UCMR2, indicating generally higher NDMA measurements at the DFR plants. Further, the two empirical CDFs were statistically significantly different (). At the high-DFR plants, 18% of the samples exceeded the California notification level of NDMA (vertical dashed line in Fig. 1). In the UCMR2 set, only 7% of samples exceeded this threshold. The exceedance frequency in the high-DFR plants was significantly higher (). Fig. 2 is a boxplot of the reported NDMA values () for all plants in the UCMR2 and for plants in the high-DFR subset; again, only sites using chloramination are shown.
EE.1943-7870.0001927/asset/60af3fdf-4a95-47c3-bc7b-c58fd8318b21/assets/images/large/figure2.jpg)
Detectable NDMA measurements at the high-DFR plants were statistically significantly higher than in the full UCMR2 set. The median detection at high-DFR plants () was significantly higher () than the median detection in the UCMR2 set (). Further, the distributions in the two sets were significantly different (), with the high-DFR distribution being right-shifted (higher NDMA) relative to the UCMR2 distribution (Fig. S2 ). The high-DFR plants had higher NDMA detection frequency, higher NDMA concentrations, and more values that exceeded the California threshold; thus, DFR appears to be associated with higher NDMA formation in chloraminating DWTPs.
High NDMA Values
Three chloraminating plants with UCMR sample average NDMA values above were considered separately; all had intakes in close proximity on the Mississippi River. Nine of the 12 samples collected by these plants were collected during periods in which flow was lower than the annual mean for the period of record. However, there was no apparent pattern in NDMA concentrations relative to the flow rate of the Mississippi in the time leading up to sampling. Samples taken during lower flow periods, where higher DFR would be likely, did not produce higher NDMA measurements. The large size of the Mississippi River lessens the effect of flow fluctuations on the DFR fraction compared to small streams where drought conditions can result in DFR constituting the majority of the flow (Rice and Westerhoff 2015). Further, the Mississippi River watershed contains widespread intensive agricultural activities, and these activities may contribute significant NDMA precursors to source waters for these three treatment plants.
Impact of WWTP Treatment on DWTP NDMA Formation
The present work is designed to evaluate whether choices at the upstream wastewater treatment plants alter the potential for NDMA formation, evaluated by the concentration detected at the maximum residence time in the distribution system. Three treatment metrics, , , and N , at the upstream WWTPs were compared through correlation analysis with measured NDMA concentrations at the maximum residence time location in the downstream DWTP distribution system. The results across all analyses for drinking-water plants using chloramine are shown in Table 1.
| Spatial range of analysis | Statistical value | Treatment metric | ||
|---|---|---|---|---|
| Fraction nitrified | Fraction denitrified | Fraction nitrogen treatment | ||
| Unbracketed | Pearson correlation | 0.119 | 0.092 | |
| -value | 0.381 | 0.898 | 0.501 | |
| 50 km | Pearson correlation | |||
| -value | 0.17 | 0.971 | 0.213 | |
| 100 km | Pearson correlation | ** | ** | |
| -value | 0.023** | 0.473 | 0.033** | |
| 150 km | Pearson correlation | *** | 0.031 | * |
| -value | 0.009*** | 0.822 | ||
Note: *** Correlation is significant at the 0.01 level; ** correlation is significant at the 0.05 level; and * correlation is significant at the 0.10 level.
No correlations were observed between any extents of treatment and NDMA in the unconfined or 50-km analyses. Correlations were seen, however, when using radii of 100 and 150 km. At 100 km, both the nitrified fraction of wastewater and the fraction of nitrogen treatment were significantly weakly inversely correlated with NDMA concentrations ( values shown in Table 1), suggesting upstream nitrification reduced NDMA precursors entering the watershed and reaching the drinking-water intake. The weak correlation reported here is likely due to additional NDMA precursor sources not considered. Alternative sources of NDMA precursors include agricultural and urban runoff, use of polymer coagulants, and variable treatment efficacy in upstream treatment plants (e.g., degree of nitrification in nitrifying plants). These sources may obscure the relationship between wastewater treatment and downstream NDMA formation at drinking-water plants.
