The Rising Tide of Brackish Water Desalination
For more than 50 years, Brockton, Mass., has struggled to produce enough water from local wells and reservoirs to meet the needs of its residents and businesses. In the spring of 2008, GE’s advanced membrane technology will help ease the community’s water worries when a new water treatment plant will start up and provide the community with a sustainable new water source produced by desalinating brackish water from the Taunton River.
Madrid-based Inima (a division of OHL) is building a $55-million desalination plant and pipeline on the tidally influenced Taunton River that is expected to provide Brockton with up to 5 million gal per day (mgd) of potable water that will augment existing raw water sources. The plant will use hollow-fiber ultrafiltration (UF) membranes and reverse osmosis (RO) to remove pollutants and salt from water—and may help lift a decades-old ban on new construction.
The heart of this innovative drinking water plant will be ZeeWeed immersed, low-pressure UF membranes made by ZENON Membrane Solutions, a part of GE Water & Process Technologies. The versatile UF membranes will serve a dual purpose in the plant—performing both pretreatment and primary filtration according to the prevailing raw water conditions. During winter and spring, when total dissolved solids (TDS) levels average less than national drinking water standards, the UF membranes will serve as the plant’s primary treatment process, according to Alfredo Andres, general manager of Inima.
When TDS levels rise in the summer, the UF system will consistently provide high-quality feedwater for the RO system, enabling it to function with minimal fouling, leading to lower cleaning requirements, higher flux rates and extended RO membrane life. ZeeWeed membrane fibers effectively remove virtually all suspended particles from the RO feedwater, regardless of influent turbidity or variable pollutant levels. This is particularly important for brackish water-and saltwater-fed facilities, where turbidity can vary greatly during seasonal changes and storm events.
Campaigning for membranes
When completed in 2008, this plant will be nearly 20 years in the making. It was originally suggested in 1991 by Jeff Hanson, a power plant engineer for a municipal electric utility. According to Hanson, the utility was looking to diversify, and the initial idea was to combine desalination and electric power generation; however, few New Englanders considered it a practical option. The governor had set up a task force to establish a feasible water plan for the state, and Hanson proposed the desalination plant at one of their meetings.
“They laughed,” Hanson said. “No one took the idea seriously—then.”
But, Hanson persevered, convinced that desalination could help solve southeast Massachusetts’ long-standing water problems. He’d researched desalination thoroughly, he said, and “there was no technical reason it wouldn’t work.”
When electric utility deregulation hit in the mid-1990s, the local utility was privatized and no longer interested in producing both power and potable water. So, Hanson joined Bluestone Energy Services, based in Braintree, Mass., as a partner and vice president to continue pursuing desalination. Bluestone provided the corporate platform he needed to promote the concept in the region. He met with local and state officials individually and collectively, explaining the process and benefits of treating saltwater in order to persuade them to support the project. In 1995, Hanson met with the Brockton City Council and eventually Jack Yunits, who had just been elected mayor of Brockton. (Yunits is now a mediator at Brockton-based Commonwealth Mediation.)
“Brockton had been struggling with water supply problems since 1952,” Yunits said. “The town had a moratorium on new construction because of water shortages. But, desalination seemed like a long shot at the time.”
Brockton officials evaluated other options, such as digging more wells, building another reservoir, or buying water from the Boston-based Massachusetts Water Resources Authority (MWRA), but all were impractical or too expensive. Yunits reconsidered desalination.
“I did the research,” he said, “and found that the math worked, the environmental benefits made sense, and the technology was improving.”
The option became even more attractive when Inima joined the project in 2001. The company has been involved in desalination projects since 1968 and currently owns and operates desalination plants in Africa, Europe, North America and South America. Inima wanted to enter the U.S. market and thought the proposed Taunton River project had potential, Andres said.
“They offered to assume all of the financial risk,” Yunits noted. “We’d already spent $5 million trying to solve our water problems, so the offer was very attractive.”
The challenge of being first
Getting political support was one challenge; securing permits for Massachusetts’ first desalination plant was another. The plant site is on the edge of a tidal basin, just down the street from a power plant, Andres said. There is no water intake south of the site, so the desalination plant won’t impinge on someone else’s water rights.
“We thought the project was fairly benign and that we had all the questions answered,” Hanson said. “We thought: ‘Nobody’s going to regulate how much water is taken from the ocean.’ We were wrong.”
The site is near environmentally sensitive areas, according to Metcalf & Eddy, an environmental engineering firm based in Wakefield, Mass., who designed the facility and did the pilot-plant testing. As a result, the permitting process for this facility took three to four years longer than expected. Environmentalists and the fishing industry were anxious about the plant’s effects on the ecosystem, and state regulators have been cautious. The project team had to acquire both freshwater- and ocean-based permits for this facility—more than 40 federal and state permits altogether.
“It took two-and-a-half years to get one water management permit,” Hanson said, noting that discharge permits were generally easier to obtain than withdrawal permits. Water is a precious commodity in Massachusetts, he said, and the state has many regulations to control water use and keep it within its basin. Any facility withdrawing more than 100,000 gal per day must be permitted.
The plant currently is licensed to produce up to 5 mgd of potable water, Andres said. Brockton is contracted to take up to 3.5 mgd, and Inima was negotiating with other communities in the area as of press time.
