Purpose and Scope

The purpose of this report is to inform the development of MassDEP’s guidelines for the collection and use of site-dependent water-chemistry data (pH, DOC, and total hardness), which are required inputs to the EPA Calculator.

The following is a list of specific objectives for this report:

This report describes the approach used to select monitoring stations and the methods used to collect water-quality samples and provides water-quality data measured in situ or analyzed from collected samples. The report presents the site-dependent aluminum criteria values generated with the EPA Calculator for receiving-water bodies near 11 water-treatment facilities in eastern and central Massachusetts during 2018–19.

Water-quality samples were collected from multiple locations near the water-treatment facilities, including (1) stations upstream from effluent discharges (for streams and rivers) and locations outside of the immediate effect of the effluent discharge (for ponds), (2) effluent discharges, (3) stations downstream from effluent discharges (for streams and rivers), and (4) other locations as directed by MassDEP. Instantaneous water-quality criteria values for total recoverable aluminum were calculated only for water-quality samples collected at stations on water bodies designated by the EPA as the receiving-water body for the water-treatment facility and for stations where samples were collected upstream from effluent discharges (for streams and rivers) or in locations outside of the immediate effect of the effluent discharge (for ponds). Instantaneous water-quality criteria values for total recoverable aluminum were not determined for water-quality data collected from effluent discharges, from stations immediately downstream from effluent discharges (for streams and rivers) or within the dilution mixing zone (for ponds), or from stations not on the receiving-water body for a facility.

The aluminum criteria values generated by the EPA Calculator for selected stations using site dependent water-chemistry parameters are included in this report to demonstrate potential applications of the EPA’s 2018 ambient water-quality criteria guidance for aluminum in freshwater; the calculated values are not regulatory effluent limits for specific facilities. The formal adoption and use of any site-dependent aluminum calculations to restore and maintain the chemical, physical, and biological integrity of surface waters under the Clean Water Act (https://www.epa.gov/laws-regulations/summary-clean-water-act) are subject to additional Federal and State regulatory frameworks.

Previous Investigations

Water quality and toxicity related to aluminum in effluent discharges have been investigated by the USGS in two recent publications (Colman and others, 2016; Besser and others, 2019). In 2011, the USGS developed a method to assess dilution of aluminum in filter-backwash effluent discharged to reservoirs from water-treatment plants (Colman and others, 2016). The dilution of aluminum discharged to reservoirs was investigated in a field study and through computer simulation. Determination of dilution is needed so that permits for discharge ensure compliance with water-quality standards for aquatic life. Dilution statistics were simulated for 13 reservoirs from 1960 to 2004 using the USGS Firm-Yield Estimator model to simulate reservoir inputs and outputs and present-day values of filter-effluent discharge and aluminum concentration (Levin and others, 2011). In addition to computing dilution factors, Colman and others (2016) determined dilution factors that would mitigate aluminum toxicity with the same statistical basis (frequency of exceedance of the chronic standard) as dilutions computed for streams at the 7-day-average, 10-year-recurrence annual low flow (7Q10) (Blum and others, 2019).

In a followup study to Colman and others (2016), the USGS (Besser and others, 2019) investigated the importance of DOC as a control on aluminum toxicity in low‐hardness waters (20 mg/L and 35 mg/L as CaCO₃). A series of 7‐day toxicity tests were completed with the water flea Ceriodaphnia dubia in dilutions of low‐hardness natural waters, which contained DOC concentrations as much as 10 milligrams per liter (mg/L). Results from the study confirm the importance of DOC as a control on aluminum toxicity in low‐hardness waters and demonstrate that total aluminum concentrations are not predictive of aluminum toxicity, except under defined water quality (pH, DOC, and total hardness) and exposure conditions (Besser and others, 2019).

Selection of Sampling Locations

Selection of sampling locations is an important study component leading to the collection of high-quality water-quality samples. The physical setting varied at each of the facilities in relation to the water bodies considered in this study. Some of the challenges encountered during the selection of station location included the need to consider the potential effects on water quality caused by impoundment from downstream dams or flood-control reservoirs, water releases from upstream reservoirs or water imports from other water bodies, induced infiltration from appreciable groundwater withdrawals, stormwater runoff, and discharges from NPDES and Superfund sites. Station selection could also be complicated by difficult access conditions or denial of permission to access private property. Factors evaluated during site selection for each station are discussed in the station descriptions in appendix 1.

Where possible, stations were established away from the outfalls of stormwater runoff pipes and away from areas receiving road runoff to reduce the possibility that metals or other constituents from stormwater runoff would directly affect water quality. Stations on ponds were established away from the shore to minimize the effects of runoff on water quality. The station locations on streams and rivers also were selected to minimize the possibility that metals or other constituents in stormwater runoff from roads would be attributed to effluent discharges. For example, for stations on streams and rivers, upstream stations were established downstream from roads, and downstream stations were established upstream from roads.

Stations on streams and rivers were selected in reaches with single channels and in areas free of eddies, with minimum turbulence and streambed irregularities that may have affected sample collection. Study protocols for sampling ponds required knowledge of pond bathymetry to identify the deepest areas of the pond. Pond bathymetry was available for some ponds from published reports, and bathymetry was measured for other ponds. Shallow and deep stations were established in a deep area so that the water-quality conditions at different depths could be compared. Stations on ponds also were established in a deep area to reduce the probability that in the summer the pond water surface would be covered with aquatic vegetation that hampers the collection of shallow and deep water-quality samples. The ponds in the study area were less than 15 feet (ft) deep and small enough in area that station location with respect to the effluent discharge was a factor during selection of monitoring station locations.

Effluent samples were collected at the ends of the outfall pipes, where possible. However, selection of appropriate locations for collection of effluent samples was dictated by the structure, route, and location of effluent outfalls. Four of the facilities had conditions that prevented collection of a representative sample directly from the end of the outfall pipe. The effluent pipes at the Marlborough and Hudson WWTFs are partially submerged in the Assabet River, and the effluent pipes at the Maynard WWTF and at the Westborough WTF join with stormwater runoff pipes before reaching the effluent outfalls. For these facilities, effluent samples were collected at the locations where the facilities collect water-quality samples, either by pumping from the effluent stream with a peristaltic pump (Marlborough and Hudson) or by using the facility’s pumps (Maynard WWTF and Westborough WTF).

The location of effluent outfalls relative to the receiving-water body varied considerably by facility. Selection of a station downstream or downgradient from effluent discharge required assessment of how and where effluent discharges would reach the receiving-water body. Some facilities discharged effluent directly into receiving-water bodies (Cohasset, Leominster, Marlborough, Hudson, Maynard, Westborough-WTF). At other facilities, effluent was routed through ditches (Westborough-WWTF); swales, intermittent streams, or riparian wetlands (Fitchburg, Hanover, Wilmington); ponds (Weymouth); or other water bodies before discharging into the receiving-water body.

A list of factors considered during selection of monitoring stations was developed (table 3). The selection factors were developed to ensure that (1) monitoring stations were selected to allow collection of discrete samples during all flow conditions; (2) stormwater runoff and known point sources (other than those from the facility being monitored) would have minimal effects on water quality; and (3) site conditions at the ambient (upstream) or ambient (pond) stations facilitated the installation of a continuous sensor, datalogger, and equipment shelter. Manmade alterations that could potentially affect water quality that needed to be considered during the site-selection process included flow alterations, such as from flood-retention dams, downstream reservoirs, surface-water diversions, groundwater withdrawals, and stormwater; or contamination, such as from NPDES discharges, EPA Superfund sites, old landfills, and stormwater runoff. Other factors that affected site selection included tributaries; isolated locations; access issues; permission to access sites through private property; and safety concerns related to road traffic, boating, or wading hazards (USGS, 2014).

Figures 3–13 show maps of the receiving water bodies and monitoring stations associated with each of the water-treatment facilities. Some of the receiving water-bodies are on small tributaries or ponds, and the text includes mention of the downstream water bodies that the receiving waters flow into. Some of the downstream water bodies in the site descriptions are outside of the map areas shown in figures 3–13 and would be too small to identify on figure 1 and, therefore, do not appear in any figure.

Wastewater-Treatment Facilities

The Westborough, Marlborough, Hudson, and Maynard WWTFs (listed in order of increasing drainage area) are within about a 20-mile (mi) reach of the Assabet River. Water quality in the Assabet River at each of these facilities is potentially affected by effluent discharge from upstream water-treatment facilities. A tributary entering the Assabet River upstream of the Westborough WWTF contains the outflow from Hocomonco Pond, which received effluent discharge from the Westborough WTF during the study (2018–19). Water quality in the Assabet River at the Marlborough, Hudson, and Maynard WTFs is potentially affected by effluent discharges from upstream WWTFs.

Westborough Wastewater-Treatment Facility

The Assabet River is the EPA-designated receiving-water body for the effluent discharge from the Westborough WWTF. The headwaters of the Assabet River are in Westborough, and the Assabet River is a small stream with a drainage area of 8.37 square miles (mi²) at the point where the stream receives the effluent discharge. Contributing drainage areas for stations in this report were determined with StreamStats (USGS, 2016).

The following three monitoring stations are associated with the Westborough WWTF: a station upstream from the effluent discharge: Assabet River near Westborough, Mass. (station 01096603); an effluent station, Westborough wastewater-treatment plant effluent (station 421651071381401); and a station downstream from the effluent discharge, Assabet River downstream from Westborough water-treatment plant at Route 9 (Turnpike Road) (station 421700071381901) (fig. 3).

Marlborough Wastewater-Treatment Facility

The Assabet River is the EPA-designated receiving-water body for the effluent discharge from the Marlborough WWTF. The Marlborough WWTF is about 6 mi downstream from the Westborough WWTF, and the drainage areas for the upstream and downstream monitoring stations are 35.3 and 35.6 mi², respectively. The Marlborough WWTF discharges effluent to the Assabet River in a river reach upstream from the Tyler floodwater-retarding dam (fig. 4) (U.S. Department of Agriculture, 2013).

The following three monitoring stations are associated with the Marlborough WWTF: a station upstream from the effluent discharge, Assabet River at Boundary Street near Northborough, Mass. (station 01096720); an effluent station, Marlborough wastewater-treatment plant effluent (station 422034071365601); and a station downstream from the effluent discharge, Assabet River, downstream Marlborough wastewater-treatment plant (station 01096725) (fig. 4).

Hudson Wastewater-Treatment Facility

The Assabet River is the EPA-designated receiving-water body for the effluent discharge from the Hudson WWTF. The facility discharges to the Assabet River upstream from the Gleasondale impoundment (fig. 5). The Hudson WWTF is about 8 mi downstream from the Marlborough WWTF, and the Assabet River at the WWTF has a drainage area of 74 mi².

The following three monitoring stations are associated with the Hudson WWTF: a station upstream from the effluent discharge, Assabet River at Cox Street near Hudson, Mass. (station 01096870); an effluent station, Hudson wastewater-treatment plant effluent (station 422406071323401); and a station downstream from the effluent discharge, Assabet River near Hudson-Stow town line (station 01096875) (fig. 5).

Maynard Wastewater Treatment Facility

The Assabet River is the EPA-designated receiving-water body for the effluent discharge from the Maynard WWTF. The Maynard WWTF is about 9 mi downstream from the Hudson WWTF, and the contributing drainage area at the WWTF effluent discharge is 117 mi². The Maynard WWTF effluent is discharged into the Powdermill impoundment, about 2,010 ft upstream from the Powdermill Dam. The Maynard WWTF is constructed on a hillslope along the left side of the Assabet River near the Maynard and Acton town line. The Acton WWTF is just downstream (0.2 mi) of the Maynard WWTF. Effluent from the Acton WWTF is discharged to rapid-infiltration beds and moves to the Assabet River as groundwater.

The following three monitoring stations are associated with the Maynard WWTF: a station upstream from the effluent discharge, Assabet River, upstream Maynard wastewater-treatment plant (station 01097021); an effluent station, Maynard wastewater-treatment plant effluent (station 422627071262301); and a station downstream from the Maynard WWTF effluent discharge, Assabet River, downstream Maynard wastewater-treatment plant (station 01097023) (fig. 6).

Water-Treatment Facilities with Pond Receiving-Water Bodies

The three WTFs discussed here (Cohasset, Fitchburg, and Westborough WTF) have receiving-water bodies that are ponds. Depending on the location of the effluent discharge relative to the pond inflow and outflow, the designation of a monitoring station as upgradient or downgradient from effluent discharge was not always possible. Consequently, monitoring stations on ponds were established in areas thought to be outside of the immediate effect of the effluent discharge.

Cohasset Water-Treatment Facility

Lily Pond is the EPA-designated receiving-water body for the effluent discharge from the Cohasset WTF. The sources of water to the Cohasset WTF are from surface water (Cohasset Water Department, 2021). Lily Pond is 51 acres in area with a maximum depth of about 8 ft (ENSR International, 2003). The drainage area to Lily Pond is 2.5 mi² but totals 8.89 mi² if the drainage area above the Bound Brook Control Structure is included (not shown). The outflow of Lily Pond (Herring Brook) joins with the outflow from Aaron River Reservoir (Aaron River) to form Bound Brook, which is a tributary to the Gulf River estuary.

The following three monitoring stations are associated with the Cohasset WTF: two stations at a deep hole in the pond, Lily Pond deep hole (shallow) (station 421326070485802) and Lily Pond deep hole (station 421326070485801); and an effluent station, Cohasset water-treatment plant backwash effluent (station 421334070490601) (fig. 7).

Fitchburg Water-Treatment Facility

The EPA-designated receiving-water body for the effluent discharge from the Fitchburg WTF facility is Wyman Pond (fig. 8). The source of water for the Fitchburg WTF is surface water. The Fitchburg WTF and Wyman Pond are both in the town of Westminster, Mass. Wyman Pond is a reservoir constructed by damming Wyman Brook. The pond is 198 acres in area with a maximum depth of about 14 ft with a contributing drainage area of 11.8 mi² at the dam. The pond is long, narrow, and sinuous in shape at its inflow end and broadest near the outflow end of the pond. Flow at the inflow end of the pond is thought to generally be in one direction from the south toward the dam at the north end of the pond. The Fitchburg WTF effluent discharges to the pond on the northwest side of Wyman Pond about 0.5 mi from the dam. The outflow of Wyman Pond is over a spillway at the dam or through the gates at the gatehouse. Wyman Pond outflows to Wyman Brook, a tributary to Flag Brook, which flows to the North Nashua River (not shown).