In the 150-km analysis, NDMA formation was not significantly () correlated with percent nitrogen treatment. However, percent nitrified treatment had an inverse correlation with NDMA significant at an alpha of 0.01. This is consistent with prior research suggesting nitrification reduces NDMA precursors in treated wastewater effluent (Krasner et al. 2009a). Although the distance assessment method is insufficient to account for wastewater effluent travel times, these results demonstrated the expected dependence of DFR impact on distance between wastewater effluent discharge and DWTP intakes. The analysis that included all WWTPs may have included wastewater that is significantly transformed in the environment prior to reaching the drinking-water intake, thus reducing NDMA precursors, whereas the analysis that included only the closest WWTPs (within 50 km) may have neglected wastewater that is reaching the intake after relatively little transformation. Fig. 3 is a scatterplot showing the observed NDMA concentrations relative to fraction of nitrification (left panel) and nitrogen treatment (right panel) for the 150-km cutoff. Plots of the results of the 50-km, 100-km, and unconfined analysis can be found in Figs. S3 –S5 , respectively.
EE.1943-7870.0001927/asset/9da5be7f-af5a-44e7-8212-5d99f3de60d2/assets/images/large/figure3.jpg)
All NDMA measures exceeding occurred at chloraminating drinking-water plants where upstream DFR was either not nitrified at all or had very low nitrification (fraction ) (left panel). This is also seen for chlorinating plants; four detects observed in the high-DFR chlorinating plants all occurred at an upstream nitrification fraction of less than 0.02 (data not shown). Considering both extents of nitrogen treatment, which include those nitrifying treatment as well as treatment including nitrification and denitrification, NDMA concentrations above are observed when a small fraction of nitrogen treatment occurs () (Fig. 3, right panel). This may suggest denitrification returns NDMA precursors to treated wastewater, as previously observed (Krasner et al. 2009a); however, the paucity of data in the low fraction of nitrification (between 0.001 and 0.2 in the left panel) precludes any comparative conclusion.
Influence of Nitrification on DWTP NDMA Formation
To further explore the relationship between upstream nitrification and DWTP NDMA formation, the median for upstream wastewater effluent was used to split the NDMA data into low and high nitrification groups (except for in the 50-km analysis). In the 50-km analysis, 57% of the data had 0% upstream nitrification. In this case, all 0% nitrification data were grouped as low nitrification. For the 150-km analysis, the split occurred such that the highest fraction of nitrification in the low nitrification group was 0.06%, whereas the lowest nitrification in the high nitrification was 12.5% of upstream WW effluent discharges. In the 50-km and 100-km analyses, all plants in the low-nitrification bin had 0% nitrification, whereas all plants in the high-nitrification bin had at least 10% nitrification. The empirical distributions for the grouped data in the 150-km analysis are shown in Fig. 4.
EE.1943-7870.0001927/asset/076a7f94-e98d-49d3-82b7-28df41668129/assets/images/large/figure4.jpg)
The CDF of samples taken in the high nitrification group closely follows that of the UCMR2, with no significant difference in distributions observed. However, the low nitrification group was significantly different (right-shifted, higher NDMA) compared to the UCMR2. This pattern was also seen for the 50-km and 100-km radii analyses but was not seen in the unconfined analysis. NDMA was more frequently detected in the low nitrification group than the high nitrification group for the 50-km (low nitrification: 69%, high nitrification: 62%), 100-km (low nitrification: 79%, high nitrification: 54%), and 150-km (low nitrification: 79%, high nitrification: 54%) analyses. The difference was significant for the 100-km and 150-km analyses () but not for the 50-km analysis ().
The median NDMA detection in the low nitrification group was significantly higher than in the high nitrification group for the 50, 100, and 150-km analyses (low nit , high nit , for each). In the 150-km analysis, about 36% of the samples for drinking-water plants where upstream nitrification was low exceeded the California Department of Public Health’s notification level of NDMA (31% and 36% for 50 and 100, respectively). This threshold was not exceeded for any drinking-water plant where upstream wastewater treatment had high nitrification.
