Ground Water Recharge Using Waters of Impaired Quality (1994)

One result of the growing competition for water is increased attention to the use of artificial recharge to augment ground water supplies. Stated simply, artificial recharge is a process by which excess surface water is directed into the ground—either by spreading on the surface, by using recharge wells, or by altering natural conditions to increase infiltration—to replenish an aquifer. Artificial recharge (sometimes called planned recharge) is a way to store water underground in times of water surplus to meet demand in times of shortage. Water recovered from recharge projects can be allocated to nonpotable uses such as landscape irrigation or, less commonly, to potable use. Artificial recharge can also be used to control seawater intrusion in coastal aquifers, control land subsidence caused by declining ground water levels, maintain base flow in some streams, and raise water levels to reduce the cost of ground water pumping.

It is useful to think of the entire artificial recharge operation as a water source undergoing a series of treatment steps during which its composition changes. The constituents of potential concern depend not only on the character of the source water, but also on its treatment prior to recharge (pretreatment), changes that occur as it moves through the soil and aquifer (soil-aquifer processes), and treatment after withdrawal for use (posttreatment).

This report discusses three types of source waters having very different characteristics—treated municipal wastewater, stormwater runoff, and irrigation return flow—that have been proposed for use in artificial recharge. Normally, each of these source waters needs to be subjected to some kind of pretreatment before being introduced into the soil or aquifer. The exact pretreatment operations required depend on the type of source water, the nature of the recharge

process, and the intended use of the recovered water. A fundamental assumption of this report is that wastewater used to recharge the ground water must receive a sufficiently high degree of treatment prior to recharge so as to minimize the extent of any degradation of native ground water quality, as well as to minimize the need for and extent of additional treatment at the point of extraction.

After pretreatment, the water is ready for recharge, either through surface spreading and infiltration through the unsaturated zone or by direct injection into ground water. Recharge by infiltration takes advantage of the natural treatment processes, such as biodegradation of organic chemicals, that occur as water moves through soil. The quality of the water prior to recharge is of interest in assessing the possible risks associated with human exposures to chemical toxicants and pathogenic microorganisms that might be present in the source water. Although one can reasonably expect that such constituents will often be reduced during filtration through the soil, as well as subsequently in the aquifer, a conservative approach to risk assessment would assume that toxicants and microorganisms are not completely removed and some are affected only minimally prior to subsequent extraction and use. Thus when recharge water is withdrawn later for another purpose, it may require some degree of posttreatment, depending on its intended use.

Taking a systems perspective that encompasses all steps from pretreatment, through recharge, through transformation and transport, to extraction, this report assesses the issues and uncertainties associated with the artificial recharge of ground water using source waters of impaired quality. In particular, the report focuses on the methodologies and nature of the recharge systems and the subsequent impacts on the native ground water quality, especially as those impacts may affect public health following use of the recovered water. Economic, institutional, and regulatory questions are examined as well. First, this chapter presents a primer on artificial recharge of ground water to give the reader an introduction to the philosophy and techniques of the field.

Water continually evaporates from the oceans and other open water bodies, moves across the land as water vapor in clouds, falls back on the land as rain and snow, and then returns to the oceans through rivers and underground pathways to start the cycle—the hydrologic cycle—again. Part of the water that fails on the land evaporates from the soil or is transpired from plants back into the atmosphere. Another part flows overland to stream channels, lakes, or the sea. The remainder seeps downward through the soil under the influence of gravity to enter the ground water system. Once in the ground water system, the water moves slowly in response to ground water slopes or hydraulic gradients until it reenters the surface part of the cycle.

The term ground water applies to water of higher than atmospheric pressure

contained below the land surface in saturated fractures, cracks, cavities, and pore, spaces in geologic formations. It is distinct from water in the unsaturated zone, which can be at or below atmospheric pressure and is contained in films and pores in the partially air-filled soil region between the ground water zone and the soil surface. This upper region containing soil, water, and air is called the unsaturated, or vadose, zone. The term recharge is used for water entering the ground water system and the term discharge applies to water leaving it. Geologic units permeable enough to yield appreciable amounts of ground water to wells are termed aquifers.

The upper surface or boundary of the zone of complete saturation in the ground water system is called the water table. In general, the water table stands higher under hills than under valleys, and these differences in water table height (also referred to as head differences) provide the hydraulic gradients that cause ground water to flow from recharge areas to areas of discharge. Ground water in the shallow geologic units containing the water table is called unconfined because the water is more or less in direct contact with the atmosphere.

Recharge to a water table aquifer occurs wherever rainfall or surface water infiltrates downward through the soil to the water table. Generally, the recharge area of an aquifer is the entire land surface overlying the aquifer, although certain portions, such as those lying under lakes or streams, may supply much of the recharge volume. Ground water in deeper geologic units separated from the water table beds by confining layers is said to be under confined, or artesian, conditions. The height of the water level in a wen open to a confined aquifer marks the position of the potentiometric surface of that aquifer. Where the height of the potentiometric surface is higher than that of the land surface, wells open to the confined aquifer will flow freely (Figure 1.1).

