Chapter 8. Treatment of drainage effluent

Need for drainage water treatment

Treating drainage water is normally one of the last drainage water management options to be considered. This is due to the high costs involved and to uncertainty about the treatment level achievable. The treatment of drainage water should be considered where all other drainage water management measures fail to guarantee safe disposal or where it is financially attractive. For subsurface drainage water containing very high levels of salinity, selenium and other trace elements, the treatment objectives are: i) reduce salts and toxic constituents below hazardous levels; ii) meet agricultural water management goals; iii) meet water quality objectives in surface waters; and iv) reduce constituent levels below risk levels for wildlife.

Treatment options

The treatment of agricultural drainage water presents a challenge due to the complex chemical characteristics of most drainage waters (Lee, 1994). Table 23 details the average chemical quality of subsurface drainage waters disposed into Kesterson Reservoir in the San Joaquin Valley as well as those disposed into evaporation ponds. The drainage waters are saline and of the NaCl-Na₂SO₄-type water. The waters conveyed by the San Luis Drain came from a single site in Westlands Water District in contrast to the evaporation pond waters that came from 27 sites.

Table 23. Average composition of agricultural tile drainage water in the San Luis Drain (drainage waters disposed into evaporation basins in parenthesis)

Source: SJVDP, 1990; and Chilcott et al., 1993.

There are numerous wastewater treatment processes for industrial and urban wastewater and for the preparation of drinking-water. Many of them offer potential for the treatment of agricultural drainage water. Treatment processes for drainage water can be divided into processes that reduce the total salinity of the drainage water and processes that remove specific ions. Methods for the removal of trace elements can be biological, physical and chemical.

Most desalinization processes also remove trace elements but their costs are often prohibitive. Less costly methods for the removal of trace elements are being developed. Lee (1994) has reviewed treatment technologies for drainage water. The SJVDIP (1999b) has reviewed treatment technologies for removing selenium from agricultural drainage water. The following is a brief summary of their findings.

Desalinization

There are numerous desalinization processes including ion exchange, distillation, electrodialysis and reverse osmosis. Of these processes, reverse osmosis is considered to be the most promising for the treatment of agricultural drainage water mainly due to its comparatively low cost.

Reverse osmosis is a process capable of removing different contaminants including dissolved salts and organics. In reverse osmosis, a semi-permeable membrane separates water from dissolved salts and other suspended solids. Pressure is applied to the feed-water, forcing the water through the membrane leaving behind salts and suspended materials in a brine stream. The energy consumption of the process depends on the salt concentration of the feed-water and the salt concentration of the effluent. Depending on the quality of the water to be treated, pretreatment might be crucial to preventing fouling of the membrane. Figure 44 describes one of several pretreatment reverse osmosis systems studied in the San Joaquin Valley. Other pretreatment steps could be lime treatment along with ion exchange.

Figure 44. Reverse osmosis system with lime-soda pretreatment

Source: after CH2M HILL, 1986.

Table 24 presents the results of a trial-run reverse osmosis using the lime-soda softening pretreatment (CH2M HILL, 1986). The permeate is the product (desalted) water and the concentrate is the brine water. The results show that TDS can be desalted from 9 800 to 640 ppm, boron from 14.5 to 7.6 ppm, and selenium from 325 to 3 ppb in a three-stage reverse osmosis system. The efficiency of removal declines with stages.

Table 24. Results of a trial-run for a three-stage reverse osmosis system, lime-soda pretreatment

Source: CH2M HILL, 1986.

The California Department of Water Resources conducted pilot-plant-scale reverse osmosis of saline drainage using cellulose acetate membranes. The bacterial and chemical fouling of the membrane was a major problem. The drainage water had to be treated with alum, and passed through a sedimentation pond and a chlorinated and filtration system. In spite of this level of pretreatment, the membranes tended to foul due to the precipitation of gypsum and calcite. The drainage waters are saturated with respect to calcite and gypsum. This same chemical fouling problem is being faced by the Yuma desalting plant off the Colorado River using drainage waters from the Wellton-Mohawk irrigation project. The estimated cost of desalting is more than US$0.81/m³, too expensive for irrigated agriculture but possibly affordable for municipalities with freshwater shortages. This cost does not include the management and disposal of the brine water. However, a potential exists for partially treating the average 10-dS/m-drainage water to about 2-3 dS/m for use by agriculture and wildlife.