Conclusion
The complete high-DFR plant set had a higher NDMA detection rate and NDMA concentrations reported than the overall UCMR. Dividing the high-DFR drinking-water plants into those with upstream wastewater plants practicing high and low degrees of nitrification shows that upstream nitrification led to high-DFR drinking-water plants that looked more similar to all drinking-water plants (in the UCMR2) in terms of NDMA formation. However, high-DFR drinking-water plants with low rates of nitrification at wastewater plants upstream had much higher detection rates (nearly 80%) and elevated NDMA measurements. Thus, when DFR rates were high, nitrification at the upstream wastewater plants appeared to reduce NDMA formation at downstream drinking-water plants affected by DFR. Other factors such as agricultural-associated NDMA precursors and treatment choices in drinking-water plants also affect NDMA formation; thus, any watershed-scale NDMA control program would need to consider these as well. The potential effect of wastewater treatment on downstream drinking-water NDMA formation may be an unaccounted benefit for the implementation of nitrogen oxidation at wastewater treatment plants. This potential benefit for installation of nitrification in wastewater treatment plants upstream of drinking-water intakes should be explored further.
Supplemental Materials
File (supplemental_materials_ee.1943-7870.0001927_cadwallader.pdf)
- Download
- 211.42 KB
Data Availability Statement
Some or all data, models, or code that support the findings of this study are available from the corresponding author upon reasonable request. Data available upon request include: paired wastewater treatment plant and drinking-water system identification numbers, including by distance cutoff; treated wastewater discharge volumes upstream of drinking-water treatment plants by treatment type and distance cutoff; and relevant data from the publicly available Clean Watersheds Needs Survey and the Third Unregulated Contaminant Monitoring Rule.
Acknowledgments
This work was supported by the National Science Foundation (NSF) Integrative Graduate Education and Research Traineeship in Nanotechnology-Environmental Effects and Policy (IGERT NEEP) fellowship program (DGE-0966227), the Heinz Endowments, Colcom Foundation, National Science Foundation Decision Center for a Desert City II: Urban Climate Adaptation (SES-0951366), and National Science Foundation Central Arizona-Phoenix Long-Term Ecological Research (BCS-102685). Support was also provided by the Center for Water Quality in Urban Environmental Systems at Carnegie Mellon University. Any opinions, findings, conclusions, or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the organizations providing financial support.
Disclaimer
This work is not a product of the United States Government or the United States Environmental Protection Agency. The authors/editors are not doing this work in any governmental capacity. The views expressed are their own and do not necessarily represent those of the United States or the USEPA.
References
Bei, E., X. Liao, X. Meng, S. Li, J. Wang, D. Sheng, M. Chao, Z. Chen, X. Zhang, and C. Chen. 2016. “Identification of nitrosamine precursors from urban drainage during storm events: A case study in southern China.” Chemosphere 160 (Oct): 323–331. https://doi.org/10.1016/j.chemosphere.2016.06.081.
California Department of Health Services. 2002. “Studies on the occurrence of NDMA in drinking water.” Accessed March 2018. https://www.waterboards.ca.gov/drinking_water/certlic/drinkingwater/documents/ndma/NDMAstudies.pdf.
Cedergren, M. I., A. J. Selbing, O. Löfman, and B. A. J. Källen. 2002. “Chlorination byproducts and nitrate in drinking water and risk for congenital cardiac defects.” Environ. Res. 89 (2): 124–130. https://doi.org/10.1006/enrs.2001.4362.
Chen, B., S. N. Nam, P. K. Westerhoff, S. W. Krasner, and G. Amy. 2009. “Fate of effluent organic matter and DBP precursors in an effluent-dominated river: A case study of wastewater impact on downstream water quality.” Water Res. 43 (6): 1755–1765. https://doi.org/10.1016/j.watres.2009.01.020.
Chen, W. H., and T. M. Young. 2009. “Influence of nitrogen source on NDMA formation during chlorination of diuron.” Water Res. 43 (12): 3047–3056. https://doi.org/10.1016/j.watres.2009.04.020.
Chuang, Y.-H., and W. A. Mitch. 2017. “Effect of ozonation and biological activated carbon treatment of wastewater effluents on formation of N-nitrosamines and halogenated disinfection byproducts.” Environ. Sci. Technol. 51 (4): 2329–2338. https://doi.org/10.1021/acs.est.6b04693.
Dai, N., T. Zeng, and W. A. Mitch. 2015. “Predicting N-nitrosamines: N-nitrosodiethanolamine as a significant component of total N-nitrosamines in recycled wastewater.” Environ. Sci. Technol. Lett. 2 (3): 54–58. https://doi.org/10.1021/acs.estlett.5b00005.