Recharge to a confined aquifer can occur if the pressure of its water is less than that in the overlying or underlying aquifers that adjoin it. Discharge of ground water takes place through springs, streams, wetlands, lakes, tidal waters, and pumped wells. Water tables and potentiometric surfaces fluctuate seasonally, generally by several feet, in response to natural variations in rates of recharge and discharge.

Ground water is constantly in motion, following hydraulic gradients from points of high head to points of low head in an aquifer system. Flow of ground water is always laminar, except near large springs or pumped wells, where it may be turbulent. The velocity of ground water flow in aquifers generally ranges from a few inches to a few feet per day and is determined by the porosity, permeability, and hydraulic gradient.

Ground water naturally contains concentrations of various mineral substances that have been dissolved from the local soil and geologic formations. In

general, the longer the water remains in the earth and the higher its temperature, the higher the concentrations of dissolved substances. Thus, deep ground waters generally tend to contain more dissolved salts than shallow ground waters. Disposal of man-made wastes adds additional substances to the ground water, sometimes degrading the quality of the water so that it no longer is potable. In coastal areas, the inland extent of salty ground water is controlled by heads in the fresh ground water system. Lowering of those heads by pumping induces saltwater intrusion.

Artificial recharge is the process of spreading or impounding water on the land to increase the infiltration through the soil and percolation to the aquifer or of injecting water by wells directly into the aquifer. Surface infiltration systems can be used to recharge unconfined aquifers only. Confined aquifers can be recharged with wells that penetrate the aquifer. Well recharge is also used for unconfined aquifers if suitable land for infiltration systems is not available.

Artificial recharge can be done using any surplus surface water. When low quality water is used for recharge, the underground formations can act as natural filters to remove many physical, biological, and chemical pollutants from the water as it moves through. Often, the quality improvement of the water is actually the main objective of recharge, and the system is operated specifically using the soil and the aquifer to provide additional treatment to the source water. Systems used in this way are called soil-aquifer treatment (SAT), or geopurification, systems.

The water extracted from SAT systems often can be used without further treatment to support recreation, landscape irrigation, and other nonpotable purposes. Potable use may require more treatment. Because aquifers usually are much comer than vadose zones, the quality improvement of the water is much less in the aquifer than in the vadose zone. Thus, recharge using wells in confined aquifers cannot be expected to produce major improvements in the quality of the water. If low-quality water is to be used for well injection, it must be treated to meet the desired reuse qualities before injection. In addition, adequate treatment of the water before recharge is necessary to reduce clogging of the recharge wells. An overview of sources of water, treatment options, recharge systems, recovery techniques, and uses of the water after recovery is given in Table 1.1.

Surface infiltration systems designed to provide artificial recharge of ground water require permeable soils (sandy loams, sands, gravels) that have relatively high infiltration rates and that can transmit the applied water without completely

FIGURE 1.2 T-levees for spreading water in the Santa Aria River, California. Except for some storm runoff events, the water is almost all treated municipal wastewater from inland cities. Credit: H. Bouwer, U.S. Water Conservation Laboratory, Phoenix, Arizona.

saturating the zone above the ground water. Conventional infiltration systems can be grouped into in-channel and off-channel systems. In-channel systems are weirs, dams, or T- or L-shaped levees that spread the water over a streambed or floodplain (Figure 1.2). Dams must be built with adequate spillways or washout sections to handle spring runoff or other periodic large flows. Inflatable rubber dams that are deflated to pass large flows also can be used. The smaller weirs and levees are often considered expendable and are easily reconstructed after damage by high flows. Off-channel systems may consist of old gravel pits or of specially built basins (Figure 1.3). In-channel and off-channel infiltration systems are common in California, where there are a large number of successful recharge projects.

The range of infiltration rates (i.e., the rate water drains into the ground when a basin is flooded) for in- and off-channel systems is about 0.3 to 3 meters

(m) (1 to 10 ft) per day, including any effects caused by clogging (Bouwer and Rice, 1984). Systems with year-round recharge and periodic drying and cleaning of the bottom typically have hydraulic loading rates of 30 to 300 m/year (98 to 980 ft/year). Evaporation rates from water surfaces and wet soils range from 0.3 m/year (1 ft/year) or less in cool, humid climates to 2.5 m/year (8.2 ft/year) or more in warm, dry climates. Thus, evaporation losses are much less than the amounts that infiltrate into the ground. If the basin bottoms are not covered by sediment or other clogging material and ground water levels are sufficiently low to not affect infiltration, infiltration rates are about the same as the vertical hydraulic conductivity of the soil, which may be about 0.3 m/day (1 ft/day) for sandy loams, 1 m/day (3.3 ft/day) for loamy sands, 5 m/day (16 ft/day) for fine sands, 10 m/day (33 ft/day) for coarser sands, and 20 to 50 m/day (66 to 160 ft/ day) for fine or clean gravel. Sand and gravel mixtures have lower hydraulic conductivities than the sand alone (Bouwer and Rice, 1984). To achieve optimal infiltration rates, a number of features need to be considered in the design process, including clogging, water depth, and ground water level.

