Abstract Introduction Inventory of Data Case Studies Summary References Tables PDF Version

Aquifer storage and recovery (ASR) in southern Florida has been proposed as a cost-effective water-supply alternative that can help meet the needs of agricultural, municipal, and recreational users and can be used for Everglades ecosystem restoration. Plans have been made to utilize ASR on an unprecedented scale in the Central and Southern Florida Comprehensive Review Study as proposed by the U.S. Army Corps of Engineers and the South Florida Water Management District (1999). This review study is also known as the Comprehensive Everglades Restoration Plan (CERP). About 330 ASR wells have been proposed for southern Florida, each with an assumed capacity of 5 Mgal/d during recharge or recovery. Pyne (1995) has described ASR as "the storage of water in a suitable aquifer through a well during times when water is available, and recovery of the water from the same well during times when it is needed."

Figure 1. Study area and locations and status of aquifer storage and recovery sites. Status is as of April 2001. [larger image]

During this time, the ASR well system is tested prior to being given a full operating permit by the Florida Department of Environmental Protection (FDEP). Three of the sites have been given an operating permit. Additionally, six sites are no longer active after experimental testing was completed (fig. 1). These sites were operated by government agencies including the U.S. Geological Survey (USGS), South Florida Water Management District (SFWMD), FDEP, and the Florida Keys Aqueduct Authority (FKAA). ASR is a relatively recent development in southern Florida, in terms of its use as a municipal or countywide source of water; 20 active sites in this category were constructed in the 1990's (with 14 of these sites having been constructed since 1996). The strategy for this use of ASR in southern Florida has been to store excess water available during the wet season and recover this water during the dry season when it is needed.

Existing and historical ASR sites in southern Florida are mostly located along the east and west coasts (fig. 1). At most sites, the proposed or planned purpose of the recovered water is to serve as a supplemental supply for municipalities. Under CERP, ASR wells will be constructed in inland areas around Lake Okeechobee, in central Palm Beach County, and along the Caloosahatchee River in Hendry County (U.S. Army Corps of Engineers and South Florida Water Management District, 2001). Recovered water is to be used for additional purposes that include maintaining water levels in Lake Okeechobee and wetland areas and reduction of surface-water flows to tide (estuarine and bay areas) during storm events.

Figure 2. Storage zone aquifers for aquifer storage and recovery sites in southern Florida. [larger image]

ASR wells are evaluated and operated through a cyclical process. Each cycle includes periods of injection (recharge) of freshwater into the ASR well, storage, and then withdrawal (recovery) with each period lasting days or months. In southern Florida, the recovery phase may commence immediately after the cessation of recharge with no period of storage, and depending on the source of water supply, municipal supply, or operational problems, the time between cycles may be extensive (months or years). After initial testing and under fully operational conditions, cycles continue but the duration of cycles and storage periods and the volume of water recharged during each cycle usually increase.

In southern Florida, ASR is largely used to store water in an aquifer that contains brackish water. Ambient ground water in the storage zone at most of the ASR sites in the study area is brackish (greater than 1,000 mg/L dissolved-solids concentration) to saline (greater than 10,000 mg/L dissolved-solids concentration); salinity appears to greatly affect the recovery of the recharged freshwater. The salinity of the recharged and recovered water is closely monitored, usually on a daily basis. Because of the high ambient water salinity of the storage zone, much of the recharged freshwater is not recovered largely due to dispersive mixing in the aquifer.

The recovery efficiency for each cycle is the total volume of water recovered, expressed as a percentage of the volume of water recharged. The salinity of water during recovery increases with time, and recovery is terminated at a salinity level that is predetermined by operational considerations. Generally, this limiting salinity level is at the potable water limit of 250 mg/L chloride concentration, or slightly higher if the recovered water is mixed with potable water at a water-treatment plant (WTP).

Few regional investigations of the hydrogeology of the Floridan aquifer system in southern Florida have been conducted, and those studies focused on issues unrelated to ASR. Lacking a regional ASR framework to aid the decision-making process, placement of ASR well sites in southern Florida have primarily been based on factors such as land availability, source-water proximity (preexisting surface-water canal systems or surficial aquifer system well fields), or proximity to a WTP. Little effort has been made to link information collected from each site into a regional hydrogeologic analysis. Additional tools and data are needed to make informed decisions that incorporate constraining hydrogeologic factors in the placement and construction of ASR sites in southern Florida.