Before designing the plant, team members spent months collecting Taunton River data at 10-second intervals to track tides and note related changes in salinity and pollutant concentrations. They found that alkalinity, color, natural organic matter, salinity and temperature can change quickly, so the treatment process had to be adaptable and robust.
The team also found that the source water is seasonally brackish and saltwater intrusion only occurs at the intake site between July and November, so RO wouldn’t be necessary during the other months. The plant’s primary treatment process would have to be flexible enough to provide high-quality feedwater to the RO system during the brackish water periods, and also meet state drinking water requirements for the remainder of the year.
With this in mind, the project team designed a desalination plant that took advantage of the tides and seasonal variations. This helped control costs, Hanson said. Then, as required under state drinking water regulations, the team pilot-tested the plant for three 10-week periods between September 2002 and June 2003. Test results showed that the treatment process handled seasonal variations effectively. The team also used this opportunity to test coagulation schemes, Metcalf & Eddy noted, and polish the process to minimize costs.
“Murphy lives in the details,” Andres remarked. “So, our approach is to start with something simple and refine it.”
The full-scale plant is designed to draw raw water from the river four times a day. The first intake cycle will begin when the rising tide tops the intake weir and ends when the TDS concentration is 8,000 mg/L. The second will begin when TDS concentrations in the falling tide drop to 8,000 mg/L, and will end when river water no longer tops the weir. Because the tide rises about twice a day, this cycle will repeat as more water is needed. TDS levels at the site can be as high as 12,000 mg/L.
The plant’s intake is being carefully designed to address environmental and fishing concerns. Commercial fishing is an important local industry, and shrinking fish populations is a major issue. The industry cares about eggs and larvae as well as fish, Hanson said, so the project team is designing an intake to protect all three. The 100- by 30-ft floating structure will include a 40-micron filter—Gunderboom spun-woven netting—to minimize slot velocity and impingement. Inima is also creating 2,800-sq-ft constructed wetland to replace a small section of tidal marsh, which was removed to accommodate the intake structure.
Filtered water will be pumped to a 3.5 million-gal raw water storage and equalization tank, which is capable of up to 12 hours of water storage. The blended water then will be sent to the ZeeWeed UF system.
A year-round solution
The ZeeWeed system consists of hollow membrane fibers that are loosely suspended in cassettes and are immersed directly in open process tanks. The cassettes are connected to permeate collection headers and aeration hoses. Permeate pumps apply a slight vacuum to the end of each membrane fiber, whose 0.04-micron pores allow water in while leaving natural solids, color, organic matter and pathogens (e.g. bacteria, Cryptosporidium, Giardia and some viruses) behind in the tank. The suction process requires less energy per gallon of treated water than pressure-driven membrane systems, as well as simplifying membrane cleaning and inspection.
The UF membranes are inherently resistant to wide variations of influent turbidity and are an ideal pretreatment system for this plant. ZeeWeed membranes provide a physical barrier against particulate matter and can consistently deliver RO feedwater with a turbidity of less than 0.1 NTU and a low silt density index (SDI), typically less than 2.5, often less than 1.5. This high-quality water protects the RO system from particulate or biological fouling better than conventional pretreatment systems, which have difficulty handling variable influent quality and producing RO feedwater with a consistently low SDI.
The UF system is so effective that, between December and June, the water would require no further treatment, Andres noted. The membranes can provide superior quality potable water with 4-log removal of waterborne pathogens, such as Giardia and Cryptosporidium, and 2.5-log virus removal.
Based on pilot-test results, the UF system will use ferric chloride in an enhanced coagulation mode, which will enable it to remove up to 71% of total organic carbon, up to 88% of UV-254 and more than 99% of color. The permeate contained less than 2.5 SDI units, less than 5 mg/L of total particles and 0.02 NTU. The overall water recovery rate (total net permeate/total feed volume) is expected to be 90%.
The project team originally intended to use clarifiers to pretreat the water before RO, Hanson said, but UF really suited the site. A wastewater treatment plant discharges upstream, so UF treatment provides a greater level of public health protection and more peace of mind than clarification. During the months that RO isn’t needed, UF provides the membrane barrier required under the surface water treatment rule, Andres said.
Between May and November, when TDS levels are higher, UF permeate will be treated by three RO units, where inorganic salts, such as calcium, sodium and chloride, will be removed.
The final treatment step will add zinc orthophosphate for corrosion control, lime for pH control, and ammonia and sodium hypochlorite for disinfection prior to releasing the water into the distribution system. The RO permeate will be blended with municipal water prior to delivery to ensure consistent TDS levels for customers.
UF sludge will be clarified and centrifuged before being landfilled. More than 99% of the water removed during dewatering should be returned to the raw water storage tank, Andres said.
The RO brine is expected to contain 16,000 mg/L of TDS, according to Metcalf & Eddy’s pilot-testing estimates. To prevent it from injuring aquatic life, the brine will be diluted with low-salinity raw water and discharged during high tide, when the river’s TDS levels are highest. It will be discharged through diffusers in the intake structure.
Construction began in early April, and is proceeding on schedule. At press time, the underground concrete process tanks for the ZeeWeed system were complete. This part of the project represents about one third of the plant’s overall footprint and includes flocculation tanks, UF membrane tanks, filtered water tank, UF reject basin, neutralization basin, raw water pumping chamber and pump room. By year’s end, construction of the main process building will be well underway.