Home Abstract Introduction Hydrogeology of S. Florida Ground-Water Movement Subsurface Storage - Oil Field Brine - Liquid Wastes > Freshwater Storage Summary and Conclusions References PDF Version

Subsurface Storage in the Floridan Aquifer System in Southern Florida: Freshwater Storage

Subsurface storage of freshwater as an alternative to surface storage has become increasingly attractive to water managers in southern Florida as urbanization and population growth have placed increasing demands on the water supply. The advantages of the subsurface storage concept are that subsurface space is free, water loss by evapotranspiration is nonexistent, and the site may be located at the point of greatest need (provided hydrogeologic conditions are favorable). The concept is particularly desirable in southern Florida, where real estate has become very expensive, the availability of water is seasonal, and underlying artesian aquifers in the intermediate (Miocene) aquifer system and Floridan aquifer system contain nonpotable saline ground water.

The source of freshwater for injection might be surplus within the surface-water storage system or the surficial aquifer system during the annual wet season. On an annual basis, the surplus freshwater would be injected through Class V wells into suitable artesian aquifers during the wet season, stored for a short period (perhaps 3 to 6 months (mo)), and then withdrawn as needed during the dry season-hence the term cyclic injection-storage-recovery. The measure of success of a cycle is recovery efficiency, which is defined as the volume of freshwater recovered before it fails to meet an established (or a prescribed) chemical standard, expressed as a percentage of the volume of freshwater that was injected. Pilot studies to date in southern Florida have assumed the standard established by EPA for chloride concentration (250 mg/L) in public water supply. Other criteria may be used depending on the particular use of the recovered water. For example, a higher chloride standard could be used if the recovered water were mixed with surface water to yield a blend that would meet drinking water standards.

Theoretical and pilot-operational studies to date indicate that recovery efficiency usually improves with successive cycles, provided that recovery ceases when the recovered water reaches the standard and that the storage period is sufficiently short to prevent significant migration of the injectant away from the point of recovery.

Pilot studies have been conducted at four sites (fig. 28) in southern Florida, with varying degrees of success (Merritt and others, 1983; Wedderburn and Knapp, 1983). Also, data on the recovery of injected wastewater (freshwater) from Class I injection wells during repairs, testing, and abandonment have yielded valuable information on recovery efficiency (McKenzie and Irwin, 1984). Aspects of the existing pilot studies are summarized in tables 18 through 21. Of the four studies, three (sites 24, 25, and 27) involved injection into water-bearing zones of the Upper Floridan aquifer and one (site 26) involved injection into water-bearing zones of the intermediate aquifer system. Plugging of the wellbore by suspended solids in the injectant was a significant problem in all four studies.

At site 24 in Palm Beach County (table 18), injection was chiefly into water-bearing zones of the Ocala Limestone and Avon Park Formation (units of the Upper Floridan aquifer). The study involved four injection-storage-recovery cycles (J.J. Plappert, Florida Department of Environmental Regulation, written commun., 1977). Recovery efficiencies ranged from 0 to 35.2 percent. The transmissivity of the injection zone(s) is probably on the order of 10,000 to 20,000 ft²/d, although data are lacking to support that assumption. The injection zones are apparently associated with zones of dissolution at or near unconformities that separate formations.

Table 18. Results of Florida Department of Natural Resources and Florida Department of Environmental Regulation injection, storage, and recovery tests of freshwater at site 24 in Palm Beach County, 1975-76 [Quantities in million gallons and rates in gallons per minute. From J.J. Plappert, Florida Department of Environmental Regulation, written commun., 1977. Location of site 24 shown in fig. 28]
Test	Cycle 1	Cycle 2	Cycle 3	Cycle 4
Quantity injected	20.5	100	306	102
Storage period (days)	15	30	30	120
Quantity of potable water recovered1	0	4.7	55.5	36.1
Percent of recovery	0	4.7	18.0	35.2
Injection rate	2,000	2,000	2,000	2,000
Withdrawal rate	1,000	1,000	1,000	1,000
Transmissivity and storage coefficient: Unknown Injected water chloride concentration: 65 mg/L (milligrams per liter) Resident water chloride concentration: 1,980 mg/L Open hole: 990 to 1,280 feet (Ocala Limestone and Avon Park Formation)
1Recovery was terminated when the chloride concentration of the recovered water reached 250 mg/L.

