Yann A. Le Gouellec, David A. Cornwell, Robert C. Cheng, Tai J. Tseng, Diem X. Vuong, Kevin L. Wattier, Catherine J. Harrison, and Amy E. Childress

OBJECTIVES:

The capability of dual-staged nanofiltration for seawater desalination to reduce operational costs was assessed through a research plan divided into four phases:

• Theoretical basis of the system (governing equations, membrane suitability, etc.)

• Operational optimization approaches (pilot tests and use of a predictive model)

• Preliminary strategies for blending desalinated water with existing finished water

• Verification of inherent redundancy of the process with viral challenge tests

BACKGROUND:

As the quantity and quality of inland water sources decline, more coastal municipalities are looking at seawater desalination as a potential source of drinking water. The Long Beach Water Department (LBWD) developed an alternative technology to desalt seawater by using dual-staged nanofiltration (NF²). This novel NF system treats the first-stage permeate through a second stage in order to produce finished water with salinity levels that meet drinking water standards.

HIGHLIGHTS:

NF² can desalt seawater to potable water levels with less energy than is theoretically needed for traditional single-pass seawater reverse osmosis (SWRO). Additionally, boron concentration in the permeate is below California's Notification Level (1 mg/L) when Stage 2 is at pH 10. The permeate will however contain bromide ions that will exert additional chlorine demand during C·T requirements, and the brominated residuals thus formed will produce brominated DBPs and deplete disinfectant residual when desalinated water is blended with surface water. Thus, controlling the effects of bromination will be essential for system implementation.

APPROACH:

Three commercially-available NF membranes were selected for this study based upon their designation as NF membranes by the manufacturers and their salt rejection characteristics. Results from the bench-scale evaluation were integrated in a performance-predicting model, which was subsequently calibrated against the results obtained with an 8-gpm pilot unit. The pilot-test plan considered the impact of temperature, pressure, and array configuration on permeate water quantity and quality. The percentage of desalinated water that could be blended into LBWD's distribution system was determined by taking into account the issues of disinfection by-product (DBP) formation, disinfectant residual, and corrosivity. Finally, viral challenge tests were considered to verify the inherent redundancy of the system and the impact that recycling streams would have on virus accumulation.

RESULTS/FINDINGS:

Pressure in Stage 2 is the key factor impacting final salinity. Because Stage 1 contributes 80 percent of the total energy required, any efforts at energy savings should focus there. The preference for divalent ion removal in Stage 1 makes the process well-suited for boron control by increasing pH in Stage 2. However, bromide ion presence in the permeate raises two issues:

• Bromide exerts chlorine demand. Therefore, the chlorine dose needs to be increased to meet C·T requirements.

• Hypobromous acid (HOBr) formed during chlorination reacts with ammonia to form bromamines, which, unlike chloramines, will react with TOC when desalinated water is blended with surface water to produce brominated DBPs and deplete disinfectant residual.

Corrosion potential when blending desalinated water into the existing distribution system can be mitigated by increasing corrosion inhibitor dosage, maintaining the desalinated water high pH resulting from the boron removal step, or keeping the proportion of desalinated water in the final blend to less than 25 percent. MS-2 phages inoculated in a seawater matrix resulted in inconsistent recoveries under laboratory conditions. Therefore, the research team determined that running pilot viral challenge tests with MS-2 would not be meaningful since log-removal calculations involve phage quantification.

IMPACT:

Additional investigation is needed to address the residual disinfection issue. In particular, the impact of maintaining a high pH for corrosion control purposes (mentioned above as an option) on chloramination needs to be assessed. Controlling the effects of bromination will be essential for system implementation. The poor recovery of MS-2 phages in seawater matrix indicates that other surrogates are needed.

RESEARCH PARTNER:

Long Beach Water Department (LBWD)

PARTICIPANT:

University of Nevada, Reno