A major operational feature of infiltration systems for artificial recharge of ground water is soil clogging caused by to accumulation of suspended solids on the bottom and banks of the infiltration facility as they settle or are strained out on the soil surface. The suspended solids can be inorganic (e.g., clays, silts, fine sands) or organic (e.g., algae, bacterial flocks, sludge particles). Also, biofilms can grow on the bottom. Some mobile bacteria actually may produce mats of polymer strands, which can then strain out fine suspended particles. Thus, clogging layers may consist of a mixture of organic and inorganic products. Their thickness may range from 1 millimeter (mm) (0.039 inch) or less to 0.3 m (1 ft) or more. They have a low permeability and, hence, they reduce infiltration rates (Bouwer, 1982).

As the clogging layer forms and its hydraulic resistance increases, the clogging layer becomes the controlling factor of the infiltration process, and infiltration rates decrease. When infiltration rates become unacceptably low, the infiltration system must be dried to restore infiltration rates. If the clogging layer is primarily organic, as, for example with treated municipal wastewater, drying alone may be sufficient to restore infiltration rates. The clogging layer then partly decomposes, cracks, and curls up to form flakes on the bottom. When the basins are flooded again, essentially normal infiltration rates are obtained until the clogging process repeats itself. The problem, then, is to find the optimal combination of flooding and drying periods that yields maximum long-term accumulated infiltration rates.

Because of the many variables involved, such optimal combinations are best found by site-specific experimentation. For treated municipal wastewater with a

low suspended solids content (for example, less than 10 milligrams per liter (mg/ 1)), such combinations typically range from 2 days of flooding and 5 days of drying to 2 weeks of flooding and 2 weeks of drying. Hydraulic loading rates (i.e., average infiltration over time, including wet, dry, and cleaning cycles) then are typically in the range of 30 to 200 m/year (98 to 660 ft/year), depending on effluent quality, soil, and climate (Bouwer, 1982). Flooding and drying schedules may also be controlled by environmental factors such as breeding of insects, protection of wildlife, formation of floating algae, odors, and recreational uses of the infiltration system.

As clogging material continues to accumulate on the bottom and banks of the infiltration facilities, it eventually reduces infiltration rates so much that it should be removed. For treated municipal wastewater this process may have to be repeated every 1 or 2 years if the municipal wastewater has had adequate pretreatment and clarification (suspended solids contents less than 10 mg/1). On the other hand, if wastewater with a very high suspended solids content is used the clogging layer may have to be removed at the end of every drying period.

Clogging layers promote unsaturated flow in the vadose zone, and they are active biofilters that can remove fine suspended solids, microorganisms, organic carbon, nitrate, and metals from the water as it moves through them. This property can be of great importance in wastewater lagoons or constructed wetlands where underlying ground water needs to be protected against pollution. Clogging layers may also be desirable for infiltration systems in very coarse sands to reduce infiltration rates and enhance SAT benefits if water of low quality is used.

The water depth in infiltration basins is selected carefully. While high hydraulic heads produced by deep water result in high infiltration rates, they also tend to compress clogging layers. Thus, contrary to intuitive expectations, deep basins can produce lower infiltration rates than shallow basins (Bouwer and Rice, 1989). Also, the rate of turnover of the water in deep basins may be less than in shallow basins, allowing suspended algae (for example, Carteria klebsii) to grow in longer exposure to sunlight. Algae causes additional clogging of the soil as the biomass is strained out by infiltration. Another undesirable effect is that calcium carbonate may precipitate because of increases in the pH of the water as photosynthesizing algae take up dissolved carbon from the water, further aggravating the clogging.

In addition to yielding higher infiltration rates, shallow basins (water depths about 20 centimeters (cm) (8 inches) or less) have the advantage that drying can start quickly (in less than 24 hours, for example) after the inflow of water into the basin has been stopped. The water then disappears quickly by infiltration, or it can be drained into a lower basin. With deep basins (for example, 10 m (33 ft)

or more, as in old gravel pits), it can take a very long time for all the water to infiltrate into the soil. Water may actually have to be pumped out of the basin to initiate drying. On the other hand, shallow basins may have more weed growth, but this can be controlled.

Another design criterion is that the ground water table must be deep enough below the infiltration system that it does not interfere with the infiltration process. This requirement applies to the mounding of the permanent water table caused by recharging, as well as to perched ground water mounds that may form over restricting layers in the vadose zone. Where infiltration rates are controlled by the clogging layer (which is the rule rather than the exception for basins and ponds), the water table must be at least 0.5 m (1.6 ft) below the bottom of the basin. This distance usually is adequate to keep the top of the capillary fringe below the basin bottom, so that infiltration rates are not restricted by underlying ground water.

Where there is no clogging layer, there is more hydraulic continuity between the water in the infiltration system and the ground water. In that case, the vertical distance between the water surface and the ground water table (at some distance from the ponds where most of the mound has dissipated) should be at least twice the width of the infiltration system (Bouwer, 1990). Thus, where ground water levels are high, maximum infiltration rates can be obtained only with long, narrow streams or basins spaced a suitable distance apart.

