top of page
  • Kelly Conway

Stormwater Management and CSO Reduction Projects in the Merrimack River Watershed

Salmon, shad and Alewives were formerly abundant here . . . until the dam, . . . and the factories at Lowell, put an end to their migrations hitherward. . . . Perchance, after a few thousands of years, if the fishes will be patient, and pass their summers elsewhere . . . nature will have levelled. . . the Lowell factories, and the Grass-ground River [will] run clear again. – Henry David Thoreau, A Week on the Concord and Merrimack Rivers, 1849.

The Merrimack River has a history of poor water quality. Since the Industrial Revolution brought textile factories and mills to Lowell, MA, the water quality was degraded by industrial and human waste discharged to the water as population growth in these mill cities boomed. The environmental legislation of the 1970s and 1980s greatly improved the water quality of the Merrimack River, but the river still often fails to meet water quality standards. Today, combined sewer overflows (CSOs) make up the main problem afflicting the Merrimack River Watershed, leading to increased levels of solid waste, nutrients, and pathogens in the water following rain events. As climate change continues to progress, the frequency and intensity of abnormal precipitation events will continue to increase. It is therefore very important that we begin work on preparing for increased volumes of stormwater entering the water systems in the coming decades in order to protect against future severe CSO events.

Combined sewer overflows occur when excessive rainfall or storm events overwhelm the combined sewer system (CSS), a sewer system in which municipal sewage and stormwater runoff are both collected in the same pipe, and causes the system to discharge untreated water containing bacteria and other nutrients and contaminants directly into the river.

In the Merrimack River Watershed, CSOs mainly stem from the cities of Lowell, Haverhill, and the Greater Lawrence Sanitary District in Massachusetts, and Nashua and Manchester in New Hampshire. The table below contains data on CSOs published in the Merrimack River Watershed Assessment study in January 2004 for the US Army Corps of Engineers by the Merrimack River Basin Community Coalition.

There are several methods of CSO prevention that I will explore in this paper. One way in which communities are tackling the issue of combined sewage overflow is by separating the sewer system into individual streams for wastewater treatment and for stormwater discharge into bodies of water. Another method is to increase capacity for storage and treatment in existing sewer systems. A third method of combating CSOs is through the implementation of green infrastructure, such as bioswales, rain gardens, and permeable pavements, which infiltrate and filter stormwater runoff to mimic natural forest habitats.

Analysis of 2013 Water Quality Data

In order to determine what is at stake when polluted runoff and CSOs enter the river, I will discuss the average water quality data that my team with the Gulf of Maine Institute (GOMI) found from monitoring a portion of the Merrimack River throughout the summer of 2013. The dates for monitoring were not selected due to their proximity to a rainfall event, so this average was intended to be representative of the average quality of the river in the summer. I also use this data to determine what may be the underlying concerns associated with water quality in average conditions, not just in the extreme rainfall events, in the hopes of bringing the river ultimate protection from contaminants and pollution.

Figure 1 below shows an aerial view of the Merrimack River as it goes through two points in Newburyport where the river’s water quality was monitored and measured by the Gulf of Maine Institute.

Figure 1: Google Earth aerial image of the Merrimack River including the Haverhill Sewage Treatment Plant, the Artichoke River mouth, the Yankee Marina monitoring station, and the Amesbury Wastewater Treatment Plant.

Some water quality data for these points is tabulated below for the summer of 2013 at the two points shown on the map, the mouth of the Artichoke River in the Merrimack River and the point on the Merrimack right off of the Yankee Marina docks. It is important to note that the Artichoke River sampling location is upstream of the Amesbury wastewater treatment plant, while the Yankee Marina sampling point is downstream of the plant. However, it is important to note that the sampling location at the mouth of the Artichoke is surrounded by much more undeveloped forestland than at the Yankee Marina, which is surrounded by residential or commercial lands characterized by mostly impervious surfaces. Impervious surfaces lead to increased polluted runoff, while forestland retains and filters runoff, discharging cleaner water into the river. The data we collected for Table # below shows that the water by the Yankee Marina docks had significantly higher phosphate levels, slightly higher nitrate levels, significantly higher specific conductivity, and considerably higher turbidity than the water at the mouth of the Artichoke River. Phosphates often enter water by runoff contaminated by human or animal waste, agricultural fertilizer, and erosion of phosphate- rich landscapes and bedrock. High levels of phosphates are also often associated with CSO events, as sewer systems cannot handle the volume of water and thus directly discharge water with human and animal waste (phosphate-rich) into the river. The EPA provides a limit of 1 mg/L for phosphates in drinking water, and both of these locations demonstrated a summer average above the limit.