The following five monitoring stations are associated with the Fitchburg WTF: two stations at the upgradient end of the pond, Wyman Pond, Leino Park Road, shallow, Westminster, Mass. (station 423132071523401), and Wyman Pond, Leino Park Road, deep, Westminster, Mass. (station 423132071523402); an effluent station, Fitchburg water treatment plant backwash effluent (station 423215071534101); and two stations at a deep hole near the outlet end of the pond, Wyman Pond shallow, Westminster, Mass. (station 423211071524701), and Wyman Pond deep, Westminster, Mass. (station 423211071524702) (fig. 8).

Westborough Water-Treatment Facility

Hocomonco Pond is the EPA-designated receiving-water body for the effluent discharge from the Westborough WTF. The source of water to the Westborough WTF is primarily from groundwater pumped from wells. Hocomonco Pond is a shallow 28-acre pond with a maximum depth of about 8 ft. A small intermittent tributary enters Hocomonco Pond from the northwestern side of the pond. The effluent from the Westborough WTF discharges into Hocomonco Pond on the western side of the pond, less than 300 ft southwest of the mouth of the intermittent tributary. The outflow from Hocomonco Pond is from the east side of the pond to a small brook that is tributary to the Assabet River upstream from the Westborough WWTF. Hocomonco Pond is an EPA Superfund site, and a pump-and-treat facility at the southeastern side of the pond pumps and treats groundwater to remove creosote from the aquifer underlying the pond (Hansen, 1993). The area is fenced on the eastern and southern sides of the pond, and permission from the EPA is needed to access the site.

The following four monitoring stations are associated with the Westborough WTF: an effluent station, Westborough water-treatment plant backwash effluent (station 421627071392401); two stations near the deep hole in the pond, Hocomonco Pond shallow, Westborough, Mass. (station 421622071385701), and Hocomonco Pond deep, Westborough, Mass. (station 421622071385702); and a station at the pond outlet, Hocomonco Pond near Otis Street, Westborough, Mass. (station 421628071384501) (fig 9).

Water-Treatment Facilities with Stream Receiving-Water Bodies

The four WTFs discussed in this section (Hanover, Leominster, Weymouth, and Wilmington) have receiving-water bodies that are streams. The effluent from the Weymouth WTF is initially discharged into a pond whose outflow is tributary to a brook that is the EPA-designated receiving-water body; therefore, Weymouth is included with facilities that have streams as receiving-water bodies.

Hanover Water-Treatment Facility

Third Herring Brook is the EPA-designated receiving-water body for the effluent discharge from the Hanover WTF. The source of water for the treatment plant is groundwater from an aquifer adjacent and underlying Third Herring and Silver Brooks (fig. 10). The effluent from the Hanover WTF does not discharge directly into Third Herring Brook. The effluent first discharges to a wetland, flows about 650 ft to Silver Brook, and then flows about 750 ft down Silver Brook to the juncture with Third Herring Brook. Downstream, Third Herring Brook flows through Old Pond Meadows about 1 mi to join with Wildcat Brook (not shown), then flows another 2 mi to join with the Indian Head River to form the North River, which discharges to the Atlantic Ocean about 10 mi downstream.

Streamflows in Third Herring Brook and in Silver Brook are affected by groundwater withdrawals by the production wells for the Hanover WTF and the Norwell WTF, which is on the northeastern side of Third Herring Brook. Silver Brook is generally dry in summer, possibly because the production wells deplete streamflow or intercept groundwater that would have discharged to the brook (Parker and Craddock, 2012, p. 33). During August 2018, Third Herring Brook dried completely in the area of the Hanover and Norwell well fields.

The following three monitoring stations are associated with the Hanover WTF: a station upstream from the effluent discharge, Third Herring Brook Pond Street near Hanover, Mass. (station 011058065); an effluent station, Hanover water-treatment plant backwash effluent (station 420754070495801); and a station downstream from the effluent discharge, Third Herring Brook, downstream water-treatment plant, near Hanover, Mass. (station 011058075) (fig. 10).

Leominster Water-Treatment Facility

Monoosnoc Brook is the EPA-designated receiving-water body for the effluent discharge from the Leominster WTF. The primary source of water for the Leominster WTF is from surface water. Monoosnoc Brook has a drainage area of 5.07 mi² at the Leominster WTF. Streamflow in the brook is altered by upstream reservoirs and water withdrawals. Downstream, Monoosnoc Brook flows into the city of Leominster where it joins with the North Nashua River (not shown).

The following three monitoring stations are associated with the Leominster WTF: a station upstream from the effluent discharge, Monoosnoc Brook upstream Leominster water-treatment plant (station 01094420); an effluent station, Leominster water-treatment plant backwash effluent (station 423258071480701); and a station downstream from the effluent discharge, Monoosnoc Brook, downstream Leominster water-treatment plant (station 01094422) (fig. 11).

Weymouth Water-Treatment Facility

The Mill River is the EPA-designated receiving-water body for the effluent discharge from the Weymouth WTF. The primary source of water to the Weymouth WTF is from surface water (fig. 12). Great Pond is in the headwaters of the Mill River. Downstream, Mill River joins with the Old Swamp River in Whitman’s Pond (not shown) and ultimately discharges to the Weymouth Back River estuary. The Weymouth WTF effluent is discharged to an embayment of Great Pond at the outlet (northern end) of Great Pond that is separated from water exchanges with the main body of Great Pond except at high pond water levels. The outflow from Great Pond flows through a control structure at the dam near Randolph Street (rebuilt in 2020) and down a small stream channel to the Mill River. Historically, the Mill River originated at Great Pond. The outflow from Great Pond is reported as the natural source of the Mill River in Wandle (1984), and the tributary paralleling the Massachusetts Bay Transportation Authority Kingston/Plymouth Commuter Rail Line is designated as the Mill River by O’Brien and others (2002). The tributary parallel to the rail line was designated as Mill River (segment MA74–04) by O’Brien and others (2002), and the station established on this tributary is designated as the ambient (upstream) station for Mill River for this report. The outflow from Great Pond joins Mill River about 0.5 mi downstream from Great Pond. Mill River is a small wetland stream at this juncture with a drainage area of 3.68 mi². A station was not established on Mill River downstream from where the outflow from Great Pond joins with the Mill River.

The following four monitoring stations are associated with the Weymouth WTF: a station upstream from where the outflow from Great Pond joins with the Mill River, Mill River near Randolph Street, South Weymouth, Mass. (station 01105587); an effluent station, Weymouth water-treatment plant effluent (station 420959070580401); and two stations on the pond embayment that receive the effluent discharge at the outlet (northern end) of Great Pond, Great Pond near outlet, shallow, South Weymouth, Mass. (station 421004070580201), and Great Pond near outlet, deep, South Weymouth, Mass. (station 421004070580202) (fig. 12).

Wilmington Water-Treatment Facility

Maple Meadow Brook is the EPA-designated receiving-water body for the Wilmington WTF (fig. 13). The water source to the Wilmington WTF is groundwater. The Wilmington WTF effluent does not discharge directly into Maple Meadow Brook. The effluent first discharges to a forested swale and flows downslope toward Sawmill Brook. The effluent moves through a riparian wetland before entering Sawmill Brook, which flows downstream through a large shrub wetland where it joins with Maple Meadow Brook. Water is impounded in the wetland behind a berm that is a remnant of an elevated reach of the Middlesex Canal, built in the early 1800s, and by a beaver dam at the stone abutment where Maple Meadow Brook passes through through the former canal crossing. Downstream, Maple Meadow Brook joins with Lubbers Brook (not shown) to form the headwaters of the Ipswich River.

The following four monitoring stations are associated with the Wilmington WTF: two stations on tributaries upstream from the effluent discharge, Sawmill Brook at Chestnut Street, Wilmington, Mass. (station 01101296), and Maple Meadow Brook, Wilmington, Mass. (station 01101294); an effluent station, Wilmington water-treatment plant backwash effluent (station 423200071100201); and a station downstream from the effluent discharge, Maple Meadow Brook at Middlesex Canal, Wilmington, Mass. (station 01101298), (fig. 13).

Collection and Processing of Water-Quality Samples

The following two types of water-quality data were collected at the monitoring stations: (1) monthly discrete water-quality samples (individual water samples collected at one discrete point in time) and associated field parameters (water temperature, pH, and specific conductance), and (2) continuous water-quality measurements (measurements recorded automatically with water-quality monitors at 15-minute intervals) of water temperature and pH.

Discrete Monthly Water-Quality Sampling

Discrete water-quality samples and field parameters were collected monthly at 38 stations for 10 to 13 months from April 2018 through May 2019. No discrete samples were collected from December 21, 2018, through January 28, 2019, because of a lapse of U.S. Government appropriations. After January 2019, discrete water-quality samples were only collected at ambient (upstream) stations and selected pond stations.

Discrete water-quality samples were collected during this study only when antecedent rainfall conditions at the stations met sampling criteria for “dry-weather” sampling to minimize the effect of stormwater runoff on water-quality conditions. Dry-weather conditions were defined as antecedent dry periods that had less than 0.1 inch (in.) of rain during a 1- to 3-day period before sampling. The length of antecedent dry period required for each station varied by water-body type: 1 day of antecedent dry conditions was required for ponds, 2 days for streams, and 3 days for rivers. All samples associated with a facility were collected on the same day. For facilities with stations on both streams and ponds (Weymouth), the longer antecedent dry period (2 days) was used. The stations on the Assabet River in Westborough and the Mill River in Weymouth (fig. 12) had small contributing drainage areas (less than 10 mi²) and were classified with stream stations (requiring 2 days of antecedent dry conditions).

A web-based Doppler rainfall map was used on the morning of each sampling day to determine which stations met the antecedent dry-period criteria. The rainfall maps were accessed from the iWeatherNet web page (iWeatherNet, 2021). The iWeatherNet rainfall maps are updated hourly from Doppler radar and provide estimates of rainfall for 24-, 48-, and 72-hour periods across Massachusetts. Rainfall from summer thunderstorms can be localized, and the Doppler rainfall maps provided more spatially representative coverage of rainfall conditions across Massachusetts than data from individual precipitation gages.

The magnitude of streamflow can affect water-quality conditions. During high-flow periods, concentrations of some water-quality constituents may decrease because of dilution or increase because of suspension of sediments and inputs from stormwater runoff. During low-flow periods, streamflow is predominantly base flow and groundwater can have a greater effect on stream water-quality conditions. In this study, water-quality samples were collected during a range of flow conditions. Streamflow conditions differ within and among years during the study period. Mean daily streamflow records from the USGS streamflow-gaging station on the Assabet River near Maynard, Mass. (station 01097000), were used to compare streamflow conditions during the study to historical streamflow conditions. A hydrograph showing mean daily streamflow during the study (water years 2018 and 2019) in comparison to the interquartile range (the area between the 25th and 75th percentile) of historical means of long-term discharge (based on the 1942–2017 period of record) indicates that streamflow conditions during the period of water-quality sampling were normal or above normal throughout most of the study period except for a few months in summer 2018 (June and July) when streamflow conditions were below normal (fig. 14).

Discrete water-quality samples were collected, processed, and analyzed for organic carbon, filtered calcium and magnesium, and total aluminum concentrations. Organic carbon was analyzed as DOC, which is defined as the organic carbon that passes through a 0.45-micrometer pore-size silver filter or 0.7-micrometer pore-size glass-fiber filter, and as TOC, which is unfiltered and includes particulates. DOC is considered the chemically reactive fraction and typically consists of a range of organic compounds, including fulvic and humic acids derived from root exudates, microbial biomass, and the decomposition of plant litter and soil organic matter (Thurman, 1985; Liu and others, 2014). Filtered calcium and magnesium were measured on samples after filtration through a 0.45-micrometer pore-size capsule filter. Hardness, which is caused primarily by the divalent cations calcium and magnesium dissolved in water, was calculated from the sum of filtered calcium and magnesium concentrations, in milliequivalents, and is reported in units of milligrams per liter (mg/L) as CaCO₃. Hardness is referred to by its EPA Substance Registry Services name of “total hardness” in this report for consistency with terminology used in the EPA Calculator (EPA, 2018). General guidelines for classification of water on the basis of hardness are as follows: 0 to 60 mg/L as CaCO₃ is classified as soft, 61 to 120 mg/L as CaCO₃ is classified as moderately hard, 121 to 180 mg/L as CaCO₃ is classified as hard, and more than 180 mg/L as CaCO₃ is classified as very hard (Hem, 1985). In general, water quality in Massachusetts tends toward lower hardness concentrations in comparison to other areas of the country containing carbonate bedrock (DeSimone, 2009). Water-quality samples collected for determination of aluminum concentrations were analyzed for total aluminum and were not filtered.

Discrete water-quality samples were collected according to standard USGS field sample- collection techniques for the collection of surface-water-quality data (USGS, [variously dated]). Before sampling, all water-quality sampling equipment was cleaned in the USGS New England Water Science Center laboratory in Northborough, Mass. Water-quality sampling bottles that required cleaning (plastic sampling bottles) were rinsed with deionized (DI) water in the Northborough laboratory on the morning of sampling. All plastic sampling bottles also were rinsed with site water before sample collection.

Water-quality samples were collected at river or stream stations using depth-integrated sampling techniques, by use of a depth-integrating hand sampler (DH–81) while wading (fig. 15A) or a depth-integrating hand-line or bridge board sampler (DH–95) while sampling from a bridge, or by pumping with a peristaltic pump (fig. 15B) or by a grab sample, depending on site conditions. When streamflow conditions were low and depth integrated sampling was not possible, grab samples were collected directly from the river using a prerinsed 2-liter (L) polyethylene bottle. In some situations, USGS protocols for sample collection were modified because of safety concerns or access issues. For example, the stations on the Assabet River in Hudson, Maynard, and Marlborough were remote and were accessed using a canoe with an electric motor. At these locations, when the river was too deep to wade and the current too swift to hold the canoe in place, samples were collected with a peristaltic pump with silicone tubing at a set depth (1.5 ft) below the water surface and by pumping directly into a prerinsed 2-L polyethylene bottle while ferrying the canoe back and forth across the river. Field parameters (water temperature, pH, and specific conductance) measured at these stations indicated that the river was well mixed, and samples are considered representative of instream conditions.

For stations on ponds, samples were collected with a peristaltic pump from a single depth and pumping directly into a prerinsed 2-L polyethylene bottle (fig. 15B). Shallow samples were collected at 1.5-ft depth. The sampling depth for deep samples varied; samples were collected as deep as possible while remaining above aquatic vegetation. Effluent grab samples were collected either directly from the discharging pipe or weir using a prerinsed 2-L polyethylene bottle or by pumping into the bottle with a peristaltic pump.