Where waters of impaired quality are used for recharge by surface infiltration systems, it may be desirable to keep ground water levels sufficiently low to create an adequate unsaturated zone below basin bottoms for aerobic processes and virus removal. Proposed California regulations, for example, require a minimum depth to ground water of 3 m (9.8 ft) below the basins (Hultquist et al., 1991). Other infiltration systems, however, such as those in the dunes of The Netherlands for pretreatment of Rhine water for potable use, operate essentially in the ground water zone with no unsaturated conditions. Also, where wells are drilled close to streams or lakes to ''pull" surface water through the aquifer for treatment prior to drinking water treatment (bank filtration systems), the processes also take place completely below the ground water table. Thus, there is no standard for minimum depth to ground water below infiltration basins for adequate quality improvement of waters of impaired quality.

Impaired quality water sources can vary in quality. For relatively unpolluted water, the most important quality parameters applicable when considering ground water recharge are suspended solids (SS) content, total dissolved solids

(TDS) content, and the concentrations of major cations such as calcium, magnesium, and sodium. As noted earlier, suspended solids cause clogging of bottoms and banks. Where the SS content is too high, presedimentation basins with possible use of coagulants may be required. On-site experimentation will determine the best combination of pretreatment and cleaning schedules for maximum hydraulic capacity and economy of operation.

High concentrations of sodium ions in soil water, particularly in water of overall low salt concentration, can break apart aggregated clay and liberate individual colloids that subsequently plug soil pores and reduce hydraulic conductivity and infiltration rate. TDS levels and concentrations of calcium, magnesium, and sodium (as reflected by the sodium adsorption ratio (SAR)) are used to determine whether a water will disperse or flocculate clay (Bouwer, 1978). Because of effects on status of clay in the clogging layer, for example, recharge basins in California typically produce higher infiltration rates for Colorado River water (high TDS, low SAR) than for Sacramento River water from the California aqueduct (lower TDS, higher SAR). Thus, TDS, calcium, and magnesium contents should be high enough and sodium content low enough in the recharge water to keep clay in the clogging layer and below in a flocculated, more permeable state. Aquifers sometimes contain clay lenses. If the recharge water has a low TDS and/or a high sodium concentration, this clay could then become dispersed and move with the ground water through the coarse layers of the aquifer, causing the water pumped from wells in the aquifer to be "muddy."

Water quality issues ultimately affect the intended use of the recharge water and whether it is suitable for potable or nonpotable purposes. When potable uses are planned, the parameters of concern include microorganisms, trace-inorganic chemicals, and anthropogenic organic chemicals including disinfection by-products. These constituents and issues related to them are examined at length in Chapter 4 in this report.

Where treated municipal wastewater or stormwater runoff are used in surface infiltration systems, the vadose zone and in some cases the aquifer act as natural, slow filters that typically reduce the concentration of various pollutants due to physical, chemical, and microbiological processes. Suspended solids are filtered out; biodegradable organic compounds are decomposed; microorganisms are adsorbed, strained out, or die because of competition with other soil microorganisms; nitrogen concentrations are reduced by denitrification; synthetic organic compounds are adsorbed and/or biodegraded; and phosphorous, fluoride, and heavy metals are adsorbed, precipitated, or otherwise immobilized. Thus, soil-aquifer treatment can be an important step in the treatment train for reuse of wastewater.

Most soil-aquifer treatment processes take place in the upper part of the

vadose zone where soils generally are finer and have a greater organic matter content than in the aquifer, the flow is unsaturated, and oxygen levels vary from aerobic to anaerobic. To protect high-quality native ground water and nearby drinking water wells against encroachment by sewage-derived water or recharge water of other impaired quality, the systems normally are designed as recharge-recovery systems, where all the recharge water is taken out of the aquifer again with strategically located wells, drains, or other interceptors (Figure 1.4).

If the recovered water is 100 percent treated municipal wastewater, as is possible in systems B and C in Figure 1.4 (assuming for system B that there are also infiltration basins on the other side of the drain), the recovered water can be treated further (posttreatment) to meet the quality requirements for the intended use. This approach permits selection of the most economical treatment train of pretreatment, SAT, and post-treatment to achieve the desired quality of the final product water.

For the United States, the typical treatment train for municipal wastewater might include primary and secondary treatment followed by disinfection, SAT, and no treatment of the withdrawn water if it is to be used for nonpotable purposes. If viruses and other pathogens are found in the water after SAT, it can be disinfected further if it is to be used for unrestricted irrigation (e.g., of crops consumed raw or brought raw into the kitchen, parks, playgrounds, golf courses, private yards) or unrestricted recreation (such as lakes for swimming). Primary treatment alone may be sufficient as pretreatment for municipal wastewater. The higher total organic carbon (TOC) content of primary effluent may enhance the quality improvement gained via SAT because it increases denitrification and biodegradation of TOC. The latter is due to the co-metabolism and secondary utilization brought on by the greater availability of organic carbon (McCarty et al., 1984). The level of pretreatment needed is highly site- and use-specific and can be selected only after careful study.