High nitrate levels also often come from human/animal waste and fertilizers as well as from CSO events. The difference in nitrate levels at the two points was not nearly as significant as that of the phosphate levels, but it is still important to recognize for analysis. Both sites had nitrate levels far below the EPA drinking water limit of 10 mg/L. The specific conductivity is also much higher at the Yankee Marina. Specific conductivity tells us about the concentration of ions in the water. In this case, it is likely that the specific conductivity at the Yankee Marina is raised by its proximity to the mouth to the Atlantic Ocean (at the far right in the aerial photo); saltwater raises the salinity and salt concentration in the water, which raises the specific conductivity. However, the high specific conductivity could also be indicative of ions in solution, including chloride, nitrate, or phosphate ions, or of the presence of runoff from clay soils with minerals that ionize in water (Fondriest Environmental). Additionally, when we took this data, we noted that the YSI equipment we used was not functioning correctly at times during the sampling, so it is important to be cautious in using this data in my analysis. Lastly, the high turbidity at the Yankee Marina suggests the presence of clay and soil particles and total suspended solids (TSS) in the water. While turbidity is not usually a health concern, it can become a health concern if the suspended solids cause metal ions to agglomerate in the water (USGS).

Included in this study performed by the GOMI team was a bacterial analysis of three beaches along the Merrimack River in Newburyport, one of which (called Back River) is in the middle of the river and surrounded by salt marshes, and the other two (Salisbury Beach and Joppa Flats) which are downstream of the main industrial and commercial area of town, and specifically downstream of the Newburyport sewage treatment plant. Our study found that the Back River beach consistently had levels of Enterococcus bacteria well below the Massachusetts state limit of 104 cfu/100 mL, while the other two beaches only had bacteria levels below this limit 40% of the time. Our study also notes that generally speaking, the high levels of bacteria at these two beaches tended to happen after rain events, although this is only anecdotal as no real data was collected on the rain events. This data therefore suggests that the bacterial levels at Salisbury Beach and Joppa Flats are higher than that of Back River due to their position downstream of the treatment plant as well as their relative lack of salt marsh to filter the contaminants.

Figure 2: Map of Newburyport and Salisbury, MA, depicting the three beaches and the Newburyport sewage treatment plant.

After analyzing these results, I suggest that in addition to CSOs, other major contributors to pollution and poor water quality in the river are runoff from impervious surfaces and discharges from sewage treatment plants in everyday function. Because this data was not collected near the date or location of a CSO event, the data suggests that typical operation of the treatment plants and typical conditions can cause significantly worse water quality even without a CSO event.

Assessment of Abatement Strategies for the Merrimack River Watershed

CSO reduction projects using one or more of the three abatement strategies have taken place all over the nation as aging infrastructure fails during heavy rainfall and storm events. However, many of the projects that have been successful cover much smaller areas. The scope of the project of the Merrimack River Watershed is significantly larger than any CSO reduction project I came across in my research. After doing research on other CSO abatement projects that used separation (Alexandria, VA, Cambridge, MA, and Nashua, NH), increased storage and treatment (Alexandria, VA), and/or green infrastructure (Seattle, WA) and performing calculations specific to the Merrimack River watershed, my primary recommendation is to combine the three methods, with an emphasis on spending on storage capacity Green infrastructure should be widely implemented, but should not be relied on as the primary stormwater management practice, and should not compose much of the total project budget. Sewer separation should take place on a case-by-case basis as aging infrastructure demonstrates need. I predict this combination plan to result in over 90% reduction of CSO volume discharge per year.