Trace element treatment

As the technology of reverse osmosis is experimental and expensive, cheaper methods of removing toxic trace elements are being pursued.

Biological processes

Conventional column reactor systems have been utilized to remove selenium from drainage waters (SJVDIP, 1999b). Selenium is microbially reduced to elemental selenium under anoxic (anaerobic) conditions in the presence of organic carbon sources (Owens, 1998).

In the initial study, the biological reactor consisted of a two-stage upflow anaerobic sludge blanket reactor followed by a fluidized bed reactor. As selenium cannot be reduced while nitrates are present, a key treatment process is the reduction of nitrates prior to enhancing selenium reduction. The sludge blanket was seeded with inoculum from sludges from ordinary sewage treatment plants. This system yielded 30 ppb selenium product water.

A subsequent large-scale pilot study examined seven different reactor systems after upflow through a conical bottom liquid-gas-solid separator with the addition of methanol as the carbon source. The conical separator was seeded with granular sludge from a bread-making bakery. This first step reduced the average nitrate concentration from 45 to 3 ppm. The waters were then fed to a number of packed bed column reactors. The best sustained results were about a 90-percent removal of selenium from 500 to 50 ppb.

Biological treatment normally refers to the use of bacteria in engineered column reactor systems for the removal or transformation of certain constituents, e.g. organic compounds, trace elements and nutrients (Owens and Ochs, 1997). However, biological treatment also includes algal-bacterial treatment processes and wetland systems. Much research has focused on the removal of selenium from drainage effluent. Box 12 describes an example of the basics of an algal-bacterial system for the removal of selenium (SJVDIP, 1999b).

Chemical processes

Chemical treatment processes refer to the use of chemicals to remove trace elements from polluted wastewater. Chemicals are frequently used for industrial wastewater treatment but are not effective in agricultural drainage water due to their often complex chemical characteristics (Lee, 1994). Chemical processes have been developed for the reduction of selenate to elemental selenium by means of ferrous hydroxide. Under laboratory conditions, ferrous hydroxide was able to reduce and precipitate selenium by 99 percent in 30 min. In field studies, although 90 percent of the selenate was reduced, the reactor time required was up to 6 h. It appeared that dissolved bicarbonate, oxygen and nitrate influenced the reduction process.

Physical processes

Physical processes involve the adsorption of ions on natural and synthetic surfaces of active materials, including ion exchange resins. Box 13 provides an example of a mini-pilot plant for the removal of heavy metals.

Flow-through artificial wetlands

Figure 45 shows the layout of a pilot project for removing selenium by flow-through wetland cells conducted in the Tulare Lake bed, a closed basin of the San Joaquin Valley (Tanji and Gao, 1999). The goal was to remove selenium from drainage waters to a bird-safe level prior to disposal into evaporation ponds.

Figure 45. Layout of pilot-scale constructed wetland experimental plots at the Tulare Lake Drainage District

Source: Tanji and Gao, 1999.

Tile drainage effluent containing about 20-ppb selenium from an adjacent farm was passed through a sand bed filter system and metered into the cells (15.2 × 76 m) with a variety of substrates (vegetation). The inflow water was measured twice a week by a totalizing meter. The water depth in Cells 1-7 was maintained at about 20 cm, and outflow was measured by v-notch weir. Cells 8-10 had variable water depths of about 20 cm, 60 cm where widgeon grass (Ruppia) was grown. The target residence time for the flowing waters was 7 days for Cells 1-7, 21 days in Cells 8 and 9, and 14 days in Cell 10. These residence times were selected after preliminary runs for optimal removal. A residence time of three days was too short for selenium removal and a residence time of more than 21 days did not increase selenium removal. Seepage rates in the cells were about 1 cm/d and evapotranspiration slightly greater than ETo (annual value about 1 600 mm).