Gerrity, D., A. N. Pisarenko, E. Marti, R. A. Trenholm, F. Gerringer, J. Reungoat, and E. Dickenson. 2015. “Nitrosamines in pilot-scale and full-scale wastewater treatment plants with ozonation.” Water Res. 72 (Apr): 251–261. https://doi.org/10.1016/j.watres.2014.06.025.
Hanigan, D., J. Zhang, P. Herckes, S. W. Krasner, C. Chen, and P. Westerhoff. 2012. “Adsorption of N-nitrosodimethylamine precursors by powdered and granular activated carbon.” Environ. Sci. Technol. 46 (22): 12630–12639. https://doi.org/10.1021/es302922w.
Krasner, S., P. Westerhoff, B. Chen, G. Amy, S. Nam, Z. Chowdhury, S. Sinha, and B. Rittmann. 2008. Contribution of wastewater to DBP formation. Denver, CO: AWWA Research Foundation.
Krasner, S. W., W. A. Mitch, D. L. McCurry, D. Hanigan, and P. Westerhoff. 2013. “Formation, precursors, control, and occurrence of nitrosamines in drinking water: A review.” Water Res. 47 (13): 4433–4450. https://doi.org/10.1016/j.watres.2013.04.050.
Krasner, S. W., S. Roback, Y. Qian, X-F. Li, R. Marfil-Vega, and Z. Bukhari. 2020. “Occurrence of nitrosamines and their precursors in North American drinking waters.” AWWA Water Sci. 2 (6): e1208. https://doi.org/10.1002/aws2.1208.
Krasner, S. W., P. Westerhoff, B. Chen, B. Rittmann, S.-N. Nam, and G. Amy. 2009a. “Impact of wastewater treatment processes on organic carbon, organic nitrogen, and DBP precursors in effluent organic matter.” Environ. Sci. Technol. 43 (8): 2911–2918. https://doi.org/10.1021/es802443t.
Krasner, S. W., P. Westerhoff, B. Chen, B. E. Rittmann, and G. Amy. 2009b. “Occurrence of disinfection byproducts in united states wastewater treatment plant effluents.” Environ. Sci. Technol. 43 (21): 8320–8325. https://doi.org/10.1021/es901611m.
Kristiana, I., D. Liew, R. K. Henderson, C. A. Joll, and K. L. Linge. 2017. “Formation and control of nitrogenous DBPs from Western Australian source waters: Investigating the impacts of high nitrogen and bromide concentrations.” J. Environ. Sci. 58 (Aug): 102–115. https://doi.org/10.1016/j.jes.2017.06.028.
Leavey-Roback, S., S. W. Krasner, and M. Suffet. 2016. “Veterinary antibiotics used in animal agriculture as NDMA precursors.” Chemosphere 164 (Dec): 330–338. https://doi.org/10.1016/j.chemosphere.2016.08.070.
Lee, C.-F. T., S. W. Krasner, M. J. Sclimenti, M. Prescott, and Y. C. Guo. 2015. “Nitrosamine precursors and wastewater indicators in discharges in the Sacramento-San Joaquin delta.” In Recent advances in disinfection by-products, 119–133. Washington, DC: American Chemical Society. https://doi.org/10.1021/bk-2015-1190.ch007.
Magnus, P., J. J. K. Jaakkola, A. Skrondal, J. Alexander, G. Becker, T. Krogh, and E. Dybing. 1999. “Water chlorination and birth defects.” Epidemiology 10 (5): 513–517. https://doi.org/10.1097/00001648-199909000-00008.
Mann, H. B., and D. R. Whitney. 1947. “On a test of whether one of two random variables is stochastically larger than the other.” Ann. Math. Stat. 18 (1): 50–60. https://doi.org/10.1214/aoms/1177730491.
Mitch, W., J. O. Sharp, R. R. Trussell, R. L. Valentine, L. Alvarez-Cohen, and D. L. Sedlak. 2003. “N-nitrosodimethylamine (NDMA) as a drinking water contaminant: A review.” Environ. Eng. Sci. 20 (5): 389–404. https://doi.org/10.1089/109287503768335896.
Mitch, W. A., and D. L. Sedlak. 2004. “Characterization and fate of N-nitrosodimethylamine precursors in municipal wastewater treatment plants.” Environ. Sci. Technol. 38 (5): 1445–1454. https://doi.org/10.1021/es035025n.