USGS techniques for trace-element sampling (Wilde and others, 1999) were used for sample collection: (1) equipment was constructed of noncontaminating materials and was cleaned rigorously before sample-collection events and between field sites, (2) equipment was handled in a manner that minimizes the chance of altering sample composition, and (3) samples were handled in a manner that prevented contamination.

Water-quality analytes and field parameters for discrete samples collected at and near the 11 water-treatment facilities and their respective USGS parameter codes, analysis methods, and minimum reporting limits are listed in table 4. Samples were filtered and acidified (for preservation) at the USGS laboratory in Northborough and shipped to the USGS National Water Quality Laboratory (NWQL) in Denver, Colorado, for analysis.

Table 4.    

Water-quality analytes and minimum reporting levels for discrete samples collected at and near seven water-treatment and four wastewater-treatment facilities in Massachusetts, 2018–19.

[Hardness is referred to in this report by its U.S. Environmental Protection Agency Substance Registry Services name of “total hardness.” USGS, U.S. Geological Survey; NWQL, National Water Quality Laboratory; µg/L, microgram per liter; mg/L, milligram per liter; NA, not available; CaCO₃, calcium carbonate; NWIS, National Water Information System. ]

Table 4.    Water-quality analytes and minimum reporting levels for discrete samples collected at and near seven water-treatment and four wastewater-treatment facilities in Massachusetts, 2018–19.

Analyte	Parameter name	USGS parameter code	Analysis Type	Analysis method	NWQL minimum reporting limit
Aluminum, unfiltered	Aluminum, water, unfiltered, recoverable, µg/L	01105	NWQL	Method I–4471-97 (Garbarino and Struzeski, 1998; Garbarino and others, 2006)	3 µg/L
Total organic carbon	Organic carbon, water, unfiltered, mg/L	00680	NWQL	Method 5310B (Clesceri and others, 1998)	0.7 mg/L
Dissolved organic carbon	Organic carbon, water, filtered, mg/L	00681	NWQL	Method 5310B (Clesceri and others, 1998)	0.23 mg/L
Calcium, filtered	Calcium, water, filtered, mg/L	00915	NWQL	Method I–1472–87 (Fishman, 1993)	0.022 mg/L
Magnesium, filtered	Magnesium, water, filtered, mg/L	00925	NWQL	Method I–1472–87 (Fishman, 1993)	0.01 mg/L
Specific conductance	Specific conductance, water, unfiltered, microsiemens per centimeter at 25 degrees Celsius	00095	Field	U.S. Geological Survey (2018)	NA
pH	pH, water, unfiltered, field, standard units	00400	Field	U.S. Geological Survey (2018)	NA
Water temperature, in degrees Celsius	Temperature, water, degrees Celsius	00010	Field	U.S. Geological Survey (2018)	NA
Total hardness	Hardness, water, mg/L as CaCO3 (calculated)	00900	Algorithm	Computation by NWIS algorithm	NA

Measurement of Field Parameters

Multiparameter water-quality monitors were used to measure field parameters including water temperature, pH, and specific conductance. Field parameters were measured in accordance with guidelines provided in Gibs and others (2007) for the use of multiparameter instruments for routine field measurements. Sensor calibration was checked the morning of each day of sampling and, if needed, sensors were calibrated before visiting field sites. Specific-conductance sensors were checked in standards covering the range in specific conductance expected at the sites to be sampled on that day based on previous field measurements. Sensors were calibrated when measurements differed from expected readings in standards by more than 5 microsiemens per centimeter (µS/cm) in standards of 100 µS/cm or less or by more than 3 percent for higher conductivity standards. The pH sensors were always calibrated when measurements in pH buffers differed from expected readings in pH standards by more than 0.2 SU and were generally calibrated if measurements differed from expected readings by more than 0.1 SU. These calibration procedures place a general limit on accuracy for pH of plus or minus 0.2 SU for the dataset when pH is considered in total. Calibration data were stored electronically in the water-quality monitors, and calibration log books were kept to help track sensor performance during the study.

Field parameters were measured using a multiparameter water-quality monitor immediately after collection of discrete water-quality samples and were measured “in situ,” meaning in the water body, and not “ex situ,” such as in water drawn from churn splitters, tubing, or sampling bottles. Monitors were submerged in site water upon arrival to a site before sample collection to allow time for equilibration, and readings were allowed to stabilize before recording a measurement. Measurements of field parameters in streams and rivers were made at six-tenths of the depth of the stream at each of the points where water-quality samples had been collected (fig. 16A). Samples were generally collected at 10 locations along a stream cross-section. For small streams (less than 5 to 10 ft wide), the number of sample locations depended on the stream width, and samples were collected at either three locations (left, center, right) or five locations (left, left center, center, right center, and right). If only three measurements were made, the value at the center measurement point was used as the final reported value. If more than three measurements were made, the median value of the cross-section measurements was used as the final value. Field measurements of water temperature, pH, and specific conductance collected across a section of a stream can be used to evaluate the magnitude of variability along the cross section and as an indicator of mixing in streams downstream from effluent discharges or tributary inputs. Measurements of field parameters in ponds (fig. 16B) were made at the same depth for which the discrete sample was collected. Field parameters were also measured along a vertical profile of the pond to evaluate the variability of water temperature, pH, and specific conductance with depth.

Measurements of field parameters associated with effluent samples were made at the location where the discrete sample was collected. Depending on the station, measurements of field parameters for effluent were measured where effluent flowed over a weir at the lagoon, at the end of the effluent outfall pipe, or in the channel downstream from the outfall if the channel contained only effluent discharge. For instances where effluent discharges were not of enough volume to submerge the monitor, field parameters were measured with a monitor inserted into a flow-through cell while pumping effluent into the cell with a peristaltic pump (fig. 16C). Data were recorded on paper field sheets and then entered in the USGS Personal Computer Field Form for facilitating data entry to the USGS National Water Information System (NWIS) database (USGS, 2020).

Continuous Monitoring of Water Quality

Regular daily fluctuations in pH can result because of the effect of photosynthesis by plants and respiration by aquatic biota and plants at night. During the day, photosynthesis by algae and aquatic vegetation produces dissolved oxygen and converts carbon dioxide to organic matter, causing dissolved carbon dioxide to decrease, increasing pH. During the night, an increase in respiration decreases dissolved oxygen and releases carbon dioxide. Increased metabolic carbon dioxide forms carbonic acid, which dissociates into bicarbonate and hydrogen ions, decreasing pH. The result can be a daily variation in pH, where the pH is typically at a minimum near dawn and at a maximum in mid-to-late afternoon (Nimick and others, 2011; Gammons and others, 2015; Andersen and others, 2017). This regular variation in pH over a 24-hour period (day plus night) is called a diel pH cycle. Slow-moving, highly productive streams, rivers, reservoirs, or ponds with open canopy cover can have large diel swings in pH (about 1 to 2 SU), whereas changes in pH may be small (about 0.1 to 0.2 SU) in forested upland streams.

In addition to photosynthesis and respiration, other factors, such as streamflow, weather (temperature, sunlight, and wind), and groundwater inflow can affect diel variations in water quality. For example, DOC concentrations can also fluctuate daily, seasonally, and with storm events. Recent development of instream optical sensors for measurement of fluorescent dissolved organic matter (FDOM) allows for continuous monitoring of FDOM, which can be used together with high frequency measurement of DOC concentrations to quantify DOC dynamics in streams (Saiers and others, 2021). However, the continuous monitoring of FDOM was not within the scope of this study.

Automatic-monitoring techniques were used to collect continuous (15-minute interval) measurements of water temperature and pH at a station near each of the 11 facilities to capture seasonal, diel, and event-driven fluctuations in water quality. Continuous measurements were recorded at stations upstream from effluent discharge for facilities where the receiving water was a stream or river, or at stations outside the immediate effect of effluent discharge for facilities where the receiving water was a pond (fig. 17A, B, table 2). Continuous measurements were recorded from June 2018 through November 2018 for all stations, with measurements for some stations beginning in April 2018 and extending into April 2019. Continuous measurements collected at the selected stations on Monoosnoc Brook (Leominster), Assabet River (Hudson), and Lily Pond (Cohasset) (table 2) were recorded for longer periods (through February 2019) to represent conditions in streams, rivers, and ponds, respectively.

Multiparameter water-quality monitors were deployed for continuous measurement of field parameters at the selected water-quality monitoring stations. Monitors were programmed to measure water temperature and pH at 15-minute intervals and reported readings to a datalogger. The datalogger was programmed to provide the power-distribution controls (turning sensors on and off) and time stamp, and to internally log all sensor data at predetermined sampling frequencies. The datalogger was interfaced with a cellular modem to allow remote-data acquisition and troubleshooting. Data were transmitted by cellular transmission at hourly intervals to the USGS New England Water Science Center Northborough office for uploading to the NWIS web. For stations on streams or rivers, the monitors were equipped with a stage sensor to help determine if the sensors were out of the water during low-flow conditions. Stations on ponds did not need stage sensors because the monitors were installed on floats, which rise and fall with changing water levels. Before deployment, all field water-quality sensors were calibrated in a laboratory-controlled environment. Water-temperature thermistors passed five-point temperature checks against a National Institute of Standards and Technology (https://www.nist.gov)-certified thermometer. Before each site visit, the field monitor’s pH sensor was checked for accuracy against known standards in the USGS Northborough laboratory and calibrated, if needed. Calibration log books were kept to aid in tracking the performance of water-quality sensors during the study.

Maintenance and calibration of continuous monitors followed standard protocols (Wagner and others, 2006; USGS, 2019, 2021a). All site monitors were visited monthly for routine calibration and maintenance. During each site visit, the water temperature, pH, and specific conductance measured by the site water-quality monitor were checked against field monitor measurements to track potentially changing environmental conditions and to aid in applying data corrections. This maintenance included cleaning the monitor and the deployment tube and checking the accuracy of the water-temperature and pH sensors. Measurements of water temperature and pH were taken before and after the sensors were cleaned to help determine the effects of sensor fouling between site visits (fig. 18A). The site monitor pH sensor was then checked in the field against pH standards (4.00, 7.00, and 10.00 SU) (fig. 18B) and calibrated, if needed (fig. 18C). Calibrations were done based on the protocols discussed in Wagner and others (2006); these protocols require calibration if readings in standards are not within 0.2 SU of standard values. If calibration of the pH sensor was needed, the sensor was calibrated to pH 7.0 SU and 4.0 SU and then checked against pH 10.0 SU standard for accuracy. Upon completion of each field visit, fouling and calibration corrections were applied to the data, if needed, in accordance with the procedures outlined in Wagner and others (2006). Data from the sensors were processed in accordance with USGS methods, including those described by Wagner and others (2006) and Pellerin and others (2013).

Water-quality monitoring data can be missing for periods because of instrument failure or environmental conditions. Three stations (Monoosnoc Brook, upstream Leominster water-treatment plant [station 01094420] [fig. 11], Assabet River near Westborough, Mass. [station 01096603] [fig. 3], and Mill River near Randolph Street, South Weymouth, Mass. [station 01105587] [fig. 12]), had short periods of missing data (about a month) because of instrument failure during the study period. In addition, water-quality data were not recorded at Wyman Pond, Leino Park Road, shallow, Westminster, Mass. (station 423132071523401) (fig. 8), during a week in October 2018 when the float and equipment were upended during the annual fall lake drawdown. Water-quality data were not recorded from Third Herring Brook in Hanover during late July, August, and early September 2018 because there was no flow in Third Herring Brook in the area of Third Herring Brook Pond Street near Hanover, Mass. (station 011058065) (fig. 10).

Quality Assurance and Quality Control

A quality-assurance program was designed to evaluate the accuracy and precision of the water-quality data collected during this study to ensure that results are representative of actual environmental concentrations. Collection and analysis of quality-control samples are mandated components of USGS water-quality field studies (Mueller and others, 2015). The goal of quality-control sampling is to identify, quantify, and document bias and variability in data that result from the collection, processing, shipping, and handling of samples and to help identify potential sources of sample contamination. Quality-assurance procedures consisted of using appropriate equipment-cleaning and sample-collection techniques and submitting quality-control samples for laboratory analysis.

Comparisons of Water Quality Among Ambient Stations

The water-quality parameters that affect aluminum toxicity (pH, DOC, and total hardness), varied among stations associated with the 11 facilities considered in this study. Water-quality conditions were often similar for one or more of the water-quality parameters (pH, DOC, or total hardness) for nearby stations that were upstream and downstream on the same river or stream or for stations with water-quality data collected from similar depths within the same pond. Although the data indicated some similarities among one or more water-quality parameters in similar water-body types (such as streams, rivers, impoundments, and ponds); ultimately, the aluminum criteria values calculated for stations were based on the local water chemistry conditions in the water body.

Boxplots of monthly discrete pH, DOC, total hardness, and aluminum in figures 19–22 show the values for the ambient (upstream) stations and pond stations outside of the immediate effect of effluent discharges. For facilities with multiple ambient (upstream) stations (Wilmington) or multiple pond stations (Cohasset, Fitchburg, and Westborough-WTF), the boxplots were developed using the combined data from those stations. Summaries of the water-quality results and comparisons of pH, DOC, total hardness, and aluminum among stations are reported below. The reported medians and minimum and maximum values were calculated from all discrete water-quality samples—no values were removed as outliers. The seasonal patterns in monthly discrete pH, DOC, total hardness, and aluminum are also discussed below, and plots showing monthly values for these parameters for each station are provided in the data release (Armstrong and others, 2022).

Comparisons of pH Among Ambient Stations

A comparison of pH values measured during the monthly discrete water-quality sampling indicated a wide range of variation of pH among stations (fig. 19). Overall, the pH values ranged from 5.0 to 8.6 SU. Median pH values for the stations varied by 1.3 SU (from 5.9 to 7.2 SU).

The pH values for the ambient (upstream) stations on the Assabet River in Westborough, Marlborough, Hudson, and Maynard (stations 01096603, 01096720, 01096870, and 01097021, respectively) had similar pH ranges that were among the highest observed during the study. Median pH values for these stations ranged from 6.9 to 7.2 SU. The pH values for stations on the Assabet River in Marlborough, Hudson, and Maynard indicated similar seasonal patterns. For these stations to have similar seasonal patterns was not unexpected because the stations are all on the Assabet River (fig. 1). Monthly pH values tended to be highest (more than 7.0 SU) from June through September, and lowest (less than 7.0 SU) from October through February. The lowest monthly pH values (6.5 SU) were measured in the headwaters of the Assabet River (Assabet River near Westborough, Mass. [station 01096603]) (fig. 3) in July, August, and September, and the highest pH value (8.6 SU) was measured in an impoundment at the most downstream station (Assabet River, downstream Maynard wastewater-treatment plant [station 01097023]) (fig. 6) during June.