Table 25 presents the performance results for the year 1999 with average weekly water selenium of 18.2 ppb, over 90 percent in the selenate form (Se+6). The residence times achieved were reasonably close to target values considering the variability in monthly ETo. The selenium concentration in the outflow waters varied from 4.6 to 12.3 ppb. The ratio of outflow to inflow selenium concentration ranged from 0.25 to 0.68 (a small ratio indicates high selenium removal). The ratio of outflow to inflow selenium on a mass basis ranged from 0.07 to 0.57 or 93 to 43 removal. The cell with open water had reduced selenium because algae and microbes naturally populated the cell and contributed to some selenium removal. In terms of performance, the ratio based on mass of selenium is a good indicator. However, in terms of potential impact on birds, the outflow concentration and ratio based on concentration are better indicators.

Table 25. Performance of the wetland cells in removing selenium from drainage water with 18.2-ppb selenium

Source: Tanji and Gao, 1999.

The control volume for each cell is the standing water, plants and the rootzone. Thus, the mass flux balance on selenium for each cell is:

(22)

The righthand-side terms of Equation 22 are mass fluxes, and mass (MSe) is defined as the product of selenium concentration and water volume, except for the volatilization term. Water inflow and outflow was monitored twice a week, water seepage estimated from the difference from inflow and outflow and ET_crop from ETo * Kc, where Kc is the crop coefficient. Volatilization of selenium by microbes and plants was monitored monthly. The DM/Dt is the mass flux of selenium accumulating in the control volume (cell) consisting of the sediments, organic detrital matter, fallen litter, standing water and standing crop.

Figure 46. Initial estimate on mass balance of selenium in ten flow-through wetland cells, 1997-2000

Figure 46 presents a summary of the mass balance on selenium in the ten wetlands cells from July 1997 to September 2000. The values reported are based on the percentage of the mass of selenium in the inflow water. On average, about 35 percent of the mass inflow of selenium remained in the treated outflow water, with smaller percentages lost through seepage and volatilization losses.

The remainder of the selenium accumulated in the cell as selenium present in the sediments, organic detrital matter, fallen litter, standing water and standing plants. The values reported are the mass of selenium found in the cells in September 2000. About 11 percent of the total selenium could not be accounted for due to errors in sampling and monitoring over a four-year period, and the difficulties of analysing for reduced forms of selenium. The sink mechanisms removing selenium from the floodwater were: adsorption of selenite (Se+4) to the mineral sediments mainly in the top 10 cm or so; selenium immobilized into elemental selenium (Se0) due to reduced conditions in the organic detrital layer; and organic forms of selenium (Se-2) tied up with the detritus and fallen litter. The principal removal mechanisms were adsorption and immobilization into elemental selenium and organic selenium.

The recommended selenium water standard to protect waterbirds is 2 ppb. None of these cells achieved that level of remediation but many cells certainly will reduce selenium toxicity. However, outflow waters from these cells contain organic selenium (17-33 percent of the total selenium), which is more toxic than inorganic forms to wildlife. These and other results are currently being reviewed to determine whether selenium removal flow-through wetland cells is a viable treatment option

Evaluation and selection of treatment options

The first steps in the selection of any drainage treatment process are: i) define the problem; ii) determine the reasons for the required treatment; and iii) determine what is to be achieved. The main reason for opting for drainage water treatment is normally the desire to reuse the drainage effluent or to conform to regulatory disposal requirements. For both purposes, specific water quality criteria apply.

In order to make a preliminary selection of suitable treatment processes, it is necessary that sufficient data be available. These data consist of historical data on the chemical constituents of the drainage water, seasonal flow variations and variations in the concentrations of the constituents of concern. Once combined with information on the targeted quality of the treated effluent, it is then possible to shortlist drainage treatment processes that are theoretically suitable.

The technical capability of the treatment process is an important factor in the selection of a treatment technology. However, it is important to consider economic, financial, social and institutional criteria in order to ensure the sustainability of the treatment facilities.