This study is part of the USGS South Florida Place-Based Studies Program, which was established for the purpose of providing physical and biological science data and information on which to base ecosystem restoration management decisions. The purpose of this study was to compile data on existing ASR sites in southern Florida and identify various hydrogeologic, design, and management factors that control the recovery of freshwater recharged into ASR wells.

The purpose of this report is to inventory well construction, hydrogeologic, and operational data on ASR sites in southern Florida and assess site performance. A secondary purpose is to identify hydrogeologic, design, or management factors that influence the success of ASR. Recovery efficiency, defined as the percent of recharged freshwater that is recovered for each cycle, is used to evaluate this performance. Four ASR case studies are described to determine possible technical factors that influence the success of ASR.

The study area includes all of southern Florida and includes Charlotte, Glades, Lee, Hendry, Collier, Monroe, Miami-Dade, Broward, Palm Beach, and Martin Counties, and parts of Okeechobee and St. Lucie Counties (fig. 1). The 27 ASR sites located in the study area represent the source of data for this study. However, this report focuses on the 23 ASR sites in which the Floridan aquifer system serves as the storage zone. Principal hydrogeologic and construction related attributes determined for each ASR site are graphically and spatially illustrated to provide a comparative analysis.

It has been nearly 20 years since Merritt and others (1983) provided a retrospective overview and status of ASR well development in southern Florida. Merritt and others (1983) presented data from three experimental ASR sites that are also included in this report, and Meyer (1989b) published additional data on experimental ASR sites in southern Florida. Other experimental ASR test data were obtained in reports or written communications for the Jupiter site (fig. 1, map no. 21; J.J. Plappert, Florida Department of Environmental Protection, written commun., 1977), the St. Lucie County site (fig. 1, map no. 27; Wedderburn and Knapp, 1983), the Lee County site (fig. 1, map no. 9; Fitzpatrick, 1986), the Hialeah site (fig. 1, map no. 15; Merritt, 1997), and the Taylor Creek/Nubbin Slough - Lake Okeechobee site (fig. 1, map no. 20; Quiñones-Aponte and others, 1996). Theoretical investigations into the feasibility of cyclic injection of freshwater in southern Florida have been described in reports by Khanal (1980) and Merritt (1985). Merritt (1997) also included numerical simulations of the salinity of recovered water in his study of Hialeah ASR site.

Some regional or local hydrogeologic studies of the Upper Floridan aquifer that encompass or include part of southern Florida are Bush and Johnston (1988), Meyer (1989a), Miller (1986), Reese (1994), Reese (2000), and Reese and Memberg (2000). The reports by Meyer (1989a), Reese (1994; 2000), and Reese and Memberg (2000) are specific to southern Florida.

Recovery of freshwater stored in brackish- to saline-water aquifers is controlled by a wide variety of factors that pertain to hydrogeologic conditions, well or well field design, and operational management. The hydrogeologic factors of a storage zone that are important to recoverability include (1) ambient salinity, (2) aquifer permeability and distribution, (3) aquifer thickness, (4) confinement, (5) ambient hydraulic gradient, and (6) structural setting. Important design and management factors to consider are (1) thickness and location of the storage zone within the aquifer, (2) volume of injected water, (3) duration and frequency of cycles and cycle storage periods, (4) well performance problems such as wellbore plugging, and (5) multiple-well configurations. Most of these factors and their control on recoverability have been numerically simulated (Merritt and others, 1983; Merritt, 1985); however, conclusions on some factors as discussed in the following sections, came from consulting reports and other literature.

Figure 3. Aquifer storage and recovery well in a confined aquifer depicting idealized flushed and transition zones created by recharge. Flushed zone contains mostly recharged water. [larger image]

During recharge of water by an ASR well, a radial zone of mixing forms around the well in the aquifer. This zone, referred to as the transition zone (Merritt, 1985), separates native water from an inner flushed zone containing mostly injected water, and this inner zone can be described as a freshwater bubble (fig. 3). The degree of mixing between the injected and native water and the width of the transition zone is primarily controlled by hydrodynamic dispersion. Hydrodynamic dispersion or dispersive mixing refers to the effects of molecular diffusion and mechanical dispersion. Mechanical dispersion results from the unevenness of flow through porous media, and at flow velocities occurring during ASR recharge and recovery, this dispersion will dominate over diffusion.