The pH values for the pond stations associated with the Cohasset and Fitchburg WTFs were lower than those for the ambient (upstream) stations on the Assabet River associated with WWTFs. The median pH values for the ambient (upstream) stations and pond stations associated with WTFs ranged from 5.9 to 6.9 SU. Various sets of stations with similar pH values were in close geographic proximity to one another (Leominster and Fitchburg; and Cohasset and Weymouth), indicating that the pH values may be affected by basin characteristics such as bedrock type, glacial geology, soils, and vegetation.

Stations on Monoosnoc Brook (Leominster), Lily Pond (Cohasset), and Wyman Pond (Fitchburg) had some of the lowest pH values observed during the monthly discrete sampling. A pH value of 5.2 SU was measured in December at Monoosnoc Brook, upstream Leominster water-treatment plant (station 01094420); a pH value of 5.0 SU was measured in November at Lily Pond deep hole (shallow) (station 421326070485802) and Lily Pond deep hole (station 421326070485801); and a pH value of 5.5 SU was measured in September at Wyman Pond, Leino Park Road, deep, Westminster, Mass. (station 423132071523402). The low monthly pH values measured in fall and winter at many stations indicate the importance of collecting water-quality data throughout the year to collect the full range of water-quality conditions at a station.

Comparisons of Dissolved Organic Carbon Among Ambient Stations

The DOC concentrations for the ambient (upstream) stations on the Assabet River in Westborough, Marlborough, Hudson, and Maynard (stations 01096603, 01096720, 01096870, and 01097021, respectively) had similar values (fig. 20). Minimum and maximum monthly DOC concentrations for these stations ranged from 2.69 to 6.46 mg/L, and median monthly DOC concentrations ranged from 4.29 to 5.32 mg/L. The monthly DOC concentrations for the ambient (upstream) stations in Marlborough, Hudson, and Maynard indicated similar seasonal patterns and were highest in July (5.81, 6.39, and 6.46 mg/L, respectively) and lowest in March (2.80, 2.69, and 2.79 mg/L, respectively). DOC concentrations began declining after leaf fall, in November, and generally did not recover to November values until May.

The DOC concentrations at the stream and pond stations associated with WTFs had a wider range of values than DOC concentrations in the Assabet River associated with the WWTFs (fig. 20). The stations with the lowest DOC concentrations were on Wyman Pond in Fitchburg and Hocomonco Pond in Westborough. The DOC concentrations at Wyman Pond stations 1–4 (stations 423132071523401, 423132071523402, 423211071524701, and 423211071524702, respectively) (fig. 8) ranged from 1.68 to 4.65 mg/L, and the median DOC concentration for all four stations was 3.10 mg/L. The DOC concentrations at Hocomonco pond stations 1 and 2 (stations 421622071385701 and 421622071385702, respectively) and at Hocomonco Pond near Otis Street, Westborough, Mass. (station 421628071384501) (fig. 9), ranged from 1.92 to 5.17 mg/L, and the median DOC concentration for all three stations was 3.30 mg/L (fig. 20). The stations with the highest DOC concentrations and greatest range of DOC were for Lily Pond in Cohasset and Third Herring Brook in Hanover. The DOC concentrations for pond stations 1 and 2 on Lily Pond in Cohasset (stations 421326070485802 and 421326070485801, respectively) (fig. 7) ranged from 6.99 to 18.9 mg/L, and the DOC concentrations for the upstream station on Third Herring Brook in Hanover (station 011058065) (fig. 10) ranged from 4.12 to 11.7 mg/L.

Comparisons of Total Hardness Among Ambient Stations

The median monthly total hardness concentrations for the ambient (upstream) stations on the Assabet River in Marlborough, Hudson, and Maynard ranged from 47.2 to 86.7 mg/L as CaCO₃ (fig. 21). Of these stations, the highest total hardness concentration (172 mg/L as CaCO₃) was measured in July at Assabet River at Boundary Street near Northborough, Mass. (station 01096720), and the lowest total hardness concentration (41.8 mg/L as CaCO₃) was detected in November at Assabet River, upstream Maynard wastewater-treatment plant (station 01097021) (fig. 6). Total hardness concentrations at these stations indicated similar seasonal patterns. Total hardness concentrations at these stations were highest from May through October and lowest from November through April.

The total hardness concentrations detected at four of the stations on streams had similar values to those from the Assabet River. The median monthly total hardness concentrations ranged from 47.1 to 87.2 mg/L as CaCO₃ at Third Herring Brook (station 011058065), the Assabet River near Westborough (station 01096603), Mill River (station 01105587), and Sawmill and Maple Meadow Brooks (stations 01101296 and 01101294, respectively).

A comparison of total hardness concentrations among stations indicated that total hardness concentrations were lowest at the ambient (upstream) station on Monoosnoc Brook (Leominster) and the pond stations on Wyman Pond (Fitchburg) (fig. 21). Total hardness concentrations were also low for the pond stations on Lily Pond (Cohasset). The median monthly total hardness concentration at Monoosnoc Brook, upstream Leominster water-treatment plant (station 01094420) was 7.05 mg/L as CaCO₃, and monthly total hardness concentrations ranged from 5.13 to 39.7 mg/L as CaCO₃. The median monthly total hardness concentrations at Wyman Pond stations 1–4 (stations 423132071523401, 423132071523402, 423211071524701, and 423211071524702, respectively) ranged from 9.98 to 13.0 mg/L as CaCO₃ and the monthly total hardness concentrations ranged from 8.77 to 13.5 mg/L as CaCO₃. Median monthly total hardness concentrations were 25.7 and 24.6 mg/L as CaCO₃ at Lily Pond stations 1 and 2 (stations 421326070485802 and 421326070485801, respectively) (fig. 7), and the monthly total hardness concentrations for these stations ranged from 15.4 to 28.9 mg/L as CaCO₃.

Comparisons of Total Aluminum Concentrations Among Ambient Stations

A comparison of aluminum concentrations among ambient stations indicated that high aluminum concentrations (greater than 100 µg/L) were consistently detected for stations on Lily Pond in Cohasset and Monoosnoc Brook in Leominster and for five monthly samples at five additional stations, but that aluminum concentrations for most stations were less than 100 µg/L (fig. 22). Aluminum concentrations detected at the ambient (upstream) stations on the Assabet River in Westborough, Marlborough, Hudson, and Maynard ranged from 6.00 to 173 µg/L with median values ranging from 37.0 to 47.0 µg/L. Aluminum concentrations at ambient stations on streams and ponds varied more than concentrations in the Assabet River. Aluminum concentrations detected at Mill River near Randolph St., South Weymouth, Mass. (station 01105587), ranged from 15.5 to 440 µg/L, with a median value of 63.0 µg/L. Concentrations at the upstream stations in Leominster and Wilmington ranged from 15.8 to 208 µg/L with median values of 45.0 and 155 µg/L, respectively. During periods of low streamflow, aluminum concentrations at downstream stations in small streams may be affected by upstream effluent discharges. For example, during a period when Third Herring Brook in Hanover (fig. 10) dried completely at the upstream station and formed a series of pools at the downstream station, aluminum concentrations measured at the downstream station (station 420754070495801) on July 3, July 20, and August 8, 2018, were 1,980, and 1290, and 3,090 µg/L, respectively.

Aluminum concentrations for the pond stations ranged widely. Aluminum concentrations for samples collected at stations on Hocomonco Pond in Westborough were among the lowest measured during the study. Aluminum concentrations were less than 12 mg/L for over half (19 of 34) of the samples collected at pond stations 1 and 2 in Hocomonco Pond (stations 421622071385701 and 421622071385702, respectively), and at the pond outlet (Hocomonco Pond near Otis Street, Westborough, Mass. [station 421628071384501]). Median aluminum concentrations at these stations ranged from 11.9 to 15 µg/L. Aluminum concentrations measured at the stations in Wyman Pond ranged from 15.1 to 67.4 µg/L with median values ranging from 28.8 to 41.1 µg/L. In contrast, aluminum concentrations measured at the stations on Lily Pond in Cohasset (fig. 7) were among the highest measured during the study. Aluminum concentrations were between 200 and 370 µg/L for about half (11 of 23) of the samples collected at the Cohasset at Lily Pond deep hole (shallow) and Lily Pond deep hole (stations 421326070485802 and 421326070485801, respectively). The backwash effluent from the Cohasset WTF discharges to Lily Pond and may not circulate out of the pond during low-flow periods of the year when water is withdrawn from Lily Pond and seasonal releases from the Aaron River Reservoir flow into Lily Pond.

Water Quality of Effluent Discharges

The monthly discrete water-quality data for pH, DOC, total hardness, and total aluminum concentrations for the effluent stations monitored during the study are presented here. Effluent discharges can be a major factor contributing to water-quality variability (Carey and Migliaccio, 2009).

Comparisons of pH Among Effluent Stations

The water quality of effluent from WWTFs and WTFs differed. The pH of effluent samples from WWTFs varied over a narrower range than those from WTFs (fig. 23A). The pH values of effluent from the four WWTFs (Westborough, Marlborough, Hudson, and Maynard) were generally near neutral pH (7.0 SU), with median monthly pH values ranging from 6.8 to 7.1 SU and monthly pH values ranging from 6.6 to 7.3 SU. In contrast, the pH values of effluent from the seven WTFs varied within a wider range, with median monthly pH values ranging between 6.3 and 7.5 SU and monthly pH values ranging from 5.8 to 8.7 SU.

The effluent discharge from WWTFs may have some capacity to buffer pH values downstream. Plots of continuous pH data for the ambient (upstream) stations in Marlborough, Hudson, and Maynard (Armstrong and others, 2022) indicated that for many days during June, July, and August, the diel cycles had consistent minimum pH values near 7.0 SU.

Quality-Assurance and Quality-Control Results

The accuracy and precision of the data collected in this study were evaluated by making quality-control measurements at each of the monitoring stations and collecting various types of quality-control samples to monitor bias and precision in data collection. Quality-control samples for the study included equipment and field blanks, replicates, and laboratory spike samples. Quality-assurance and quality-control data are available in a table in a data release (Armstrong and others, 2022).

Westborough Wastewater-Treatment Facility

Stations near the Westborough WWTF (fig. 3) were visited monthly from April 13, 2018, through May 2, 2019, to collect discrete water-quality samples and measure water-quality field parameters. Most monthly discrete pH values for the ambient (upstream) station (Assabet River near Westborough, Mass. [station 01096603]) were near neutral pH (7.0 SU), ranging from 6.5 to 7.3 SU with a median value of 7.0 SU. Lowest pH values were from 6.5 to 6.6 SU from July through October. Continuous pH data were measured at the upstream station from April 26, 2018, through November 26, 2018. The continuous pH data indicate diel variations in pH were generally small. Diel variations in pH were greatest during May and least during November (median daily range in pH per month of 0.3 and 0.1 SU, respectively). Monthly discrete pH values for the ambient (downstream) station (Assabet River downstream from Westborough wastewater-treatment plant [station 01096725]) ranged from 6.6 to 7.2 SU with a median value of 7.0 SU.

Monthly DOC concentrations at the upstream station ranged from 3.15 to 7.56 mg/L with a median concentration of 5.32 mg/L. DOC concentrations were highest in October and November and lowest in March and April. Monthly DOC concentrations at the downstream station ranged from 3.72 to 6.46 mg/L with a median concentration of 4.34 mg/L.

Monthly total hardness concentrations at the upstream station ranged from 38.8 to 73.3 mg/L as CaCO₃ with a median total hardness concentration of 47.2 mg/L as CaCO₃. All monthly samples were in the range from 38 to 55 mg/L as CaCO₃. except for July 2018 (73.3 mg/L as CaCO₃). Monthly total hardness concentrations at the downstream station ranged from 67.6 to 402 mg/L as CaCO₃, with a median total hardness concentration of 144 mg/L as CaCO_3.

Aluminum concentrations at the upstream station ranged from 14.7 to 129 µg/L with a median concentration of 47.0 µg/L. Aluminum concentrations at the downstream station ranged from 28.6 to 166 µg/L, with a median concentration of 69.1 µg/L.

The pH values for the effluent station (Westborough wastewater-treatment plant effluent [station 421651071381401]) (fig. 3) were near neutral (7.0 SU) and varied within a similar range as the ambient stations, ranging from 6.7 to 7.2 SU with a median value of 7.0 SU. Effluent DOC concentrations were relatively constant and ranged from 3.01 to 4.13 mg/L, with a median value of 3.74 mg/L. The effluent total hardness concentrations varied more than effluent pH and DOC values. The effluent total hardness concentrations ranged from 140 to 664 mg/L as CaCO₃, with a median value of 192 mg/L as CaCO_3. High total hardness concentrations at the downstream station during August and December followed a similar pattern of total hardness concentrations in the effluent. Aluminum concentrations in effluent discharge ranged from 51.9 to 95.0 µg/L, with a median concentration of 75.7 µg/L. Aluminum concentrations in the effluent discharge were generally higher than concentrations at the upstream station (a median of 34 µg/L higher) and likely contributed to higher aluminum concentrations at the downstream station.

Marlborough Wastewater-Treatment Facility

Stations near the Marlborough WWTF (fig. 4) were visited monthly from May 4, 2018, through May 3, 2019, to collect discrete water-quality samples and measure water-quality field parameters. Monthly discrete pH values for the ambient (upstream) station (Assabet River at Boundary Street near Northborough, Mass. [station 01096720]) ranged from 6.7 to 7.6 SU, with a median value of 7.2 SU. The monthly pH values for the upstream Marlborough station were among the highest of all stations, with pH values higher than 7.0 SU for 75 percent of the samples. The lowest pH value (6.7 SU) was measured at the upstream station during November. Continuous pH data were measured at the upstream station from May 24, 2018, through February 4, 2019. The continuous pH data indicated diel variations in pH in the Assabet River in Marlborough were larger than variations in pH measured upstream at the Assabet River stations associated with the Westborough WWTF. The daily pH values for the upstream Marlborough station varied by as much as 1.5 SU. These diel variations in pH were likely caused by photosynthesis and respiration of algae, aquatic vegetation, and periphyton in the large wetland reach of the Assabet River just upstream from the Marlborough monitoring stations. The greatest diel pH variation (1.7 SU) was measured on July 17–18, 2018. Diel variations in pH were greatest during June and least during November (median daily range in pH per month of 0.7 and 0.1 SU, respectively). Monthly discrete pH values for the downstream station (Assabet River, downstream Marlborough wastewater-treatment plant [station 01096725]) (fig. 4) ranged from 6.8 to 7.3 SU with a median value of 7.1 SU. Monthly discrete pH values were similar at the upstream and downstream stations and did not indicate a seasonal pattern.

