UC Berkeley

Water reuse and desalination are increasingly considered as viable sources of potable water because improvements in materials and designs have decreased the cost of reverse osmosis (RO) membranes and their operation. Although most contaminants are efficiently rejected by reverse osmosis membranes, compounds with neutral charge and low molecular weight have proven to be difficult to remove. Depending on the characteristics of the membrane and the feed water, some contaminants may be present in reverse osmosis permeate at concentrations that are high enough to compromise water quality.

When chemical disinfection is applied in desalination systems, compounds that pose potential risks to the human health and aquatic ecosystems or impact the aesthetic quality of drinking water may be formed. In particular, several compounds of concern are produced when chlorine is used as pretreatment. The formation and speciation of chlorination byproducts in desalination systems is affected by the elevated concentrations of bromide and iodide in seawater and desalinated product water. To gain insight into byproducts most likely to be formed in desalination systems, disinfection byproduct formation studies conducted in saline source waters, coastal power stations and existing desalination systems were reviewed. These prior studies suggested that chlorine, chloramine and chlorine dioxide all pose potential risks in desalinated water systems. Chlorination of seawater intakes to prevent membrane fouling and disinfection of blended product water both pose potential risks to water quality.

To assess the formation and fate of chlorination byproducts under different conditions likely to be encountered in desalination systems, trihalomethanes, dihaloacetonitriles, haloacetic acids, and bromophenols were analyzed in water samples from a pilot-scale seawater desalination plant. In the pilot plant, the rejection of neutral, low-molecular-weight byproducts ranged from 45% to 92%, while charged species of similar molecular weights ranged between 77% to 97% rejection. Bench-scale chlorination experiments conducted on seawater from various locations indicated significant temporal and spatial variability for chlorination byproduct formation that could not be explained by bulk measurements of dissolved organic carbon concentration and UV absorbance.

When desalinated water was blended with freshwater, elevated concentrations of bromide in the blended water enhanced dihaloacetonitrile formation through a shift in the active disinfecting agent from hypochlorous acid (HOCl) to hypobromous acid (HOBr). In most situations, data from the pilot plant and bench-scale studies indicated that chlorination byproducts formed from continuous chlorination of seawater or blending of desalinated water and freshwater will not compromise water quality or pose significant risks to aquatic ecosystems. However, blends of desalinated seawater with water rich in humic substances could lead to higher-than-expected production of haloacetonitriles and other chlorination byproducts.

When reverse osmosis is used for the treatment of municipal wastewater effluent, compounds that exhibit low taste and odor thresholds could compromise water quality. To assess potential for odors in wastewater effluent to compromise potable water reuse schemes, we evaluated odors in secondary effluent using flavor profile analysis and gas chromatography with olfactometry detection (GC/Olfactometry or GC/Olf). The primary odor reported in secondary effluent samples was classified as earthy/musty and was typically present at an intensity well above the odor threshold. Using GC/Olfactometry on samples prior to reverse osmosis, we identified sixteen peaks present at high intensity in more than 80% of the wastewater effluent samples. Odor descriptors reported in GC/Olfactometry analysis of secondary effluent were categorized as fragrant, sulfide, rancid, and hydrocarbon/chemical. Potential odorants associated with olfactometry peaks were identified by comparing the odorant with sensory descriptors and gas chromatography and mass spectrometry (GC/MS) analysis of an authentic standard of the putative compound. Other than organosulfide, aldehydes and volatile acid odorants previously identified in wastewater treatment, compounds including 2-pyrrolidone, lactones, chlorophenol, and vanillins were also identified as odorants associated with olfactometry peaks.

Potent odorous compounds were detected in secondary effluent by quantitative GC/MS. The most prominent compounds were 2,4,6-trichloroanisole (median concentration 9.5 ng/L) and geosmin (median 7 ng/L). Both compounds exhibited earthy/musty sensory profiles at these concentrations. During advanced treatment, olfactometry peaks exhibited variable fate depending on their abundance, molecular structures and odor thresholds. Reverse osmosis significantly decreased the concentrations of low molecular weight odorous compounds in wastewater, but did not eliminate all odors. Odor peaks were typically reduced to below their odor thresholds during advanced oxidation processes (i.e., UV/H2O2¬) but also in a system employing biologically activated carbon (BAC) with ozone pretreatment. Odors can be removed from secondary effluent by applying multiple barrier treatment trains that combine reverse osmosis or another physical treatment method with chemical oxidation.