The ambient salinity of water in the storage zone is of primary importance in controlling recovery of freshwater because of mixing with this water and potential buoyancy stratification. Buoyancy stratification occurs where the ambient salinity is high, provided permeability in the aquifer is also high (Merritt, 1985); the injected freshwater moves upward and flows out over the native ground water. During the recovery phase, such stratification increases mixing. Buoyancy stratification should be considered possible when the ambient ground water has a dissolved-solids concentration greater than 5,000 mg/L (Pyne, 1995); in the Floridan aquifer system of southern Florida this equates to about 2,500 mg/L chloride concentration (Reese, 1994). On the basis of numerical simulation, recovery efficiency has been shown to decrease with increasing salinity in saline aquifers only because of dispersive mixing in the transition zone - no buoyancy stratification (Merritt, 1985). Ambient water salinities modeled in Merritt's study, as defined by chloride concentration, were 2,000, 7,000 to 8,000, and 19,000 mg/L (seawater-like salinity).

The permeability or hydraulic conductivity of the storage zone may greatly affect recoverability. The probability of buoyancy stratification increases as permeability increases (Merritt, 1985). Additionally, mechanical dispersion is related to the distribution of permeability within the storage zone. Higher permeability can equate to higher dispersive mixing, and an increase in this dispersion lowers recovery efficiency (fig. 3). Thus, recovery could be better in a sand aquifer of uniform permeability where dispersion results primarily from flow through intergranular pore spaces, as opposed to a limestone aquifer having diffuse and conduit flow components, particularly if thin zones of high permeability occur within the limestone aquifer.

Loss of injected freshwater could occur if a storage zone is not well confined. Injected water may move upward or downward out of the storage zone, or saline water may move up into the storage zone during recovery.

Recovery efficiency is greater in a thin aquifer than in a thick aquifer because of the lower vertical extent of the transition zone along which mixing occurs. However, this effect can be partially offset by increasing the volume of water recharged during a cycle. Minimizing the thickness of the storage zone within a thick aquifer can also be beneficial depending on the aquifer's distribution of vertical hydraulic conductivity.

Downgradient movement of a bubble of recharged water due to the background hydraulic gradient could reduce recovery efficiency. Based on an estimated gradient at the Hialeah ASR site in the Upper Floridan aquifer, reduction in recovery due to this effect was simulated to be minor for a storage period of 6 months, but not for 5 years (Merritt and others, 1983). The average velocity of ambient flow, referred to as the average linear velocity, is a function of both hydraulic conductivity and porosity as well as the background hydraulic gradient (Freeze and Cherry, 1979).

The structural setting of the storage zone at an ASR site could be important to recovery (Water Resources Solutions, Inc., 1999a). Freshwater recovery at a site located in an area that is structurally high or where the dip is low could be more favorable than in an area that is in a structural depression or where the dip is relatively high due to the tendency of the bubble of recharged water to move updip because of buoyancy forces. This factor is likely to be more important as the contrast in salinity and fluid density increases. Structural deformation may influence storage zone confinement due to fracturing, faulting, or vertical dissolution features.

The location of the storage zone relative to the aquifer may be important. If a storage zone extends over only a portion of an aquifer's thickness, this could negatively affect recovery. Merritt (1985) simulated recovery in a case where the ASR storage zone extended only over the lower part of the important flow zone (zone with high permeability) near the top of the Upper Floridan aquifer. Results indicated that recovery efficiency was virtually unaffected compared to the case with the well open to the full thickness of the zone. However, the low ambient salinity (1,200 to 1,300 mg/L chloride concentration) and the moderate hydraulic conductivity values that were used in the simulation prevented any appreciable buoyancy effects from occurring (effects that could cause vertical flow and mixing to increase).

The volume of injected water affects the recovery efficiency. On a per cycle basis, recovery efficiency generally increases as the total volume of injected water increases (Merritt, 1985). However, the effect is much less beneficial when interlayer dispersion (the transverse dispersion between layers of differing hydraulic conductivity in the aquifer) increases. Interlayer dispersion causes mixing between injected and ambient waters in addition to the mixing in the transition zone.