At site 25 in Dade County (table 19), injection of freshwater chiefly was into water-bearing zones of the Suwannee Limestone although the injection well tapped parts of the Tampa Limestone and Avon Park Formation (all units of the Upper Floridan aquifer). The study involved three injection-storage-recovery cycles. Recovery efficiencies ranged from 32.9 to 47.8 percent. A decline in efficiency was recorded for the third cycle which probably was related to migration of the injectant downgradient from the injection recovery well during the 181 days (d) of storage. The transmissivity of the injection zone(s) is 10,950 ft²/d. The results of the tests at this site were the basis for theoretical studies that used a mathematical model to evaluate the effects of varying aquifer characteristics, fluid density, regional flow, well arrays, and operating schedules on recovery efficiency (Merritt, 1985).

Table 19. Results of U.S. Geological Survey injection, storage, and recovery tests of freshwater at site 25 in Dade County, 1975-80 [Quantities in million gallons and rates in gallons per minute. Location of site 25 shown in fig. 28]
Test	Cycle 1	Cycle 2	Cycle 3
Quantity injected	41.9	85	208
Storage period (days)	2	54	181
Quantity of potable water recovered	13.8	40.7	80.1
Percent of recovery	32.9	47.8	38.5
Injection rate	1440-780	854	800
Withdrawal rate2	330	494	450
Transmissivity3: 10,950 ft2/d (feet squared per day) Storage coefficient3: 8.4 X 10-5 Injected water chloride concentration: 65mg/L (milligrams per liter) Resident water chloride concentration4: 1,200 mg/L Open hole: 955 to 1,105 feet (Tampa Limestone, Suwannee Limestone, and Avon Park Formation).
1Progressive decline due to wellbore plugging. 2Natural artesian flow. 3Estimated by computer simulation of well G-3062 pumping test. 4Multilevel composite, range from 800 to 2,000 mg/L.

At site 26 in Lee County (table 20), injection of freshwater was into water-bearing zones in limestone of the Hawthorn Formation (unit of the intermediate aquifer system). The study involved three injection-storage-recovery cycles. Recovery efficiencies ranged from 9.7 to 38.7 percent. The efficiency of the first cycle, which had the greatest efficiency value, is probably not representative of the true efficiency because of the small amount injected and the short storage period. The value for the third cycle (30.4 percent) probably represents the efficiency and the storage capability of the aquifer. The transmissivity of the injection zone(s) is about 750 ft²/d.

Table 20. Results of U.S. Geological Survey injection, storage, and recovery tests of freshwater at site 26 in Lee County, 1980-82 [Quantities in million gallons and rates in gallons per minute. From Fitzpatrick, 1985. Location of site 26 shown in fig. 28]
Test	Cycle 1	Cycle 2	Cycle 3
Quantity injected	0.571	16.831	229.026
Storage period (days)	0	47	99
Quantity of potable water recovered	.221	.663	8.819
Percent of recovery	38.7	39.7	30.4
Injection rate	170-350	300	300
Withdrawal rate4	95-110	165-175	150
Transmissivity: 700-800 ft2/d (feet squared per day) Storage coefficient: About 1 x 10-4 Injected water chloride concentration: Cycle 1, 60mg/L (milligrams per liter); cycle 2, 150 to 350 mg/L5; cycle 3, 80 to 100 mg/L (finished water) and 60 mg/L (raw water). Resident water chloride concentration: 550 mg/L Open hole: 447 to 600 feet (limestone of the Hawthorn Formation)
1Estimated after loss of water due to equipment failure. 28.548 x 104 gallons of finished water, followed by 20.873 x 106 gallons of raw river water. 3Low due to relatively high chloride concentration of injected water. Purpose was to test well after acidification. 4Natural artesian flow; improvement in cycle 2 due to acidification of well. 5Abnormally high due to record low flows in Caloosahatchee River source; decreased during injection.

At site 27 in St. Lucie County (table 21), injection of freshwater chiefly was into water-bearing zones of the Ocala Limestone and the Avon Park Formation of the Upper Floridan aquifer. The water-bearing zones are associated with zones of dissolution near formation contacts. The study involved one injection-storage-recovery cycle, for which the recovery efficiency was only 3 percent. The low efficiency was due to the high chloride concentration (200 mg/L) of the injectant. The 3-percent recovery efficiency represented a 79-percent blend of the injectant (chloride concentration of 200 mg/L) with native ground water (chloride concentration of 1,000 mg/L). A recovery efficiency of 33 percent would have been realized had the chloride concentration of the injectant been 50 mg/L (based on the indicated rate of mixing and the limit of 250 mg/L for chloride in drinking water). The transmissivity of the injection zone(s) is about 6,000 ft²/d.