The monthly DOC concentrations at the upstream station ranged from 2.80 to 5.81 mg/L, with a median concentration of 4.31 mg/L. Monthly DOC concentrations at the downstream station were similar to those at the upstream station and ranged from 3.00 to 5.58 mg/L, with a median concentration of 4.60 mg/L. The DOC concentrations were lowest (less than 3 mg/L) from February through March.

Total hardness concentrations were similar at the upstream and downstream stations. Monthly total hardness concentrations at the upstream station ranged from 60.1 to 172 mg/L as CaCO₃, with a median total hardness concentration of 86.7 mg/L as CaCO₃ (station 01096725). Monthly total hardness concentrations at the downstream station ranged from 61.5 to 174 mg/L as CaCO₃, with a median total hardness concentration of 114 mg/L as CaCO₃. Total hardness concentrations were highest from May through October (greater than 100 mg/L as CaCO₃).

Aluminum concentrations at the upstream station ranged from 10.2 to 86.0 µg/L with a median concentration of 42.0 µg/L. Aluminum concentrations at the downstream station ranged from 28.1 to 82.0 µg/L with a median concentration of 43.6 µg/L.

The pH values for the effluent station (Marlborough wastewater-treatment plant effluent [station 422034071365601]) (fig. 4) ranged from 6.6 to 7.0 SU with a median of 6.8 SU. The pH of the effluent tended to be lower (median 0.4 SU lower) than values for the upstream station during most months. Effluent DOC concentrations were relatively constant and ranged from 3.11 to 4.68 mg/L with a median of 4.04 mg/L. The total hardness concentration of the effluent ranged from 195 to 284 mg/L as CaCO₃ with a median of 264 mg/L as CaCO₃. The total hardness values in the effluent were typically about 100 mg/L higher than total hardness in the river. Monthly aluminum concentrations in the effluent ranged from 4.1 to 10.0 µg/L with a median of 6.5 µg/L and were generally lower (median 31 µg/L lower) than those at the upstream station.

Hudson Wastewater-Treatment Facility

Stations near the Hudson WWTF (fig. 5) were visited monthly from April 24, 2018, through May 3, 2019, to collect discrete water-quality samples and measure water-quality field parameters. The monthly discrete pH values for the ambient (upstream) station (Assabet River at Cox Street near Hudson, Mass. [station 01096870]), ranged from 6.8 to 7. 3 SU with a median of 7.0 SU. The lowest pH values for the upstream station (6.8 SU) were measured from October through December 2018. Continuous pH data were measured at the upstream station from April 20, 2018, through April 29, 2019. The diel variability of pH was greatest during June through September, when pH varied by as much as 0.7 to 1.2 SU. Diel variations in pH were greatest during June and least during February (median daily range in pH per month of 0.4 and 0.1 SU, respectively). The monthly discrete pH values for the ambient (downstream) station (Assabet River near Hudson-Stow town line [station 01096875]) ranged from 6.8 to 7.3 SU with a median of 6.9 SU.

The monthly DOC concentration at the upstream station ranged from 2.69 to 6.39 mg/L with a median of 4.29 mg/L. The monthly DOC concentrations at the downstream station had similar values to those at the upstream station and ranged from 3.31 to 6.30 mg/L with a median of 4.46 mg/L. The DOC concentrations were lowest (less than 4.0 mg/L) at the upstream station from December through March. The monthly total hardness concentrations at the upstream station ranged from 53.0 to 140 mg/L as CaCO₃ with a median of 75.7 mg/L as CaCO_3.

Monthly total hardness concentrations at the downstream station ranged from 54.6 to 130 mg/L as CaCO₃ with a median of 83.9 mg/L as CaCO₃. Total hardness concentrations were highest (greater than 80 mg/L as CaCO₃) from May through October and lowest (less than 60 mg/L) as CaCO₃ in February.

Aluminum concentrations at the upstream station ranged from 22.3 to 173 µg/L with a median value of 42.0 µg/L. Aluminum concentrations at the downstream station ranged from 18.1 to 77.4 µg/L with a median value of 41.0 µg/L.

The pH values for the effluent station (Hudson wastewater-treatment plant effluent [station 422406071323401]) were near neutral and ranged from 6.9 to 7.3 SU with a median of 7.1 SU. Effluent DOC concentrations ranged from 2.90 to 4.57 mg/L with a median of 3.38 mg/L. Effluent total hardness concentrations varied little monthly and ranged from 66.2 to 82.3 mg/L as CaCO₃ with a median of 73.1 mg/L as CaCO₃. Aluminum concentrations in the effluent were low and ranged from 3.00 to 10.5 µg/L with a median of 5.65 µg/L. Effluent aluminum concentrations were always lower (median 36 µg/L lower) than aluminum concentrations in the river at the upstream and downstream stations.

Maynard Wastewater-Treatment Facility

Stations near the Maynard WWTF (fig. 6) were visited monthly from April 24, 2018, through May 3, 2019, to collect discrete water-quality samples and measure water-quality field parameters. The monthly discrete pH values for the Maynard stations were among the highest of all stations. The pH values for the ambient (upstream) station (Assabet River, upstream Maynard wastewater-treatment plant [station 01097021]) (fig. 6) ranged from 6.6 to 8.4 SU with a median of 7.2 SU. The pH values indicated a seasonal pattern—pH values were lowest (less than 7.0 SU) from October through February, and highest (8.4 SU) in June. Continuous pH data were measured at the upstream station from May 3, 2018, to February 8, 2019. The diel variability of pH was greatest from July through September, when pH varied by as much as 1.5 to 1.8 SU. The daytime maximum pH values tended to result in early afternoon and were maintained for short periods of 1 to 2 hours, in comparison to the nighttime minimum pH values that were often maintained for 6 hours or more, from near midnight to shortly after dawn. The highest pH measured was 8.9 SU on August 6, 2018. The lowest pH values were measured during September (pH 6.5–6.6 SU, September 21–24, 2018), coinciding with heavy rains and flooding caused by the remnants of Hurricane Florence (https://www.weather.gov/ilm/HurricaneFlorence). Diel variations in pH were greatest during June (median daily range in pH per month of 0.6 SU) and least during November and December (median daily range in pH per month of 0.1 SU). The pH values for the ambient (downstream) station (Assabet River, downstream Maynard wastewater-treatment plant [station 01097023]) ranged from 6.5 to 8.6 SU with a median of 7.1 SU.

The monthly DOC concentrations at the Maynard upstream station (fig. 6) ranged from 2.79 to 6.46 mg/L with a median of 4.57 mg/L The monthly DOC concentrations at the Maynard downstream station were similar to those at the upstream station and ranged from 3.14 to 6.52 mg/L with a median of 4.96 mg/L. The DOC concentrations had a seasonal pattern and were lowest (less than 5 mg/L) from November through April and highest (greater than 5 mg/L) from July through October.

The monthly total hardness concentrations at the upstream station ranged from 41.8 to 101 mg/L as CaCO₃ with a median of 64.5 mg/L as CaCO_3. The monthly total hardness concentrations at the downstream station ranged from 42.0 to 101 mg/L as CaCO₃ with a median of 72.6 mg/L as CaCO_3. Total hardness concentrations indicated a slight seasonal pattern; total hardness values were higher from May through October 2018 and lower from December 2018 through May 2019. The total hardness was highest (101 mg/L as CaCO₃) in September 2018 and lowest (42 mg/L as CaCO₃) in November 2018.

Aluminum concentrations at the upstream station ranged from 6.00 to 67.8 µg/L with a median of 37.0 µg/L. Aluminum concentrations at the downstream station ranged from 20.2 to 79.8 µg/L with a median of 41.5 µg/L.

The pH values for the effluent station (Maynard wastewater-treatment plant effluent [station 422627071262301]) (fig. 6) ranged from 6.6 to 7.2 SU with a median of 6.9 SU. The pH values of the effluent tended to be a few tenths lower than the pH values in the river most months (average 0.4 SU lower). Effluent DOC concentrations ranged from 4.06 to 6.48 mg/L with a median of 5.44 SU. Effluent total hardness concentrations ranged from 111 to 138 mg/L as CaCO₃ with a median of 127 mg/L as CaCO₃. Total hardness concentrations in the effluent were consistently about 50 mg/L higher than the total hardness values in the river for the upstream station; however, the higher concentrations were not reflected at the downstream station. Aluminum concentrations in effluent ranged from 70.3 to 341 µg/L with a median of 155 µg/L. Aluminum concentrations in the effluent were consistently higher (median 120 µg/L higher) than those at the upstream station. These higher aluminum concentrations were not generally reflected at the downstream station, likely because of dilution of effluent in the Powdermill impoundment (fig. 6).

Cohasset Water-Treatment Facility

Stations near the Cohasset WTF (fig. 7) were visited monthly from April 27, 2018, through May 1, 2019, to collect discrete water-quality samples and measure water-quality field parameters. The monthly pH values for Lily pond station 1 (Lily Pond deep hole [shallow] [station 421326070485802]) ranged from 5.0 to 7.9 SU with a median of 6.2 SU. Monthly pH values for Lily Pond station 2 (Lily Pond deep hole [station 421326070485801]) ranged from 5.0 to 7.4 SU with a median of 5.8 SU. The monthly pH values for Lily Pond stations 1 and 2 were among the lowest values of all stations in the study; more than one-half of the measurements collected at Lily Pond station 2 were less than 6 SU. The monthly pH values for the pond varied but did not indicate a distinct seasonal pattern. The lowest pH values (5.0 SU) were measured during November 2018 at both the shallow and deep stations. Continuous pH data were measured at Lily Pond station 1 from May 16, 2018, through December 19, 2018. The continuous pH data indicated large diel variations (between 1.0 and 1.5 SU) during September 2018. These daily cycles in pH are caused by photosynthesis and respiration of aquatic plants and algae. Although the range of diel variations in pH decreases during the fall, examination of continuous pH data in November indicates that large diel variations in pH can still result late in the growing season. Diel variability in pH was greatest during September and least during December (median daily range in pH per month of 1.2 and 0.2 SU, respectively). During 2018, the daily maximum pH value at Lily pond station 1 was 8.0 SU or greater for 11 days, including 10 days during September (8.0 to 8.8 SU) and 1 day during October (8.8 SU). During October and November, the daily minimum pH values were 5.5 SU or less (from 4.9 to 5.5 SU) for 9 days and 6.0 SU or less for 42 days. More than one-half (28) of the days with low pH (6.0 SU or less) were in the fall, including 8 days during October, 14 days during November, and 6 days during December.

The monthly DOC concentrations at Lily Pond stations 1 and 2 (fig. 7) ranged from 6.99 to 18.9 mg/L with a median of 10.6 mg/L and from 7.22 to 18.4 mg/L with a median of 11.2 mg/L, respectively. The monthly DOC concentrations at Lily Pond stations 1 and 2 were among the highest measured during the study. More than half of the monthly samples had DOC concentrations of 10.0 mg/L or higher. The DOC concentrations had a seasonal pattern and were lowest (less than 8.0 mg/L) at the Lily Pond station 1 during April and highest (18.9 mg/L) during November 2018.

The total hardness concentrations at Lily Pond stations 1 and 2 ranged from 17.5 to 28.9 mg/L as CaCO₃ with a median of 25.6 as CaCO₃ and from 15.4 to 28.7 mg/L as CaCO₃ with a median of 24.6 as CaCO₃, respectively. Total hardness concentrations did not indicate a distinct seasonal pattern and were lowest (less than 20 mg/L as CaCO₃) in November 2018 and May 2019.

Aluminum concentrations at Lily Pond station 1 ranged from 43.3 to 370 µg/L with a median of 164 µg/L. Aluminum concentrations at Lily Pond station 2 ranged from 27.3 to 357 µg/L with a median of 194 µg/L.

The pH values for the effluent station (Cohasset water-treatment plant backwash effluent [station 421334070490601]) (fig. 7) ranged from 6.1 to 6.9 SU with a median of 6.3 SU. The effluent DOC concentrations ranged between 3.35 and 5.36 mg/L with a median of 3.82 mg/L. The pH values and DOC concentrations in the effluent were less variable than those in Lily Pond. The total hardness concentrations in the effluent ranged from 23.3 to 35.3 mg/L as CaCO₃ with a median of 27.7 mg/L as CaCO₃. The total hardness concentrations in the effluent were generally within the same range as total hardness concentrations in Lily Pond.

Effluent aluminum concentrations collected from April 27 through December 14, 2018, ranged from 101 to 207 µg/L with a median value of 186 µg/L. Most of these effluent samples were not collected during a filter backwash but were collected from lagoon supernatant by lifting the top board of the weir. Two effluent samples collected during scheduled sampling but under different discharge conditions had higher aluminum concentrations. A sample collected in February 2019 from a stream of effluent leaking through boards in the weir had an aluminum concentration of 8,770 µg/L. A sample collected at the effluent outfall in March 2019 after a filter backwash when the effluent did not have an opportunity to settle in the lagoon because the lagoons were frozen had an aluminum concentration of 68,900 µg/L.

Fitchburg Water-Treatment Facility

Stations near the Fitchburg WTF (fig. 8) were visited monthly from April 23, 2018, through April 25, 2019, to collect discrete water-quality samples and measure water-quality field parameters. The pH values for pond station 1 (Wyman Pond, Leino Park Road, shallow, Westminster, Mass. [station 423132071523401]) ranged from 5.6 to 7.3 SU with a median of 6.2 SU. The pH values for pond station 2 (Wyman Pond, Leino Park Road, deep, Westminster, Mass. [station 4231320715234012]) had similar values as those at Wyman Pond station 1 and ranged from 5.5 to 7.3 SU with a median of 6.2 SU. The monthly pH values varied but did not indicate a seasonal pattern. The lowest pH values were measured during September (5.5 SU) and November (5.6 SU) at pond station 1. The pH values for pond station 3 (Wyman Pond shallow, Westminster, Mass. [station 423211071524701]) ranged from 6.1 to 6.4 SU with a median of 6.2 SU. The pH values for pond station 4 (Wyman Pond deep, Westminster, Mass. [station 423211071524702]) had similar values as those at Wyman Pond station 3 and ranged from 5.7 to 6.5 SU with a median of 6.3 SU.