Where the water after SAT is to be used for drinking, posttreatment may be necessary to remove residual TOC and possibly pathogens that have survived SAT. The treatment could include activated carbon filtration, reverse osmosis or other membrane filtration, and disinfection. Also, more pretreatment may be done (e.g., nitrogen removal, filtration, and disinfection) where required by local regulations, where soils and aquifers are too coarse to provide adequate treatment, or where the water after SAT is pumped from a random or nonsystematic layout of wells, which makes treatment after SAT unfeasible.

Sometimes, dilution with native ground water is relied on to allow potable use of the recovered water without further treatment. Proposed California regulations, for example, require that well water from SAT systems using treated municipal wastewater consist of not more than 20 or 50 percent sewage (depending on the level of pretreatment and site conditions) (Hultquist et al., 1991).

FIGURE 1.4 Schematic showing lines of flow in recharge - recovery SAT systems with (A) natural drainage of renovated water into stream, lake, or low area, (B) collection of renovated water by subsurface drain, (C) infiltration areas in two parallel rows and line of wells midway between, and (D) infiltration areas in center surrounded by a circle of wells. Source: H. Bouwer.

Infiltration systems for artificial recharge of ground water or SAT systems for treatment and storage of waters of impaired quality must be tailored to local hydrogeology, quality of input water, and climate. In general, basin water depths should be less than 30 cm (1 ft), and the basins should be hydraulically indepen-

dent so that each can be flooded, dried, and cleaned according to its best schedule. Inlet structures must not cause soil erosion that could clog basin bottoms. Drying periods should be started before infiltration rates have reached low values so that drying can be achieved by natural infiltration, and pumping or draining the basins is not necessary. There should also be a sufficiently large number of basins to permit flexible operation (variable flooding, drying, and cleaning cycles), with some basins in reserve to handle maximum water flows or flows during periods of low infiltration rates. Rates can be low, for example, in the winter when the water is cold, drying is slow, and infiltration recovery is incomplete, or in the summer when algae and bottom biofilms grow faster.

Where there is no local experience with artificial recharge, adequate site investigations and local experimentation with a pilot or test project are necessary, especially if the source water is treated municipal wastewater or other low-quality water. The results from such pilot projects are then used to develop design and management criteria for optimal performance of the full-scale system. Even when the full-scale system is constructed and in operation, fine-tuning may be necessary to improve its performance. Finally, infiltration systems must be closely managed to monitor their performance and to allow for quick action when something goes wrong.

When recharge systems using treated municipal wastewater are underdesigned or not properly managed (e.g., not dried and cleaned frequently enough), their infiltration rates go down and operators eventually have to fill all the basins with no time for drying. This situation causes further declines in infiltration rates while the wastewater keeps coming in and water depths in the basins increase. This compresses the clogging layer and causes more growth of algae in the basins, and infiltration rates are further reduced until eventually the whole system is ineffective. To prevent this reduction in efficiency, there must always be sufficient basin area to maintain shallow water depths and to allow regular drying and cleaning for maintaining infiltration rates.

Ground water recharge with surface infiltration systems is not feasible where permeable surface soils are not available, land is too costly, vadose zones have restricting layers or undesirable natural or synthetic chemicals that can leach out, or aquifers have poor-quality water at the top or are confined. For those conditions, ground water recharge with recharge wells is an option, These wells are similar to regular pumping wells. For unconsolidated aquifers (sand, gravel), they consist of a casing, screen, gravel pack, grouting, and a pipe to apply water to the well for infiltration into the aquifer (Figure 1.5). For consolidated aquifers (sandstone, fractured rock, limestone with secondary porosity) the portion of the well in the rock is completed as an open borehole without screen or envelope.

FIGURE 1.5 Schematic of injection well. Source: Modified from Schneider, B. J., H.F.H. Ku, and E. T. Oaksford. 1987. Hydrologic effects of artificial recharge at East Meadow, Long Island, New York. U.S. Geological Survey Water-Resources Investigations Report 85-4323, Figure 8, p. 19.

Some recharge wells have several injection pipes to recharge several confined aquifers.

Beyond their increased costs, the major problem with injection wells is clogging of the aquifer around the well, especially at the borehole interface between gravel envelope and aquifer where suspended solids can accumulate and bacterial growth tends to concentrate. Injection wells are much more vulnerable to clogging than surface infiltration systems because the infiltration rates into the aquifer around the borehole are much higher than in infiltration basins. In addition, remediation of clogging in wells is much more difficult than in surface infiltration systems. For injection wells, clogging effects can be remediated by

periodic pumping of the wells to reverse the flow and dislodge clogging materials. When recharge wells are pumped, the first water coming out typically is brown and odorous, and must be treated as wastewater or recycled through the water treatment plant. Pumping schedules may range from 20 minutes each day to a few times per year, depending on how fast recharge rates decline. If pumping does not restore recharge rates, redevelopment of the well by surging, jetting, or other conventional well development technique is necessary. Clogging effects can also be overcome by increasing the pressure of the water inside the well. However, increasing the injection pressures too much may actually exacerbate clogging and can cause upward flow of water around the casing or grouting of the well and piping.