Table 21. Results of South Florida Water Management District injection, storage, and recovery tests of freshwater at site 27 in St. Lucie County, 1982-83 [Quantities in million gallons and rates in gallons per minute. From Wedderburn and Knapp, 1983. Location of site 27 shown in fig. 28]
Test	Cycle
Quantity injected	1.488
Storage period (days)	37.5
Quantity of potable water recovered	.041
Percent of recovery1	2.76
Injection rate	331
Withdrawal rate	140-190
Transmissivity: 6,000 ft2/d (feet squared per day) Storage coefficient: 1.6 x 10-4 Injected water chloride concentration: 200 mg/L (milligrams per liter) Resident water chloride concentration: 1,000 mg/L Open hole: 600 to 775 feet (limestone of the Hawthorn Formation, Ocala Limestone, and Avon Park Formation).
1Percent of recovery would have been 33 percent if the chloride concentration of the injected water were 50 mg/L.

During 1975-77, the U.S. Geological Survey, in cooperation with the Florida Department of Environmental Regulation, conducted a study of the quality of recovered secondary-treated wastewater from subsurface storage in the Upper Floridan aquifer at site 14 in Broward County (table 22). The injection system consisted of two wells that were in operation from 1959 to 1975. Injection ceased in January 1975 when the plant's function was transferred to the Broward County North Regional wastewater treatment plant. Recovery of the injected treated wastewater began in April 1975 and ended in March 1977, when the chloride concentration reached 250 mg/L. The recovery efficiency based on reaching a chloride concentration of 250 mg/L was only 2 percent, which was much less than expected for the great volume (3.4 Ggal) that was injected during the 16 yr of operation. The transmissivity of the injection zone was not determined but was probably greater than for previously discussed freshwater storage pilot studies. Records of the construction of one injection well suggest that injection occurred to a greater depth (perhaps as deep as 1,600 ft) than previously reported. The low recovery efficiency is probably a result of higher aquifer transmissivity, higher chloride concentration (and, hence, higher density) of the resident water, and construction problems. As with the previous pilot studies, plugging of the wellbore by suspended solids was a significant problem.

Table 22. Results of U.S. Geological Survey wastewater recovery tests of freshwater at site 14 in Broward County, 1975-77 [Quantities in million gallons and rates in gallons per minute. From McKenzie and Irwin, 1984. Location of site 14 shown in fig. 28]
Test	Cycle
Quantity injected	3,401
Storage period (years)	16
Quantity of potable water recovered	69.2
Percent of recovery	2
Injection rate	400
Withdrawal rate	4-132
Transmissivity and storage coefficient: Not determined Injected water chloride concentration: 84 mg/L (milligrams per liter) Resident water chloride concentration: 2,360 mg/L Open hole: 995 to about 1,250 feet (Avon Park Formation)

Unpublished data collected by the Florida Department of Environmental Regulation on the amount of freshwater effluent from abandoned injection wells at sites 16 and 17 in Dade County suggest that recovery efficiency for wells that tap the Boulder Zone of the Lower Floridan aquifer is virtually nonexistent. The injection well at site 16 was abandoned in 1983 after 13 yr of operation and after 18.4 Ggal of effluent were injected; the injection well at site 17, also abandoned in 1983, was operated for 11 yr, during which time 7.8 Ggal of effluent were injected. At both sites, the chloride concentration of the injectant was about 60 mg/L, and the chloride concentration of the resident water was about 19,200 mg/L. For both sites, the amount of effluent recovered before the chloride concentration exceeded 250 mg/L did not exceed 1 Mgal. The recovery tests indicate that there is no potential for recovering freshwater stored in the highly transmissive Boulder Zone.

Dissolution zones at erosional unconformities between the Suwannee and Ocala Limestones and the Avon Park Formation probably offer the best opportunity for large-scale storage of freshwater in the subsurface of southern Florida. Detailed maps of the dissolution zones are unavailable, but maps showing the configuration of the top of the middle and upper Eocene rocks are shown in chapter B (Miller, 1986, pls. 6, 8) of this professional paper. The surface is irregular and shows the effects of large-scale erosion at the close of the Eocene Epoch. Erosion removed the Ocala Limestone from much of southeastern Florida and exposed the underlying (older) Avon Park Formation. Zones of dissolution are prominent near this erosion surface; therefore, the maps in Miller (1986) may be used to estimate the depth at which favorable injection zones may be present.