Fig. 3 Graph of cost estimate and CSO reduction comparison for the 3 strategies studied.


Buote, Brenda J. “Waste-water project to begin.” Boston Globe. 27 March 2014. Web. Cambridge DPW. “Sewer Separation.” The Cambridge Department of Public Works. 2017. Web.

City of Manchester. “Combined Sewer Overflow – Stormwater.” City of Manchester Department of Environmental Protection, 2017. Web.

Fondriest Environmental, Inc. “Conductivity, Salinity & Total Dissolved Solids.” Fondriest Environmental, Inc. 2017. Web.

Greeley & Hansen. “Alternatives Evaluation: Sewer Separation.” City of Alexandria, VA, Department of Transportation and Environmental Services. October 2015. Web.

Greeley & Hansen. “CSS LTCPU Report.” City of Alexandria and Virginia Department of Environmental Quality. December 2016. Web.

Merrimack River Basin Community Coalition and CDM. “Merrimack River Watershed Assessment Study.” New England District, U.S. Army Corps of Engineers. January 2004. Web.

MWRC. “Improve Water Quality and Quantity.” Merrimack River Watershed Council. 2017. Web. Nashua DPW. “Wastewater Capital Projects.” City of Nashua Division of Public Works. 2 April 2014.


Northern Middlesex Clean Waters. “Water Quality.” 2017. Web.

Normandeau Associates, Inc. “Historical Water Quality and Selected Biological Conditions of the Upper Merrimack River, New Hampshire.” December 2011. Web.

Odefey, Jeffrey, et al. “Banking on Green: A Look at How Green Infrastructure Can Save Municipalities Money and Provide Economic Benefits Community-wide.” American Society of Landscape Architects, April 2012. Web.

Sullivan, Jim. “Merrimack River makes Top 10 list of nation’s most endangered waterways.” Eagle- Tribune. 17 April 2016. Web.

Tetra Tech, Inc. “Combined Sewer Overflow Program: 2010 CSO Reduction Plan Amendment.” Seattle Public Utilities. May 2010. Web.

“The Merrimack River.” National Park Service, U.S. Department of the Interior. 2017. Web.

U.S. Geological Survey (USGS). “Turbidity.” U.S. Department of the Interior. 2 December 2016. Web.


Kelly is an undergraduate senior at Columbia University studying earth and environmental engineering. During her time at Newburyport High School, she was the Newburyport GOMI team lead for the monitoring of marine invasive species and an active participant in water quality testing projects, pepperweed pulls, and community and youth outreach events. She was the primary author of the 2012 GOMI Marine Invasive Species report and a major contributor to the 2012 GOMI Water Quality report. During her senior year of high school, Kelly was selected to be a Henry David Thoreau Scholar for her demonstration of environmental dedication and leadership. Her experience with environmental conservation and stewardship as a member of GOMI led her to pursue environmental engineering studies in college, specifically in the field of the water-energy nexus: the intersection of water resources, energy resources, and sustainability principles.

At Columbia, Kelly joined a laboratory group conducting research on technologies for the water-energy-environment nexus. Her research has focused on developing desalination technologies, including the study of novel nanofiltration membranes for reverse osmosis desalination. Kelly also spent a summer in Newburyport working as an intern in GZA GeoEnvironmental, Inc.’s office on a variety of environmental engineering projects around the North Shore. In November 2016, she was inducted to the New York Alpha chapter of Tau Beta Pi, a national engineering honor society. She plans to pursue further graduate study in the field of water treatment processes following her graduation from Columbia in May 2018. In her free time, Kelly enjoys distance running, hiking, playing with her intramural basketball team, and exploring new restaurants and attractions in New York City.

Kelly’s article is a edited down version of a longer term paper she did for her Solid Hazardous Waste Management class. Her study was conducted on the Merrimack River Watershed, Massachusetts.

206 views0 comments
bottom of page