Munch, J., and M. Bassett. 2004. Method 521 determination of nitrosamines in drinking water by solid phase extraction and capillary column gas chromatography with large volume injection and chemical ionization tandem mass spectrometry (MS/MS). Cincinnati: National Exposure Research Laboratory, Office of Research and Development, USEPA.
Pearson, K. 1895. “Note on regression and inheritance in the case of two parents.” Proc. R. Soc. London 58 (387–352): 240–242. https://doi.org/10.1098/rspl.1895.0041.
Pearson, K. 1900. “On the criterion that a given system of deviations from the probable in the case of a correlated system of variables is such that it can be reasonably supposed to have arisen from random sampling.” London Edinburgh Dublin Philos. Mag. J. Sci. 50 (302): 157–175. https://doi.org/10.1080/14786440009463897.
Peto, R., R. Gray, P. Brantom, and P. Grasso. 1984. “Nitrosamine carcinogenesis in 5120 rodents: Chronic administration of sixteen different concentrations of NDEA, NDMA, NPYR and NPIP in the water of 4440 inbred rats, with parallel studies on NDEA alone of the effect of age of starting (3, 6 or 20 weeks) and of species (rats, mice or hamsters).” IARC Sci. Publ. 57: 627–665.
Plewa, M., E. Wagner, S. Richardson, A. Thruston, Y. Woo, and A. McKague. 2004. “Chemical and biological characterization of newly discovered iodoacid drinking water disinfection byproducts.” Environ. Sci. Technol. 38 (18): 4713–4722. https://doi.org/10.1021/es049971v.
Plewa, M. J., Y. Kargalioglu, D. Vankerk, R. A. Minear, and E. D. Wagner. 2002. “Mammalian cell cytotoxicity and genotoxicity analysis of drinking water disinfection by-products.” Environ. Mol. Mutagen. 40 (2): 134–142. https://doi.org/10.1002/em.10092.
Plewa, M. J., E. D. Wagner, M. G. Muellner, K.-M. Hsu, and S. D. Richardson. 2008. “Comparative mammalian cell toxicity of N-DBPs and C-DBPs.” In Disinfection by-products in drinking water, 36–50. Washington, DC: American Chemical Society. https://doi.org/10.1021/bk-2008-0995.ch003.
Prescott, M., S. W. Krasner, and Y. C. Guo. 2017. “Estimation of NDMA precursor loading in source water via artificial sweetener monitoring.” J. Am. Water Works Assn. 109 (6): E243–E251. https://doi.org/10.5942/jawwa.2017.109.0037.
Rice, J., S. H. Via, and P. Westerhoff. 2015. “Extent and impacts of unplanned wastewater reuse in US rivers.” J. Am. Water Works Assn. 107 (11): E571–E581. https://doi.org/10.5942/jawwa.2015.107.0178.
Rice, J., and P. Westerhoff. 2015. “Spatial and temporal variation in de facto wastewater reuse in drinking water systems across the U.S.A.” Environ. Sci. Technol. 49 (2): 982–989. https://doi.org/10.1021/es5048057.
Rice, J., A. Wutich, and P. Westerhoff. 2013. “Assessment of de facto wastewater reuse across the U.S.: Trends between 1980 and 2008.” Environ. Sci. Technol. 47 (19): 11099–11105. https://doi.org/10.1021/es402792s.
Richardson, S. D., M. J. Plewa, E. D. Wagner, R. Schoeny, and D. M. DeMarini. 2007. “Occurrence, genotoxicity, and carcinogenicity of regulated and emerging disinfection by-products in drinking water: A review and roadmap for research.” Mutat. Res. 636 (1–3): 178–242. https://doi.org/10.1016/j.mrrev.2007.09.001.
Roccaro, P., F. G. Vagliasindi, and G. V. Korshin. 2011. “Quantifying the formation of nitrogen-containing disinfection by-products in chlorinated water using absorbance and fluorescence indexes.” Water Sci. Technol. 63 (1): 40–44. https://doi.org/10.2166/wst.2011.006.
Russell, C. G., N. K. Blute, S. Via, X. Wu, and Z. Chowdhury. 2012. “Nationwide assessment of nitrosamine occurrence and trends.” J. Am. Water Works Assn. 104 (3): E205–E217. https://doi.org/10.5942/jawwa.2012.104.0037.