Figure 4. Simulated improvement of potable water recovery efficiency with successive injection and recovery cycles for a variety of dispersion models. The Upper Floridan aquifer at the Hialeah aquifer storage and recovery site was used in the design of the model (modified from Merritt, 1985). [larger image]

Recovery efficiency increases with repeated cycles. Twelve successive cycles of injection and recovery, with recovery of up to only 250 mg/L chloride concentration for each cycle, were simulated for a variety of longitudinal and transverse dispersivity coefficients. Recovery efficiency improved substantially for all cases with repeated cycles, but the rate of improvement diminished with increasing cycles (fig. 4). Recovery efficiency improves with repeated cycles because much of the recharged water from a previous cycle is left in the aquifer, and during the next cycle, recharged water mixes with water of a lower salinity.

Well plugging can occur during recharge in the Upper Floridan aquifer, reducing the recharge rate and freshwater recovery. This plugging is usually caused by deposition of particulate matter in the injected water or by the formation of a precipitate or sludge caused by reactions that occur at the wellbore face or in the aquifer. One method used to restore formation injectivity is periodic backflushing of the well during the recharge phase. At the Hialeah site, well backflushing produced very fine particles of calcite and an iron compound that had precipitated (Merritt, 1997). Plugging at the Lee County site is attributed to suspended material in the injected water and bacteriological growth at the open borehole face (Fitzpatrick, 1986). Well plugging may affect one flow zone in an open-hole interval more than another, reducing overall recovery. During recovery, the less affected zone contributes most of the flow, and the salinity of water from this zone exceeds the limiting salinity level before all the recoverable freshwater from the plugged zone is obtained.

Various numbers and configurations of multiple storage wells at a site were modeled by Merritt (1985). In that study, the number of wells were varied from one to nine, and the well patterns were varied from a linear array to eight wells in an octagonal pattern with an additional well in the center. Greatest recovery efficiencies were attained in arrays consisting of a central well surrounded by perimeter wells. Though in all cases, the recovery efficiencies for the multiple-well configurations were no better than the single-well case injecting the same total volume as the array of wells. Recovery efficiency could improve, however, when the total volume injected increases as the number of wells injecting at a site increases.

Figure 5. Generalized geology and hydrogeology of Lee, Hendry, and Collier Counties (modified from Reese, 2000). [larger image]

The three principal hydrogeologic units in southern Florida are the surficial, intermediate, and Floridan aquifer systems. These aquifer systems in the western part of the study area (Lee, Hendry, and Collier Counties) are described in figure 5. Water-bearing rocks in the intermediate aquifer system grade or pinch out to the east, and in southeastern Florida the intermediate aquifer system becomes the intermediate confining unit. The Floridan aquifer system consists of the Upper Floridan aquifer, middle confining unit, and Lower Floridan aquifer. Three of the aquifers used for ASR in southern Florida are shown in figure 5; namely, the sandstone and mid-Hawthorn aquifers of the intermediate aquifer system and the Upper Floridan aquifer.

The Upper Floridan aquifer is 500 to 1,200 ft thick in southern Florida (fig. 5; Reese, 1994 and Reese and Memberg, 2000). This aquifer is well confined above by thick units in the Hawthorn Group consisting of clay, marl, silt, or clayey sand; hydraulic head in the aquifer is above land surface. The middle confining unit of the Floridan aquifer system underlies the Upper Floridan aquifer and provides good to leaky confinement. This confining unit consists of micritic limestone (wackstone to mudstone), dense dolomite, and in some areas, beds of gypsum (fig. 5). The upper and lower boundaries of the middle confining unit are difficult to define, but its thickness has been estimated to range from 500 to 800 ft in southwestern Florida.

In southwestern Florida, the Upper Floridan aquifer includes the lower part of the Hawthorn Group, Suwannee Limestone, Ocala Limestone, and in some areas, the upper part of the Avon Park Formation (fig. 5). In southeastern Florida, the Suwannee Limestone and Ocala Limestone are commonly absent (Reese, 2000; Reese and Memberg, 2000). In both eastern and western areas, the top of the Upper Floridan aquifer usually is contained within a basal Hawthorn unit, which is defined by an overlying marker unit composed of micritic limestone or marl (fig. 5; Reese and Memberg, 2000). In some areas along the east coast, the Suwannee Limestone is either interpreted as being absent (Miller, 1986; Reese and Memberg, 2000) or present in the lower part of this basal Hawthorn unit.