Continuous pH data were measured at Wyman Pond station 1 from May 24, 2018, through November 27, 2018. This station is upgradient of effluent discharges to Wyman Pond. The continuous pH data indicated that diel variations were largest (0.5–1.0 SU) in late July and August 2018. Diel variations in pH were greatest during August and least during October (median daily range in pH per month of 0.5 and 0.1 SU, respectively). The lowest observed pH value (6.0 SU) was measured during late September. Minimum pH values were less than 6.2 SU for 2 days in early September, 11 days in late September and early October, and 5 days in late November.

The monthly DOC concentrations at Wyman Pond stations 1 and 2 ranged from 1.68 to 4.71 mg/L with a median of 2.84 mg/L and from 1.72 to 4.65 mg/L with a median of 3.26 mg/L, respectively. The monthly DOC concentrations at Wyman Pond stations 3 and 4 ranged from 2.16 to 3.89 mg/L with a median of 3.11 mg/L and from 2.17 to 4.10 mg/L with a median of 3.09 mg/L, respectively. The monthly DOC concentrations at the Wyman Pond stations were among the lowest during the study, with 90 percent of the monthly DOC concentrations less than 4 mg/L. Although low, the DOC concentrations had a distinct seasonal pattern; values generally increased from spring into fall, peaked during September, and then decreased from fall to early spring, with the lowest value (1.68 mg/L) in April.

Total hardness values for Wyman Pond stations 1 and 2 (fig. 8) ranged from 8.77 to 13.0 mg/L as CaCO₃ with a median of 9.98 mg/L as CaCO₃ and from 8.78 to 12.8 mg/L as CaCO₃ with a median of 11.4 mg/L as CaCO₃, respectively. Total hardness values for Wyman Pond stations 3 and 4 ranged from 10.1 to 13.5 mg/L as CaCO₃ with a median of 12.9 mg/L as CaCO₃ and from 10.8 to 13.5 mg/L as CaCO₃ with a median of 13.0 mg/L as CaCO₃, respectively. Total hardness concentrations at the Wyman Pond stations were among the lowest measured during the study, with 90 percent of the monthly values less than 13.5 mg/L as CaCO₃. The lowest concentrations (from 8.77 to 8.82 mg/L as CaCO₃) were measured at Wyman Pond stations 1 and 2 during November 2018 and April 2019, respectively.

Aluminum concentrations at Wyman Pond stations 1 and 2 ranged from 23.6 to 67.4 µg/L with a median of 41.0 µg/L and from 25.3 to 64.1 µg/L with a median of 31.7 µg/L, respectively. Aluminum concentrations at Wyman Pond stations 3 and 4 ranged from 15.1 to 62.0 µg/L with a median of 28.9 µg/L and from 16.2 to 62.0 µg/L with a median of 28.8 µg/L, respectively. Aluminum concentrations at the Wyman Pond stations were lowest (less than 30 µg/L) in June, July, and August. Although aluminum concentrations differed slightly between Wyman Pond stations 1 and 2 and Wyman Pond stations 3 and 4, the monthly aluminum concentrations at the paired shallow and deep stations tended to be similar.

The pH values for the effluent station (Fitchburg water-treatment plant backwash effluent [station 423215071534101]) (fig. 8) ranged from 6.4 to 7.6 SU with a median of 6.7 SU. Effluent pH values were higher than pH values in Wyman Pond. Effluent DOC concentrations ranged from 2.06 to 2.54 mg/L with a median of 2.18 mg/L. Effluent DOC concentrations were lower than those in Wyman Pond. Effluent total hardness concentrations ranged from 12.0 to 14.0 mg/L as CaCO₃ with a median of 13.0 mg/L as CaCO₃. Effluent total hardness concentrations were similar to concentrations measured in Wyman Pond.

Effluent aluminum concentrations ranged from 318 to 10,800 µg/L with a median of 806 µg/L. The effluent aluminum concentration of 10,800 µg/L was measured during May 2018 during a period when the lagoons were being cleaned. The aluminum concentrations in the effluent were higher than concentrations measured in Wyman Pond. The effluent from the Fitchburg WTF lagoons is discharged into an intermittent stream containing some beaver dams, potentially allowing for additional settling of aluminum before discharging to Wyman Pond. The relatively large size of Wyman Pond, compared to other ponds considered in this study, likely also provides for dilution of effluent discharge.

Westborough Water-Treatment Facility

Stations near the Westborough WTF (fig. 9) were visited monthly from April 18, 2018, through April 25, 2019, to collect discrete water-quality samples and measure water-quality field parameters. The pH values for pond station 1 (Hocomonco Pond shallow, Westborough, Mass. [station 421622071385701]) ranged from 6.2 to 7.5 SU with a median of 6.8 SU. The pH values for pond station 2 (Hocomonco Pond deep, Westborough, Mass. [station 421622071385702]) had similar values as those at Hocomonco Pond station 1 and ranged from 6.0 to 7.2 SU with a median of 6.8 SU. The pH values varied monthly with some seasonal differences. The lowest pH values (6.0–6.2 SU) were measured during October 2018. Continuous pH data were measured at the Hocomonco Pond station 1 from May 4, 2018, through December 4, 2018. The largest diel pH variations (1.0–2.0 SU) were measured during July 2018. These daily cycles in pH are likely caused by photosynthesis and respiration of aquatic plants and algae. Diel variations in pH were greatest during July and least during November (median daily range in pH per month of 1.4 and 0.1 SU, respectively). During 2018, the daily maximum pH value was 8.5 SU or higher for 28 days, 22 of which were in July. The highest pH values were in the afternoons of July 16 and July 31, 2018 (9.1 and 9.0 SU, respectively). The lowest daily minimum pH values ranged from 6.3 to 6.5 SU and were observed during September and again during October. The pH values for Hocomonco Pond station 3 (Hocomonco Pond near Otis Street, Westborough, Mass. [station 421628071384501]) (fig. 9) ranged from 6.3 to 7.0 SU with a median of 6.6 SU.

The DOC concentrations for Hocomonco Pond (fig. 9) were among the lowest measured during the study. The monthly DOC concentrations at Hocomonco Pond stations 1 and 2 ranged from 2.05 to 3.52 mg/L with a median of 3.29 mg/L and from 1.92 to 5.17 mg/L with a median of 3.40 mg/L, respectively. The monthly DOC concentrations at Hocomonco Pond station 3 ranged from 2.03 to 4.28 mg/L with a median of 3.28 mg/L. The DOC concentrations had only a slight seasonal pattern, with the highest DOC concentrations (4 to 5.2 mg/L) measured in July through October 2018 and the lowest DOC concentrations (less than 2.5 mg/L) measured in February through April 2019.

Monthly total hardness concentrations at Hocomonco Pond stations 1 and 2 ranged from 60.9 to 89.4 mg/L as CaCO₃ with a median of 73.9 mg/L as CaCO₃ and from 62.4 to 93.6 mg/L as CaCO₃ with a median of 73.7 mg/L as CaCO₃, respectively. Monthly total hardness concentrations at Hocomonco Pond station 3 ranged from 62.2 to 91.2 mg/L as CaCO₃ with a median of 76.9 mg/L as CaCO₃. The total hardness concentrations did not indicate a seasonal pattern. The lowest values were measured during December (less than 63 mg/L as CaCO₃), and the highest values were measured during September (ranging from 89 to 94 mg/L as CaCO₃).

Aluminum concentrations at Hocomonco Pond stations 1 and 2 ranged from 3.00 to 36.3 µg/L with a median of 15.0 µg/L and from 7.10 to 122 µg/L with a median of 15.0 µg/L, respectively. Aluminum concentrations at Hocomonco Pond station 3 ranged from 3.10 to 41.5 µg/L with a median of 11.9 µg/L.

The pH values for the effluent station (Westborough water-treatment plant backwash effluent [station 421627071392401]) (fig. 9) ranged from 6.7 to 8.7 SU with a median of 7.4 SU. Effluent pH values were generally higher than pH values in the pond. Effluent DOC and total hardness concentrations were always lower than the DOC and total hardness concentrations in the pond. Effluent DOC concentrations ranged from 1.55 to 2.31 mg/L with a median of 1.82 mg/L. Effluent total hardness concentrations ranged from 35 to 63.3 mg/L as CaCO₃ with a median of 54.4 mg/L as CaCO₃. Aluminum concentrations in the effluent ranged from 133 to 1,540 µg/L with a median of 333 µg/L. Aluminum concentrations in the effluent were always higher than aluminum concentrations in Hocomonco Pond. These higher aluminum concentrations were not reflected at pond stations 1–3, likely because of dilution of effluent in Hocomonco Pond.

Hanover Water-Treatment Facility

Stations near the Hanover WTF (fig. 10) were visited monthly from May 3, 2018, through May 2, 2019, to collect discrete water-quality samples and measure water-quality field parameters. The pH values for the ambient (upstream) station (Third Herring Brook Pond Street near Hanover, Mass. [station 011058065]) ranged from 6.2 to 7.6 SU with a median of 6.4 SU. The pH values in Third Herring Brook varied throughout the year with no distinct seasonal pattern. The lowest monthly pH value (6.2 SU) was measured at the upstream station in May 2018. Continuous pH data were measured at the upstream station from May 29, 2018, through February 5, 2019. Low flows affected the collection of water-quality data in Third Herring Brook during summer of 2018. No continuous pH data were measured from July 29 through September 12, 2018, because Third Herring Brook was dry at the upstream station during this period. Diel variations in pH were small (median daily range in pH per month of about 0.1 SU) and were less variable than pH variations between months. For example, pH values in fall 2018 ranged from 6.0 SU in September to 7.1 SU in November. The pH values for the ambient (downstream) station (Third Herring Brook downstream water-treatment plant near Hanover, Mass. [station 011058075]) (fig. 10) ranged from 5.8 to 7.9 SU with a median of 6.2 SU.

The monthly DOC concentrations at the upstream station ranged from 4.12 to 11.7 mg/L with a median of 8.40 mg/L. The monthly DOC concentrations detected at the upstream station were lowest in February and highest during October 2018. The monthly DOC concentrations at the downstream station ranged from 4.79 to 16.9 mg/L with a median of 9.76 mg/L. The monthly DOC concentrations detected at the downstream station during July and October were greater than 16 mg/L and were among the highest measured during the study. The brook flows through a large wetland in this reach that dried to a series of shallow pools during these months.

Total hardness concentrations at the upstream station ranged from 36.4 to 85.6 mg/L as CaCO₃ with a median of 47.1 mg/L as CaCO₃. Total hardness concentrations at the downstream station ranged from 33.7 to 173 mg/L as CaCO₃ with a median of 54.8 mg/L as CaCO₃. The total hardness concentrations at the upstream station were highest (greater than 50 mg/L as CaCO₃) from May through October.

Aluminum concentrations at the upstream station ranged from 34.3 to 90.0 µg/L with a median of 65.5 µg/L. Aluminum concentrations at the downstream station ranged from 79 to 3,090 µg/L with a median of 225 µg/L. The highest aluminum concentrations at the downstream station were measured during July (1,980 and 1,290 µg/L) and August (3,090 µg/L), when streamflow in Third Herring Brook at the downstream station dried to a series of pools.

The pH values for the effluent station (Hanover water-treatment plant backwash effluent [station 420754070495801]) (fig. 10) ranged from 6.9 to 8.2 SU with a median of 7.5 SU. Effluent pH values were always higher than values in Third Herring Brook. Effluent DOC concentrations ranged from 2.06 to 6.19 mg/L with a median of 4.03 mg/L. Monthly DOC concentrations in effluent were always lower than DOC concentrations in the Third Herring Brook. The total hardness of the effluent ranged from 133 to 245 mg/L as CaCO₃ with a median of 156 mg/L as CaCO₃. Monthly total hardness values in effluent were always higher than total hardness values in Third Herring Brook. Aluminum concentrations in the effluent were higher than concentrations at the upstream station and ranged from 210 to 114,000 µg/L with a median of 1,140 µg/L. Effluent discharges likely contributed to high aluminum concentrations at the downstream station during the period from July 29 through September 12, 2018, when Third Herring Brook was dry at the upstream station and during low-flow periods leading up to and after the period when the brook was dry.

Leominster Water-Treatment Facility

Stations near the Leominster WTF (fig. 11) were visited monthly from April 23, 2018, through May 6, 2019, to collect discrete water-quality samples and measure water-quality field parameters. The pH values for the ambient (upstream) station (Monoosnoc Brook, upstream Leominster water-treatment plant [station 01094420]) ranged from 5.2 to 6.9 SU with a median of 5.9 SU. The monthly pH values for November and December 2018 (5.3 and 5.2 SU, respectively) and March 2019 (5.4 SU) were among the lowest pH values measured during the study. The pH at the upstream station had a distinct seasonal pattern, with higher values (greater than 6.5 SU) from June through August, and lower values (less than 6.0 SU) from October through February. Continuous pH data were measured at the ambient (upstream) station (Monoosnoc Brook, upstream Leominster water-treatment plant [station 01094420]) from May 4, 2018, through April 29, 2019. No data were collected from May 7 through June 7, 2018, because of a monitor malfunction. The continuous pH data indicate that Monoosnoc Brook had small diel variations in pH that were generally less than 0.4 SU. Diel variations in pH were greatest during June and least during November. Variations in pH between or within months were generally greater than diel variations of pH. The pH values for the ambient (downstream) station (Monoosnoc Brook, downstream Leominster water-treatment plant [station 01094422]) ranged from 5.4 to 7.2 SU with a median of 6.4 SU.

The monthly DOC concentrations at the upstream station ranged from 2.00 to 7.37 mg/L with a median of 4.21 mg/L. The monthly DOC concentrations at the downstream station ranged from 2.07 to 7.32 mg/L with a median of 4.03 mg/L. The monthly DOC concentrations in Monoosnoc Brook had a seasonal pattern and were lowest (less than 3 mg/L) during July and August and highest (greater than 5.0 mg/L) from October through December.

Total hardness concentrations at the upstream station ranged from 5.13 to 39.1 mg/L as CaCO₃ with a median of 7.05 mg/L as CaCO₃. Total hardness concentrations at the downstream station ranged from 5.19 to 35.2 mg/L as CaCO₃ with a median of 9.12 mg/L as CaCO₃. The total hardness concentrations in Monoosnoc Brook were among the lowest measured during the study. Total hardness had a seasonal pattern; values were higher (greater than 10.0 mg/L as CaCO₃) from June through August and lower (less than 10.0 mg/L as CaCO₃) during the remainder of the year.