The best strategy for dealing with clogging of injection wells is to prevent it by proper treatment of the water before injection. This means removal of suspended solids, assimilable organic carbon, nutrients such as nitrogen and phosphorus, and microorganisms. Also, chlorine is added to maintain a residual chlorine level in the well to minimize microbiological activity. Clogging parameters such as the membrane filtration index, assimilable organic carbon, and clogging in test columns with much higher velocities than in the actual recharge well system (Peters and Castell-Exner, 1993) are useful for identifying relative clogging potentials of various waters. The parameters cannot be used to predict clogging and declines in recharge rates for planned wells because actual clogging often is erratic, seasonal, and sensitive to small changes in water quality and must be evaluated with test wells.

Other processes that can decrease recharge rates in wells are precipitation of calcium carbonate, iron oxides, and other compounds in the aquifer, dispersion and swelling of clay, and air binding. The last can occur when the recharge water is cooler than the aquifer water. As the temperature of the recharge water increases in the aquifer, dissolved air goes out of solution and forms air pockets in the aquifer. These block the pores and reduce the hydraulic conductivity and, hence, the recharge rates. For this reason, dissolved air concentrations in the recharge water should always be as small as possible, and free-falling water in the recharge well should be avoided to prevent air entrainment.

Where municipal wastewater is used for ground water recharge with injection wells, it must undergo extensive pretreatment, including advanced wastewater treatment (AWT) processes. This is necessary to achieve a quality that is essentially the same as that of the drinking water standards before recharge because aquifers generally are too coarse for significant soil-aquifer treatment. The only treatment processes that can be expected in aquifers are some additional TOC removal, removal of some microorganisms, improvement in taste and odor, and similar "aging" and "polishing" effects. In addition, AWT is necessary to minimize clogging in the well. Where the AWT includes reverse osmosis (RO), the water will have a very low TDS concentration, which makes it "hungry" and therefore corrosive. The interaction between this water and the

receiving aquifer must then be well understood to make sure that the recharge water does not mobilize undesirable chemicals from minerals and other solid phases of the aquifer. Thus blending of the water after RO with water of a higher TDS content before recharge may be necessary.

A rapidly growing practice in artificial recharge is the use of aquifer storage recovery (ASR) wells, which combine recharge and pumping functions. They are used for recharge when surplus water is available, and are pumped when the water is needed. ASR wells typically are used for seasonal storage of drinking water in areas where water demands are significantly greater in summer than in winter, or vice versa. With these wells drinking water treatment plants can be built to meet average, rather than peak, demands.

Dry wells are boreholes in the vadose zone, usually about 10 to 50 m (33 to 160 ft) deep and about 1 to 1.5 m (3.3 to 4.9 ft) in diameter. They are widely used for infiltration and disposal of stormwater runoff in areas without storm sewers or combined sewers and, hence, they produce incidental recharge of ground water. There is concern that this causes pollution of ground water, but so far studies have not been able to document this. There is also concern about illegal disposal of waste fluids through the dry wells (by so called "night dumpers"). Dry wells normally are drilled into permeable formations in the vadose zone that can accept the runoff water at sufficient rates. Where ground water is relatively deep, dry wells are much cheaper than injection wells and, hence, it is tempting to use dry wells to recharge the ground water instead of injection wells that must go all the way down to the aquifer. To provide adequate recharge, the dry wells should penetrate permeable formations for a substantial distance.

The main problem with dry wells is clogging of the walls. It is impossible to remediate such clogging by pumping or redeveloping because the dry well is in the vadose zone and ground water cannot flow into it. Thus, clogging must be prevented or minimized. One way to do this is to protect the water in the well against the slaking and sloughing of clay layers that could make the water in the well muddy. Slaking can be avoided by filling the well with sand and placing a perforated pipe in the center to carry the water for recharge.

In addition, the water must be pretreated before recharge to remove all clogging agents, including suspended solids, biodegradable organic carbon, nutrients, and microorganisms, and it must be disinfected to maintain a residual chlorine level. If clogging still occurs (and long-term clogging is always a possibility), it will then mostly be caused by bacterial cells and organic metabolic products such as polymers on the well wall (biofouling). Thus, while such

clogging cannot be remediated by pumping, cleaning, or redevelopment, a long drying period perhaps could produce significant biodegradation of the clogging material to restore the dry well for recharge.

The choice between using dry wells in the vadose zone or injection wells in the aquifer is governed by economics. The costs of installing the wells, the pretreatment requirements of the water, the useful lives of the wells, and the costs of well replacement and maintenance and remediation need to be compared. If the vadose zone contains undesirable chemicals that can be leached out, dry wells should not be used.