The Upper Floridan aquifer generally consists of several thin water-bearing zones of high permeability (flow zones) interlayered with thick zones of much lower permeability. Commonly, only one or two major flow zones provide the bulk of the productive capacity. These flow zones are often less than 20 ft thick each and tend to be in the upper part of the Upper Floridan aquifer, typically at or near the top of the Suwannee Limestone, Ocala Limestone, and Avon Park Formation. Unconformities that formed at the end of the Oligocene and Eocene Epochs are present at these contacts (Miller, 1986), and zones of dissolution occur in association with these unconformities in southern Florida (Meyer, 1989a). In southwestern Florida, the most important flow zone tends to be associated with the top of the Suwannee Limestone, whereas in southeastern Florida it is the top of the Avon Park Formation or, if present, the top of the Ocala Limestone. In both of these areas, the basal Hawthorn unit lies above this contact.

Figure 6. Trace of hydrogeologic section in Palm Beach County (modified from Reese and Memberg, 2000). [larger image]

The basal Hawthorn unit is shown in an east-west hydrogeologic section that extends across Palm Beach County near the southern end of Lake Okeechobee (figs. 6 and 7). This unit is thickest along the coast and thins toward the center of the peninsula. Also shown on the section (fig. 7) are the depths of the saltwater interface in the Floridan aquifer system and a unit composed mostly of dolomite and dolomitic limestone referred to as the dolomite unit (Reese and Memberg, 2000). The saltwater interface (fig. 7) is defined as the depth below which total dissolved solids concentration is greater than 10,000 mg/L.

The dolomite unit of the Floridan aquifer system generally is considered to be within the uppermost permeable unit of the Lower Floridan aquifer in southern Florida (fig. 7; Meyer, 1989a). In some areas of Palm Beach County, however, the top of this unit is as high as 1,200 to 1,300 ft below sea level, as shown (for example) by wells PB-1172 and PB-1173 in figure 7. In these areas, it is uncertain whether all of the dolomite unit would be included in the Lower Floridan aquifer.

Figure 7. Hydrogeologic section extending east-west across Palm Beach County (modified from Reese and Memberg, 2000). [larger image]

The altitude of the basal contact of the Hawthorn Group (same as the base of the basal Hawthorn unit) was mapped for most of southern Florida in three previous studies (Reese, 1994, fig. 6; Reese, 2000, fig. 7; and Reese and Memberg, 2000, fig. 6). Determination of the depth of this contact was primarily based on lithology and gamma-ray geophysical log patterns. As described above, this contact does not necessarily correspond with the top of the Upper Floridan aquifer, but the most important flow zone(s) in the Upper Floridan aquifer is typically associated with the contact. The altitude of this contact varies considerably in southern Florida, ranging from less than 600 ft to greater than 1,200 ft below sea level. Local relief can be as much as several hundred feet, particularly in southwestern Florida.

Complex structure in the Hawthorn Group has been identified in Lee and Hendry Counties along the Caloosahatchee River (Cunningham and others, 2001). The wavy configuration patterns of seismic reflection data show this structure, and these patterns are probably related to karstic collapse of deeper limestone that could be in the Floridan aquifer system.

A number of individuals, private consulting firms, water utilities, and regulatory agencies assisted in this study by providing data and technical input. Maintenance supervisor John Reynolds of the Boynton Beach East WTP and lead operators Guy Bartolotta (Broward County WTP 2A), John Cargill (Fiveash WTP), and Howard Erlick (Springtree WTP) were very helpful in providing information and conducting tours of their sites. Steve Evans, Water Quality Supervisor at the Boynton Beach East WTP, was especially helpful in providing detailed water-quality records of all cycles for the ASR well. Offices of the Underground Injection Control Program of FDEP in West Palm Beach, Ft. Myers, and Tallahassee graciously provided additional ASR technical information and data. Mark Pearce of Water Resource Solutions, Inc., Cape Coral, Fla., provided helpful technical input.