Aluminum concentrations at the upstream station ranged from 26.7 to 208 µg/L with a median of 155 µg/L. Aluminum concentrations at the downstream station ranged from 86.1 to 287 µg/L with a median of 198 µg/L.

Monoosnoc Brook contained low ionic-strength water during portions of the study period. Specific conductance at the upstream station ranged between 107 and 426 µS/cm for the first half of the water year (April through September 2018). From October 2018 through April 2019, the specific conductance was less than 100 µS/cm. Seasonal differences in water quality at the Monoosnoc Brook stations are potentially related to the sources of water to the brook. Specific conductance was highest (about 425 µS/cm) during low flows in July and August when streamflow is primarily base flow (groundwater discharge). Specific conductance was lowest (less than 100 µS/cm) during periods of the year when streamflows were higher (October through April).

The pH values for the effluent station (Leominster water-treatment plant backwash effluent [station 423258071480701]) (fig. 11) ranged from 6.6 to 8.1 SU with a median of 7.5 SU. The effluent pH was always higher than the pH of Monoosnoc Brook at the upstream station, and the effluent discharge likely contributed to higher pH values for the downstream station. The pH at the downstream station averaged 0.3 SU higher than the pH at the upstream station. The effluent DOC concentrations ranged from 2.17 to 4.35 mg/L with a median of 3.15 mg/L. Effluent total hardness concentrations ranged from 5.07 to 10.4 mg/L as CaCO₃ with a median of 7.71 mg/L as CaCO₃. Effluent aluminum concentrations ranged from 326 to 792 µg/L with a median of 453 µg/L. Monthly aluminum concentrations in effluent were always higher than aluminum concentrations in Monoosnoc Brook. During periods of low flow, effluent discharges likely contributed to high aluminum concentrations at the downstream station.

Weymouth Water-Treatment Facility

Stations near the Weymouth WTF (fig. 12) were visited monthly from May 3, 2018, through May 2, 2019, to collect discrete water-quality samples and measure water-quality field parameters. The pH values for the ambient (upstream) station (Mill River near Randolph Street, South Weymouth [station 01094420]) ranged from 5.2 to 6.9 SU with a median of 5.9 SU. The monthly discrete pH values for Mill River near Randolph Street, South Weymouth, Mass. (station 01105587) ranged from 6.1 to 6.7 SU with a median of 6.3 SU. The pH values did not have a distinct seasonal pattern. Continuous pH data were measured at the upstream station from April 17, 2018, through February 4, 2019. Data were not collected between December 19, 2018, and January 17, 2019, because of a monitor malfunction. Daily minimum pH values were lowest during November (8 days with pH less than 6.0 SU), and daily maximum pH values were highest during May (16 days with pH greater than 6.8 SU). Diel variations in pH were greatest during August and least during December (median daily range of pH per month of 0.2 and 0.1 SU, respectively). Water-quality conditions at the shallow and deep stations on Great Pond were similar but differed from conditions in the Mill River. The pH values for Great Pond station 1 (Great Pond near outlet, shallow, South Weymouth, Mass. [station 421004070580201]) (fig. 12) ranged from 6.3 to 6.5 SU with a median of 6.4 SU. The pH values for Great Pond station 2 (Great Pond near outlet, deep, South Weymouth, Mass. [station 421004070580202]) ranged from 6.3 to 6.5 SU with a median of 6.4 SU.

The monthly DOC concentrations at the Mill River station ranged from 3.26 to 11.1 mg/L with a median of 5.29 mg/L. The monthly DOC concentrations at the Great Pond stations were lower than the DOC concentrations in Mill River. The monthly DOC concentrations at Great Pond stations 1 and 2 ranged from 2.35 to 3.53 mg/L with a median of 2.82 mg/L and from 2.32 to 3.09 mg/L with a median of 2.84 mg/L, respectively.

The total hardness concentrations at the Mill River station ranged from 61.5 to 111 mg/L as CaCO₃ with a median of 80.7 mg/L as CaCO₃. The total hardness concentrations had a slight seasonal pattern that was higher (greater than 90 mg/L as CaCO₃) from late May through September and lower (less than 90 mg/L as CaCO₃) the remainder of the year. The lowest total hardness concentration (61.5 mg/L as CaCO₃) was measured during December 2018. These seasonal differences in water quality are potentially related to the water quality of base flow in summer when streamflow in the Mill River is primarily groundwater discharge. The total hardness concentrations at the Great Pond stations 1 and 2 ranged from 17.1 to 28.1 mg/L as CaCO₃ with a median of 25.2 mg/L as CaCO₃ and from 20.2 to 28.1 mg/L as CaCO₃ with a median of 25.4 mg/L as CaCO₃, respectively.

Aluminum concentrations at the Mill River station ranged from 15.5 to 440 µg/L with a median of 63 µg/L. Aluminum concentrations in Great Pond were higher at the deeper station than at the shallow station. Aluminum concentrations at Great Pond stations 1 and 2 ranged from 27.5 to 71 µg/L with a median of 47.1 µg/L and from 30.1 to 192 µg/L with a median of 58.6 µg/L, respectively.

The pH values for the effluent station (Weymouth water-treatment plant effluent [station 420959070580401]) (fig. 12) ranged from 5.8 to 7.2 SU with a median of 6.25 SU. The effluent DOC concentrations ranged from 1.90 to 4.14 mg/L with a median of 3.06 mg/L. Effluent total hardness concentrations ranged from 16.0 to 29.3 mg/L as CaCO₃ with a median of 27.8 mg/L as CaCO₃. Effluent aluminum concentrations ranged from 194 to 12,700 µg/L with a median of 971 µg/L. In addition to the high aluminum values in the effluent discharge, there were some physical indications (clear water and a delta of gelatinous sediments of unusual composition at the effluent outfall) that coagulant from the plant may be reaching the pond.

Wilmington Water-Treatment Facility

Stations near the Wilmington WTF (fig. 13) were visited monthly from May 4, 2018, through May 6, 2019, to collect discrete water-quality samples and measure water-quality field parameters. The pH values for the two ambient (upstream) stations on Sawmill and Maple Meadow Brooks varied over a similar pH range. The pH values for the Sawmill Brook station (Sawmill Brook at Chestnut Street, Wilmington, Mass. [station 01101296]) ranged from 6.5 to 7.0 SU with a median of 6.8 SU. The pH values for the upstream Maple Meadow Brook station (Maple Meadow Brook, Wilmington, Mass., [station 01101294]) ranged from 6.6 to 7.1 SU with a median of 6.9 SU. The monthly pH values did not indicate a consistent seasonal pattern. Continuous pH data were measured at the Sawmill Brook station from June 6, 2018, through February 7, 2019. Diel variations in pH were greatest during August and September (median daily range in pH per month of 0.2 and 0.1 SU, respectively. The lowest daily minimum pH values were near 6.0 SU and were measured over 1- to 2-day periods during July, August, and September. The highest daily maximum pH values (7.0–7.2 SU) were measured during January. The pH values for the downstream Maple Meadow Brook station (Maple Meadow Brook at Middlesex Canal, Wilmington, Mass., [station 01101298]) ranged from 6.1 to 6.7 SU with a median of 6.2 SU.

Monthly DOC concentrations at the Sawmill Brook station (fig. 13) ranged from 3.24 to 6.66 mg/L with a median of 4.71 mg/L. The monthly DOC concentrations at the upstream Maple Meadow Brook station ranged from 2.33 to 4.50 mg/L with a median of 2.81 mg/L. The DOC concentrations were lowest (less than 3.00 mg/L) from May through August and from December through April. The monthly DOC concentrations at the downstream Maple Meadow Brook station ranged from 4.09 to 8.08 mg/L with a median of 6.83 mg/L. Monthly DOC concentrations at the downstream station were always higher than at the two upstream stations, potentially because of the effect of the large wetland between the upstream and downstream stations.

Total hardness concentrations at Sawmill Brook station ranged from 57.9 to 115 mg/L as CaCO₃ with a median of 79.8 mg/L as CaCO₃. The total hardness concentrations at the upstream Maple Meadow Brook station (fig. 13) ranged from 62.4 to 117 mg/L as CaCO₃ with a median of 87.2 mg/L as CaCO₃. The total hardness concentrations at the upstream stations had a seasonal pattern and were higher (greater than 90 mg/L as CaCO₃) from May through October and lower (less than 90 mg/L as CaCO₃) during the remainder of the year. The total hardness concentrations at the downstream Maple Meadow Brook station ranged from 37.9 to 97.0 mg/L as CaCO₃ with a median of 77.6 mg/L as CaCO₃.

The aluminum concentrations at the Sawmill Brook station ranged from 15.8 to 161 µg/L with a median of 46.4 µg/L. The aluminum concentrations at the upstream Maple Meadow Brook station ranged from 28.0 to 91.3 µg/L with a median of 45.0 µg/L. Aluminum concentrations at the two upstream stations on Sawmill Brook and Maple Meadow Brook were generally higher than aluminum concentrations at the downstream Maple Meadow Brook station. The aluminum concentrations at the downstream Maple Meadow Brook station ranged from 4.10 to 80.0 µg/L with a median of 19.0 µg/L.

The pH values for the effluent station (Wilmington water-treatment plant backwash effluent [station 423200071100201]) (fig. 13) ranged from 6.9 to 7.8 SU with a median of 7.0 SU. Effluent DOC concentrations ranged from 2.71 to 7.50 mg/L with a median of 4.93 mg/L. Effluent total hardness ranged from 99.7 to 131 mg/L as CaCO₃ with a median of 117 mg/L. Effluent aluminum concentrations ranged from 320 to 119,000 µg/L with a median of 1,540 µg/L. In November and December 2018, the southern lagoon was cleaned, and there was little effluent outflow while the lagoon was refilling. Monthly effluent samples collected while the lagoons were refilling were collected from streams of effluent flowing through cracks between boards on the weir and had high aluminum concentrations (119,000, 9,930, and 26,900 µg/L). The high aluminum concentrations likely resulted because the effluent leaks were not a supernatant discharge and included suspended solids. The high aluminum concentrations in the effluent were not reflected at the downstream station possibly because of the slow movement of water and long flow paths through the wetland.

Westborough Wastewater-Treatment Facility

Monthly aluminum criteria value calculations for the Assabet River in Westborough were generated using water-quality data collected at the ambient (upstream) station Assabet River near Westborough, Mass. (station 01096603) (fig. 9). The monthly aluminum CMC values for the upstream station ranged from 1,100 to 1,800 µg/L, and the monthly aluminum CCC values ranged from 430 to 610 µg/L. The CMC minimum, 5th percentile, and 10th percentile values were 1,100, 1,100, and 1,100 µg/L, respectively, and the CCC minimum, 5th percentile, and 10th percentile values were 430, 442, and 450 µg/L, respectively. The aluminum concentrations measured in the Assabet River at the downstream station, Assabet River downstream from Westborough water-treatment plant at Route 9 (station 421700071381901), ranged from 28.6 to 166 µg/L and were always lower than the minimum, 5th percentile, and 10th percentile of the monthly aluminum CMC and CCC values.

Marlborough Wastewater-Treatment Facility

Monthly aluminum criteria value calculations for the Assabet River in Marlborough were generated using water-quality data collected at the ambient (upstream) station Assabet River at Boundary Street near Northborough, Mass. (station 01096720) (fig. 4). The monthly CMC values for the upstream station ranged from 1,300 to 2,800 µg/L, and the monthly CCC values ranged from 460 to 830 µg/L. Monthly aluminum CMC and CCC values were highest during May through October and lowest during November and February. The CMC minimum, 5th percentile, and 10th percentile values were 1,300, 1,300, and 1,380 µg/L, respectively. The CCC minimum, 5th percentile, and 10th percentile values were 460, 466, and 486 µg/L, respectively. The aluminum concentrations measured in the Assabet River at the downstream station, Assabet River, downstream Marlborough wastewater-treatment plant (station 01096725), ranged from 61.5 to 174 µg/L and were always lower than the minimum, 5th percentile, and 10th percentile of the monthly aluminum CMC and CCC values.

Hudson Wastewater-Treatment Facility

Monthly aluminum criteria value calculations for the Assabet River in Hudson were generated using water-quality data collected at the ambient (upstream) station Assabet River at Cox Street near Hudson, Mass. (station 01096870) (fig. 5). The monthly aluminum CMC values for the upstream station ranged from 1,300 to 2,500 µg/L, and the monthly aluminum CCC values ranged from 470 to 710 µg/L. The aluminum calculations indicated seasonal patterns. Monthly CMC and CCC values were highest during June through October and lowest in December. The CMC minimum, 5th percentile, and 10th percentile values were 1,300, 1,360, and 1,420 µg/L, respectively, and the CCC minimum, 5th percentile, and 10th percentile values were 470, 476, and 486 µg/L, respectively. The aluminum concentrations measured in the Assabet River at the downstream station, Assabet River near Hudson-Stow town line (station 01096875), ranged from 18.1 to 77.4 µg/L and were always lower than the minimum, 5th percentile, and 10th percentile of the monthly aluminum CMC and CCC values.

Maynard Wastewater-Treatment Facility

Monthly aluminum criteria value calculations for the Assabet River in Maynard were generated using water-quality data collected at the ambient (upstream) station Assabet River, upstream Maynard wastewater-treatment plant (station 01097021) (fig. 6). The monthly aluminum CMC values for the upstream station ranged from 1,000 to 3,600 µg/L, and the monthly aluminum CCC values ranged from 420 to 1,700 µg/L. The aluminum CMC and CCC calculations indicated seasonal patterns. Monthly CMC and CCC values were highest during June through September and lowest in November and February. The CMC minimum, 5th percentile, and 10th percentile values were 1,000, 1,060, and 1,140 µg/L, respectively, and the CCC minimum, 5th percentile, and 10th percentile values were 420, 420, and 428 µg/L, respectively. The aluminum concentrations measured in the Assabet River at the downstream station, Assabet River, downstream Maynard wastewater-treatment plant (station 01097023), ranged from 20.2 to 79.8 µg/L and were always lower than the minimum, 5th percentile, and 10th percentile of the monthly aluminum CMC and CCC values.