Schreiber, I. M., and W. A. Mitch. 2006. “Occurrence and fate of nitrosamines and nitrosamine precursors in wastewater-impacted surface waters using boron as a conservative tracer.” Environ. Sci. Technol. 40 (10): 3203–3210. https://doi.org/10.1021/es052078r.
Sgroi, M., F. G. A. Vagliasindi, S. A. Snyder, and P. Roccaro. 2018. “N-nitrosodimethylamine (NDMA) and its precursors in water and wastewater: A review on formation and removal.” Chemosphere 191 (Jan): 685–703. https://doi.org/10.1016/j.chemosphere.2017.10.089.
Singer, P. C. 1994. “Control of disinfection by-products in drinking water.” J. Environ. Eng. 120 (4): 727–744. https://doi.org/10.1061/(ASCE)0733-9372(1994)120:4(727).
Smirnov, N. 1948. “Table for estimating the goodness of fit of empirical distributions.” Ann. Math. Stat. 19 (2): 279–281. https://doi.org/10.1214/aoms/1177730256.
Swayne, M., G. Boone, D. Bauer, and J. Lee. 1980. Wastewater in receiving waters at water supply abstraction points. Washington, DC: USEPA.
USEPA. 1987. “N-nitrosodimethylamine CASRN 62-75-9.” Accessed April 2018. https://cfpub.epa.gov/ncea/iris2/chemicallanding.cfm?substance_nmbr=45.
USEPA. 2002. Effects of water age on distribution system water quality. Washington, DC: USEPA.
USEPA. 2006. National primary drinking water regulations: Stage 2 disinfectant and disinfection byproduct rule. Washington, DC: USEPA.
USEPA. 2007. Unregulated contaminant monitoring regulation (UCMR) for public water systems revisions. Washington, DC: USEPA.
USEPA. 2008. Clean watersheds needs survey 2004 report to . Washington, DC: USEPA.
USEPA. 2009. Drinking water contaminant candidate list 3—Final. Washington, DC: USEPA.
USEPA. 2012a. Organic chemicals, sampling and analytical requirements. Washington, DC: USEPA.
USEPA. 2012b. UCMR 2 (2008–2012) occurrence data. Washington, DC: USEPA.
USEPA. 2014. Announcement of preliminary regulatory determinations for contaminants on the third drinking water contaminant candidate list. Washington, DC: USEPA.
USEPA. 2016. Drinking water contaminant candidate list 4—Final. Washington, DC: USEPA.
USGS. 2018. “National water information system: Web interface.” Accessed April 1, 2018. https://waterdata.usgs.gov/nwis/rt.
Villanueva, C. M., et al. 2006. “Bladder cancer and exposure to water disinfection by products through ingestion, bathing, showering and swimming in pools.” Am. J. Epidemiol. 165 (2): 148–156. https://doi.org/10.1093/aje/kwj364.
Villanueva, C. M., K. P. Cantor, S. Cordier, J. J. K. Jaakkola, W. D. King, C. F. Lynch, S. Porru, and M. Kogevinas. 2004. “Disinfection byproducts and bladder cancer: A pooled analysis.” Epidemiology 15 (3): 357–367. https://doi.org/10.1097/01.ede.0000121380.02594.fc.
Zeng, T., C. M. Glover, E. J. Marti, G. C. Woods-Chabane, T. Karanfil, W. A. Mitch, and E. R. V. Dickenson. 2016. “Relative importance of different water categories as sources of N-nitrosamine precursors.” Environ. Sci. Technol. 50 (24): 13239–13248. https://doi.org/10.1021/acs.est.6b04650.
Information & Authors
Information
Published In
Journal of Environmental Engineering
Volume 147 • Issue 12 • December 2021
Copyright
This work is made available under the terms of the Creative Commons Attribution 4.0 International license, https://creativecommons.org/licenses/by/4.0/.
History
Received: Mar 4, 2021
Accepted: Jul 13, 2021
Published online: Sep 16, 2021
Published in print: Dec 1, 2021
Discussion open until: Feb 16, 2022
Authors
Metrics & Citations
Metrics
Citations
Download citation
If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download.
Cited by
- B. Coskun, N. Bilgin-Saritas, E. Aydin, E. Pehlivanoglu-Mantas, Identification of transformation products during Doxylamine chloramination for NDMA mitigation, Environmental Technology, 10.1080/09593330.2022.2135462, (1-16), (2022).