The environmental effects of ground water recharge vary from site to site, and there can be both beneficial and harmful impacts. In general, however, the types of environmental effects that should be considered when planning recharge facilities range from ecological effects on soil, hydrologic, and aquatic ecosystems, to effects on species dependent on riparian habitats, and to possible effects on people's use of the water resources for recreation. As with any use of water, planners must take care to recognize that the impacts of their actions will affect not only local conditions, but also conditions downstream (third-party effects).

The ecological effects of ground water recharge are, for the most part, relatively straightforward and predictable, at least in a qualitative sense. If water is diverted directly from a stream or other surface water source, the reduction in downstream flow will have the same ecological consequences as a diversion for any other purpose that results in a reduction in streamflow with the same timing and quantity. Ecological effects often are difficult to quantify. In almost all cases, they are site specific and difficult to generalize.

During the actual process of recharge, recharge basins produce some significant changes in the ecosystem, both at the surface of the soil and in the soil profile. The nature of these changes depends on whether the recharge basin is a natural stream bed, a lagoon constructed specifically for the purpose, or a natural off-stream basin. The actual management of the recharge basin will often be established so as to manipulate the soil ecology to optimize the recharge process. Bed permeability is always a consideration, and aeration of the bed by intermittent recharge affects the population dynamics of the soil ecosystem. Degradation of organic matter and denitrification are frequent objectives.

The long-term ecological effect on the recharge basin depends on the adsorption of organics and inorganics in the upper soil horizons and on the treatment of the basin when recharge is ceased. Whether the surface topography of the basin is altered to minimize or to maintain inadvertent recharge also can be important.

Downgradient ecological effects depend on the rate of recharge and the nature of the receiving aquifer. If recharge is into a relatively deep aquifer, and

if the water table does not approach within a meter of two of the soil surface within the watershed, few ecological consequences are to be expected. On the other hand, if the water table rises anywhere within the basin to within a meter of the soil surface, one can confidently predict sufficient capillary rise to maintain a higher rate of production of vegetation at the surface than would be present otherwise, especially in arid regions. The soil ecology would be similarly affected. Again, the results will be highly site specific, and whether the consequences are considered desirable or undesirable depends on the nature of the effect and the objectives of the recharge. Normally, losses due to evapotranspiration would be regarded as counterproductive because of the loss of stored water and the increase in ground water salinity, which would subtract from other uses of the water.

Along the banks of many rivers in the western United States, the natural recharge results in substantial growth of phreatophytes, such as mesquite. Herbicides, cutting, and other methods to reduce plant growth have been used in attempts to increase streamflow. Although vegetation removal makes more water available, it usually produces a loss rather than a gain in terms of ground water storage because the water moves downstream quickly with little time for infiltration. The removal of vegetation and subsequent increased flow also facilitates erosion. Because the value of riparian habitat, especially in dry regions, is increasingly recognized, many experts believe that the harm of vegetation removal outweighs any potential benefits.

Beyond the ecological impacts at individual sites, it is important to step back and consider a broader perspective. Because ground water recharge is an option typically pursued where water is scarce, planners must be aware that water would be similarly scarce for nonhuman components of the ecosystem. Thus, the decisions that are made about the water source for recharge and about how the project will be managed could cause other components of the ecosystem to face the consequences of reduced supplies and thus have widespread implications.

By adding to the storage and flow of ground water, artificial recharge modifies the hydrologic cycle. Water from the surface environment that otherwise would not have entered the ground water reservoir is emplaced underground through infiltration or injection techniques. This modification of the local water regime may have either beneficial or adverse consequences.

In areas where the base flow of streams is supported by ground water discharge, additions to the storage and flow of ground water by recharge may result in higher sustained streamflows during low flow or drought conditions. The flow of springs might also be sustained at higher levels through dry periods by the higher ground water heads that would result from artificial recharge. The increased ground water discharge to springs and streams would dampen somewhat the amplitude of their cyclical flow fluctuations, thus helping to sustain associated wetland environments. Wetlands not associated with surface water

courses might also be sustained by the higher stand of the water table. In coastal areas the increased flow of streams and of direct ground water recharge to estuaries might help maintain less variable salinity conditions and to counter seasonal migration of the freshwater-saltwater interface in the stream channel and aquifer.

The extraction of ground water from some hydrogeologic settings can cause irreversible compaction of fine-grained beds of silt and clay in the aquifer, which in turn causes the land surface to subside. Through artificial recharge, ground water heads can be restored to or maintained at levels that can help prevent or reduce subsidence.

An indirect environmental impact may result from the fact that as ground water heads are raised by artificial recharge, less energy is used to pump a given quantity of water. This may result in a net savings in energy needs, depending on the recharging method employed and the energy required by the recharge operation. In addition, if the water used for recharge is of a higher quality than the ambient water in the aquifer, the quality of the recovered water may improve, resulting in a reduction in treatment requirements at the point of withdrawal. This, too, could result in some energy savings.