Cohasset Water-Treatment Facility

Monthly aluminum criteria value calculations for Lily Pond in Cohasset were generated using water-quality data collected at Lily Pond station 1 (Lily Pond deep hole [shallow] [station 421326070485802]) and Lily Pond station 2 (Lily Pond deep hole [station 421326070485801]) (fig. 7). When values from Lily Pond stations 1 and 2 are combined, the monthly CMC values ranged from 30 to 3,300 µg/L, and the monthly CCC values ranged from 19 to 1,100 µg/L. The aluminum CMC and CCC calculations at Lily Pond stations 1 and 2 varied but did not indicate distinct seasonal patterns. Monthly CMC and CCC values were lowest during November. The CMC minimum, 5th percentile, and 10th percentile values determined from the combined monthly aluminum criteria values were 30, 32, and 102 µg/L, respectively, and the CCC minimum, 5th percentile, and 10th percentile values determined from the combined monthly aluminum criteria values were 19, 20, and 66 µg/L, respectively. The low aluminum criteria values for these stations were affected primarily by the low pH values measured in Lily Pond. The aluminum concentrations at Lily Pond stations 1 and 2 ranged from 43.3 to 370 µg/L, and from 27.3 to 357 µg/L, respectively. The aluminum concentrations measured at pond stations 1 and 2 exceeded the minimum, 5th percentile, and 10th percentile of the monthly aluminum CMC and CCC values for 8 of the 10 months for which CMC and CCC values were calculated.

Fitchburg Water-Treatment Facility

Monthly aluminum criteria value calculations for Wyman Pond in Fitchburg were generated using water-quality data collected at four stations on Wyman Pond in Fitchburg, pond stations 1 and 2 (stations 423132071523401 and 423132071523402, respectively), and pond stations 3 and 4 (stations 423211071524701 and 423211071524702, respectively) (fig. 8). The monthly CMC values for pond stations 1 and 2 ranged from 49 to 1,632 µg/L, and the monthly CCC values ranged from 31 to 450 µg/L. The monthly CMC values for pond stations 3 and 4 ranged from 95 to 390 µg/L, and the monthly CCC values ranged from 59 to 220 µg/L. The monthly aluminum criteria value calculations for the stations on Wyman Pond varied but did not indicate seasonal patterns. The low aluminum criteria values for Wyman Pond are a result of concurrently low pH and DOC concentrations for sampling events. The monthly CMC and CCC values were lowest at pond station 1 during November, pond station 2 during September, pond station 3 during May and November, and pond station 4 during September.

When monthly CMC and CCC values from all four Wyman Pond stations (pond stations 1–4) are combined, the minimum, 5th percentile, and 10th percentile of the monthly aluminum CMC values were 49, 89, and 95 µg/L, respectively, and the minimum, 5th percentile, and 10th percentile of the monthly aluminum CCC values were 31, 55, and 60 µg/L, respectively. The aluminum concentrations at pond stations 1 and 2 ranged from 23.6 to 67.4 µg/L, and the aluminum concentrations at pond stations 3 and 4 ranged from 15.1 to 62 µg/L. The highest monthly aluminum concentrations detected in Wyman Pond exceeded the minimum of the monthly aluminum CMC values (49 µg/L) at all four stations. There were no pond stations where aluminum concentrations exceeded the 5th percentile (89 µg/L) and 10th percentile (95 µg/L) of the monthly aluminum CMC values. The monthly aluminum concentrations measured in Wyman pond during September (67.4 µg/L) or November (67 µg/L) exceeded the minimum, 5th percentile, and 10th percentile of the monthly aluminum CCC values for all four stations.

Westborough Water-Treatment Facility

Monthly aluminum criteria value calculations for Hocomonco Pond in Westborough were generated using water-quality data collected at three stations on Hocomonco Pond, pond stations 1 and 2 (stations 421622071385701 and 421622071385702, respectively), and pond station 3 (station 421628071384501) (fig. 9). When values from pond stations 1, 2, and 3 are combined, the monthly CMC aluminum criteria values ranged from 620 to 2,300 µg/L, and the monthly CCC values ranged from 290 to 750 µg/L. The aluminum concentration calculations at the pond stations varied with the lowest values in fall and winter. The CMC minimum, 5th percentile, and 10th percentile values when all three pond stations are combined were 620, 706, and 769 µg/L, respectively. The combined CCC minimum, 5th percentile, and 10th percentile values were 290, 336, and 341 µg/L, respectively. The aluminum concentrations detected in Hocomonco Pond ranged from 3 to 122 µg/L and were always lower than the minimum, 5th percentile, and 10th percentile of the monthly aluminum CMC and CCC values for pond stations 1–3.

Hanover Water-Treatment Facility

Monthly aluminum criteria value calculations for Third Herring Brook in Hanover were generated using water-quality data collected at the ambient (upstream) station Third Herring Brook Pond Street near Hanover, Mass. (station 011058065) (fig. 10). The monthly CMC values for the upstream station ranged from 900 to 2,900 µg/L, and the monthly CCC values ranged from 380 to 880 µg/L. The aluminum CMC and CCC values indicated seasonal patterns. The monthly CMC and CCC values were low for January through March with the lowest values during March (CMC and CCC values of 900 and 380 µg/L, respectively). The CMC minimum, 5th percentile, and 10th percentile values were 900, 900, and 909 µg/L, respectively, and the CCC minimum, 5th percentile, and 10th percentile values were 380, 386, and 397 µg/L, respectively. The aluminum concentrations measured in Third Herring Brook at the downstream station, Third Herring Brook, downstream water-treatment plant, near Hanover, Mass. (station 011058075), ranged from 79 to 3,090 µg/L, and the monthly aluminum concentrations measured during July (1,980 and 1,290 µg/L) and August (3,090 µg/L) exceeded the minimum, 5th percentile, and 10th percentile of the monthly aluminum CMC and CCC values.

Leominster Water-Treatment Facility

Monthly aluminum criteria value calculations for Monoosnoc Brook in Leominster were generated using water-quality data collected at the ambient (upstream) station, Monoosnoc Brook, upstream Leominster water treatment plant (station 01094420) (fig. 11). The monthly CMC values for the upstream station ranged from 15 to 880 µg/L, and the monthly CCC values ranged from 9 to 360 µg/L. The low aluminum CMC and CCC values were caused by low pH values measured in Monoosnoc Brook in combination with low DOC and total hardness concentrations. The monthly CMC and CCC values for the upstream station were lowest in December (15 and 9 µg/L, respectively). The CMC minimum, 5th percentile, and 10th percentile values were 15, 19, and 21 µg/L, respectively, and the CCC minimum, 5th percentile, and 10th percentile values were 9, 11, and 13 µg/L, respectively. The aluminum concentrations measured in Monoosnoc Brook at the downstream station Monoosnoc Brook, downstream Leominster water-treatment plant (station 01094422) ranged from 86.1 to 287 µg/L and were always higher than the minimum, 5th percentile, and 10th percentile of the monthly aluminum CMC and CCC values.

Weymouth Water-Treatment Facility

Monthly aluminum criteria value calculations for the Mill River in Weymouth were generated using water-quality data collected at the ambient (upstream) station Mill River near Randolph Street, South Weymouth, Mass. (station 01105587) (fig. 12). The monthly CMC values for the upstream station ranged from 840 to 1700 µg/L, and the monthly CCC values ranged from 360 to 770 µg/L. Monthly CMC and CCC values for the Mill River station varied throughout the year and were lowest during February. The CMC minimum, 5th percentile, and 10th percentile values were 840, 870, and 896 µg/L, respectively. The CCC minimum, 5th percentile, and 10th percentile values were 360, 372, and 382 µg/L, respectively. The aluminum concentrations measured at the upstream station on Mill River ranged from 15.5 to 440 µg/L and were always lower than the minimum, 5th percentile, and 10th percentile of the monthly aluminum CMC values. The aluminum concentration measured at the upstream station on Mill River in February 2019 (440 µg/L) exceeded the minimum, 5th percentile, and 10th percentile of the monthly aluminum CCC values. During the selection process for stations associated with the Weymouth WTF, a pair of stations (stations 421004070580201 and 421004070580202) were selected in the pond downgradient of the effluent discharge from the Weymouth WTF, but a monitoring station was not selected on Mill River (the receiving-water body) downstream from where the effluent discharge reaches the Mill River. Aluminum concentrations in the effluent could reach high values (from 194 to 12,700 µg/L). However, aluminum concentrations measured in the embayment of Great Pond that receives the effluent discharge were always lower than the CMC and CCC minimum, 5th percentile, and 10th percentile values. Aluminum concentrations measured at the shallow pond station, Great Pond near outlet, shallow, South Weymouth, Mass. (station 421004070580201), ranged from 27.5 to 71 µg/L; and aluminum concentrations measured at the deep pond station, Great Pond near outlet, deep, South Weymouth, Mass. (station 421004070580202), ranged from 30.2 to 192 µg/L.

Wilmington Water-Treatment Facility

Monthly aluminum criteria value calculations for Maple Meadow Brook in Wilmington were generated using water-quality data collected at two ambient (upstream) stations, Sawmill Brook at Chestnut Street, Wilmington, Mass. (station 01101296), and Maple Meadow Brook, Wilmington, Mass. (station 01101294) (fig. 13). Monthly CMC values for the upstream station on Sawmill Brook ranged from 1,300 to 1,900 µg/L, and the monthly CCC values ranged from 460 to 610 µg/L. The monthly CMC values for the upstream station on Maple Meadow Brook ranged from 930 to 1,700 µg/L, and the monthly CCC values ranged from 360 to 540 µg/L. Monthly CMC and CCC values were lowest during August and December at the Sawmill Brook station and during December at the upstream Maple Meadow Brook station. The CMC minimum, 5th percentile, and 10th percentile values for the combined aluminum CMC values values for Sawmill and Maple Meadow Brooks were 930, 1,120, and 1,250 µg/L, respectively, and the CCC minimum, 5th percentile, and 10th percentile values were 360, 435, and 455 µg/L, respectively. Some high aluminum concentrations (from 320 to 119,000 µg/L) were detected in effluent leaking from the boards of the weir when the lagoon was filling. However, the aluminum concentrations measured in Maple Meadow Brook at the downstream station Maple Meadow Brook, at Middlesex Canal, Wilmington, Mass. (station 01101298), ranged from 4.1 to 80 µg/L and were always lower than the minimum, 5th percentile, and 10th percentile of the monthly aluminum CMC and CCC values.

Assessment of Variations in pH Measurement on Aluminum Criterion Maximum Concentration and Criterion Continuous Concentration Calculations

Discrete in situ pH measurements were compared to concurrent continuous (15-minute interval) pH measurements (fig. 25). Discrete pH measurements were less than the concurrent continuous pH measurements recorded by water-quality monitors for six of the stations, including Leominster, Fitchburg, Hanover, and, to a lesser extent, Marlborough, Maynard, and Wilmington. The 91 discrete measurements made at the 11 stations where continuous data were collected were on average 0.2 SU less than concurrent continuous measurements (differences ranged from −1.1 to 1.0 SU). A review of the discrete and continuous pH data did not definitively determine the reasons for specific differences. The following are possible sources of these differences: (1) differences in measurement location, (2) differences among measurements in rapidly moving and relatively still water, (3) inaccuracies resulting from measurement in low ionic-strength water using electrodes designed for general use, (4) use of different instruments to measure pH, and (5) errors from temperature differences between calibration buffers and field measurements. Differences in discrete and continuous pH measurements are discussed below using the ambient (upstream) Leominster station (fig. 11) as an example.

Assessment of the Effects of Diel Variations in pH on Aluminum Criterion Calculations

The daily cycles of photosynthesis and respiration create variations in carbon dioxide concentration throughout the day and night that generate corresponding diel variations in pH. The lowest pH values typically result in the morning in the hours shortly after sunrise, and highest pH values typically result in the middle of the afternoon. In general, diel variations in pH are smaller in shaded forested streams and rivers and higher in low-gradient and impounded rivers or ponds that have an open canopy with appreciable algae and aquatic vegetation. The daily range in pH values is greater during the growing season (April or May through September or October) than in the nongrowing season (October or November through March or April). Time-series plots of pH values for the stations considered in this study with continuous (15-minute interval) pH data are available in Armstrong and others (2022).

The timing of sample collection can affect the pH values detected during sampling and, consequently, the aluminum CMC and CCC values generated by the EPA Calculator for a given sampling event. Water-quality samples for this study were generally collected at stations associated with two facilities per day. Early morning hours were used to calibrate instruments, rinse sample bottles, load vehicles, and travel to a site. Typically, the first set of water-quality samples was collected in the middle to late morning and the second set of samples was collected in the early afternoon. For the morning samples, the median sample-collection time was 11 a.m. with 90 percent of the samples collected between 10 a.m. and 11:45 a.m. For the afternoon samples, the median sample-collection time was 1 p.m. with 90 percent of the samples collected between 12 p.m. and 3 p.m.

To assess how diel variations in pH could potentially affect aluminum CMC and CCC calculations, the minimum and maximum pH values from the continuous dataset were determined for each day that the water body was sampled. These minimum and maximum pH values were substituted into the EPA Calculator along with the DOC and total hardness concentrations collected during the discrete sampling. The resulting aluminum CMC and CCC values were then compared to the aluminum calculations determined by use of the continuous pH measurement that corresponded in time to the discrete pH sample. During the growing season, when diel variations in pH are largest, application of the minimum and maximum daily pH values to the EPA Calculator was determined to cause differences in monthly CMC values that ranged by several hundred µg/L for most stations. Differences of several thousand µg/L were determined in two instances, where differences between minimum and maximum diel pH values varied by about 1 to to 2 SU. Minimum and maximum pH values of 6.5 to 7.5 SU in Wyman Pond in Fitchburg (station 423132071523401) during August 2018 led to differences in calculated CMC and CCC values of 1,210 and 490 µg/L, respectively; and minimum and maximum pH values of 6.6 to 8.8 SU in Lily Pond in Cohasset (station 421326070485802) during October 2018 led to differences in calculated CMC and CCC values of 3,200 and 2,100 µg/L, respectively. Diel pH variations vary throughout a year, however. For many stations, the lowest CMC and CCC instantaneous aluminum criteria values resulted during the nongrowing season, when diel variations in pH were much smaller, thus minimizing the effect that diel pH variation may have on the minimum and calculated 5th and 10th percentile statistical CMC and CCC aluminum values. For most stations considered in this study, continuous pH data were collected during an 8- to 9-month period from April or May through December. Over this period, diel variations in pH at the 11 stations were determined to correspond to differences in the 10th percentile of CMC values by a median of 160 µg/L, ranging from 0 to 610 µg/L, and differences in the 10th percentile of CCC values by a median of 40 µg/L, ranging from 15 to 210 µg/L. A dataset with at least a full year of record or longer would improve the evaluation and comparison of the potential effect of diel pH variations and sampling time on aluminum CMC and CCC calculations.