On the negative side, the diversion for ground water recharge of surface water or wastewater normally discharged to surface waters may result in a reduction of downstream flows, especially if the diversion is to another drainage basin or if the ground water reservoir does not discharge to the streams. Consequently, the reduced streamflow could result in undesirable changes to wetland environments and to salinity conditions in estuaries. Increased salinization of an estuary could result from the reduction in fresh streamflow and cause an undesirable change in the ecology of the estuary. San Francisco Bay and Florida Bay are examples of estuaries where ecological conditions have been altered by increases in salinity caused by reductions in fresh water inflows, although the diversions of fresh water in these areas are not attributable to ground water recharge facilities but are caused more broadly by increased demands for water. Stream pollution may increase and downstream appropriation rights may be jeopardized, as well, if recharge is not planned with an eye to the comprehensive needs of the region.

The raised level of the water table caused by artificial recharge sometimes can have deleterious consequences. If a water table is allowed to rise to the soil zone, soils may become water-logged and salinized and agricultural crops or native vegetation might be affected adversely. Underground structures, pipelines, gravel pits, and other facilities built during low stands of the water table could become inundated if the water table rises to their level. Dewatering schemes would be required to depress the water table locally and keep substructures dry should inundation or other adverse effects occur.

Perhaps the least predictable and the most difficult to remedy of all the potential environmental impacts of artificial recharge with wastewater is the

likelihood of degrading ground water quality. Recharge with source waters of impaired quality could introduce microbial, inorganic, and organic chemical constituents into ground water, with the potential to cause environmental problems. As mentioned earlier, biochemical and geochemical reactions between the source water and the resident ground water and/or the aquifer materials could result in mobilization of chemical constituents that are part of the mineral framework of the aquifer. In addition, artificial recharge can leach anthropogenic contaminants from the vadose zone to ground water and move pollution plumes in aquifers to where they are not wanted. The resulting water quality degradation may require that the recovered ground water receive treatment not previously needed. Natural discharges to surface waters of ground water whose quality has been altered by wastewater recharge could also be damaging to the ecology of the receiving surface water body. The inability to identify all of the organic compounds in the recharge water, coupled with the difficulty of predicting the biochemical and geochemical changes in the subsurface, creates uncertainty with respect to the potential for degradation of ground water quality and the resulting environmental and ecological consequences.

The growing competition for water in the United States will bring more and more attention to ground water resources. Artificial recharge of ground water is an established practice, and a long history of experimentation and actual full-scale recharge projects exists and clearly shows the benefits this technology can bring. Considerable experience-based, practical knowledge and experience about the problems and potentials of this technology is available (Asano, 1985). Thus, artificial recharge is likely to receive increased emphasis in the future, particularly regarding its role in the treatment and storage of municipal wastewater and other low-quality water for reuse. Given the broad body of knowledge available, the issues that arise with artificial recharge of ground water generally can be addressed through careful preproject planning and ongoing management.

There are still several operational issues that must be addressed on a site-specific basis. These concerns are related to project sustainability, treatment needs, public health impacts, and economic and institutional constraints. In the short-term, project sustainability is controlled by operating and managing the system so as to prevent or control clogging. Long-term sustainability is dependent on finding the best combination of pretreatment, soil-aquifer treatment, and posttreatment for determining whether the source waters will exceed the treatment and removal capacity of the soil-aquifer treatment system. These issues are discussed in the following chapters.

Asano, T., ed. 1985. Artificial Recharge of Groundwater. Boston, Mass.: Butterworth. 767 pp.

Bouwer, H. 1978. Ground Water Hydrology. New York:McGraw-Hill. 480 pp.

Bouwer, H. 1982. Design considerations for earth linings for seepage control. Ground Water 20(5):531-537.

Bouwer, H., and R. C. Rice. 1984. Hydraulic properties of stony vadose zones. Ground Water 22(6):696-705.

Bouwer, H., and R. C. Rice. 1989. Effect of water depth in ground water recharge basins on infiltration rate. J. Irrig. Drain. Eng. 115(4):556-568.

Bouwer, H. 1990. Effect of water depth and ground water table on infiltration from recharge basins. Pp. 377-384 in: Proceedings of the 1990 National Conference of the Irrigation and Drainage Division. American Society. of Civil Engineers, S.C. Harris ed. Durango, Colo. July 11-13, 1990.

Hultquist, R. H., R. H. Sakaji, and T. Asano. 1991. Proposed California regulations for ground water recharge with reclaimed municipal wastewater. Pp. 759-764 in: Proceedings of the 1991 Specialty Conference, Environmental Engineering, American Society for Civil Engineers. Reno, Nev. July 1991.

McCarty, P. L., B. E. Rittmann, and E. J. Bouwer. 1984. Microbiological processes affecting chemical transformations in ground water. Pp. 89-116 in Ground Water Pollution Microbiology, G. Bitton and C. P. Gerba eds. New York: John Wiley.

Peters, J. H., and C. Castell-Exner, eds. 1993. Proceedings Dutch-German Workshop on Artificial Recharge of Groundwater, Sept. 1993, Castricum, The Netherlands. Nieuwegein, The Netherlands: Keuringsinstituut voor Waterleidingartikelen.