Summary Introduction Methods Geologic Setting Results Rock Analysis Water Chemistry Ground Water Contamination QC/QA Conclusions Future Studies Acknowledgments References Appendices

Cored wells described in Appendix B show lateral facies changes that were unexpected. Overall facies in the Pleistocene limestone tend to reflect overlying Holocene sediment facies. For example, Holocene sediments are generally arranged in facies belts such as reefs, carbonate sand, and wackestone/mudstone. These facies belts parallel the Keys and platform margin. Landward of the outer reef margin is the 2-mile-wide White Bank carbonate sand belt and farther landward is the 2- to 3-mile-wide mud/wackestone belt known as Hawk Channel. The youngest Pleistocene limestone just beneath tends to parallel these Holocene facies. For example, in KL-5 and OR-5, the Holocene coral reef accumulation is similar to the underlying Pleistocene coral reef accumulation. The only difference is absence of the branching coral Acropora palmata in the Pleistocene. Holocene grainstone and packstone sediments of White Bank generally overlie Pleistocene grainstone, packstone, and wackestone facies, as demonstrated by cores KL-3, the top 12 ft (4 m) of KL-4 and OR-3 and 4. KL-3 was the deepest core drilled (65 ft or 22 m) and coral is conspicuously absent from this well. Corals were also absent from OR 3 and 4. However, these wells are shallow and therefore are not entirely comparable with the KL wells. Support for this observation of facies mimicry between Holocene and Pleistocene was found in wells between the KL and OR transect that were drilled earlier (Shinn et al., 1977; Shinn, 1981). The Lower Keys transect (SB wells) also shows that grainstone/packstones underlie the Hawk Channel area but the transect does not extend seaward far enough to determine if Pleistocene reef facies underlie Holocene reef facies.

Porosity is generally high, at least 45% and higher. Where core recovery is low or nonexistent (see core descriptions), the cause, as discussed previously, is most likely due to extremely high porosity. Porosity of the more resistant zones was measured in the 20% range. Actual porosity is probably higher because the volume-displacement method does not measure fine non-connected pores. Zones of lower porosity, and for the most part lower permeability, are restricted to mud/wackestone facies and the altered zones associated with subaerial unconformities. Significantly, the top 5 ft (1.5 m) of the Pleistocene in all deep wells were less porous and permeable than the underlying 5-ft interval. The top 1 ft (0.3 m) of the first 5-ft interval has even lower porosity and permeability. Porosity and permeability reduction near the unconformity is diagenetic and related to post-depositional diagenesis.

Figure 4. Printout of tide-induced pressure from well KL-1. Bold line is tide fluctuation; narrow line is pressure within the well expressed as the difference between outside and inside pressure. [larger image]

Throughout the remainder of the report, discussion of well transects and water chemistry will be in the following order: Lower Keys, middle Key Largo, and north Key Largo. Tables have also been arranged in that order. This was the sequence in which the wells were drilled, completed and sampled.

Figures 5, 6, and 7 show the locations of the three well transects. A transect was located in the Lower Keys for three reasons: 1) to include typical Pleistocene oolite facies; oolite generally has lower permeability than Key Largo Limestone and, according to Vacher et al. (1992), the permeability difference between oolite and Key Largo Limestone is an order of magnitude; 2) previous studies involving the use of groundwater monitoring wells had been conducted in the Lower Keys oolite (Hanson 1980; Lapointe et al., 1990), and 3) the location of the transect ties in with an ongoing onshore monitoring project initiated by the State DEP office in Marathon, Florida (Rios, pers. commun., 1992). The ongoing DEP study is situated in close proximity to a new RV complex that has a treatment plant and two 90-ft (27 m) injection wells. The four monitoring wells, three of which were placed approximately 100 ft (30 m) from the two injection wells, were established in 1989 before the RV complex and treatment plant went into operation. The fourth well was located approximately 900 ft (275 m) from the disposal wells. All four wells were drilled in oolite but encountered Key Largo facies at ~30 ft (9 m) according to the drillers' logs. The wells are screened at approximately the same interval as our wells SBB-1, 2, and 3 described below. The four onshore monitoring wells along with the six offshore wells drilled for this project will provide an important monitoring network for assessing the effects of future development and possible subsurface-contaminant buildup. We sampled the DEP wells during the initial round of water sampling. Results will be discussed later.

Figure 5. Map of Saddlebunch Keys transect. Location of DEP onshore monitoring wells is shown in inset. SBB wells are in Miami Oolite and SB-1A&B begin in Key Largo Limestone. SB-2 and SB-3 were drilled on Holocene patch reefs, but underlying Pleistocene limestone consist of ooids and skeletal grainstones. [larger image]

Figure 6. Middle Key Largo transect showing two onshore wells and five offshore wells. KL-1 is shown in Figure 2B. Rodriguez Key Bank is a Holocene sediment accumulation. Contours do not show topography of Pleistocene limestone. [larger image]

Figure 7. North Key Largo transect (OR series for Ocean Reef Club, which occupies the north end of Key Largo). Wells ORO-1A&B are located at the south end of airport runway. OR-2 and 3 are shallow wells as shown in Figure 3D. OR-4 and 5 penetrate >20 ft (6m) of Holocene accumulation before penetrating underlying Pleistocene limestone. [larger image]

Well SBB-1, the first well in this transect, is located on oolitic rock in the bay in ~1 ft (30 cm) of water about 600 ft (200 m) from the two closest onshore DEP monitoring wells, MW-1D and 3D (listed as MO-171, and 173, in Table IV). SBB-1 and nearby SBB-2 were both drilled in oolite. The oolite, exposed at the surface, grades downward into skeletal and pelletal grainstone. An unconformity interpreted as the Q3 lies at 24 ft (7.3 m) in SBB-1 and 23 ft (7.5 m) in SBB-2. The well screens are between 31 and 35 ft (9.5-10.7 m) in both wells.

SBB-3 began in oolite that changed to coral facies at 20 ft (6 m). A distinct Q3 soilstone crust was not recovered, although brown calcareous carbonate infilling of voids was encountered at 24 ft (7.3 m), which we believe represents the Q3 horizon. Unlike at SBB- 1 and 2, coral is present below the Q3 unconformity. The Q3 at nearby Big Pine Key has been shown to be Key Largo coralline facies (Schroeder et al., 1958; Hanson, 1980; Kindinger, 1986; Shinn et al., 1989). The well screen in SBB-3 was placed below the unconformity between 31 and 35 ft (see Table I and core description in Appendix B). Lunar tides exceeding a few centimeters were not observed during the drilling of SBB-1, 2 and 3. However, slight tidal pumping was observed in all three wells.

SB-1A and B were drilled ~6 ft (2 m) apart on the ocean side of the Saddlebunch Keys in ~3 ft (1 m) of water. Tidal fluctuations of approximately 3 ft (1 m) were observed during the 4 days required to drill and complete these wells. Tidal pumping at this site was sufficiently strong to prevent quartz sand from entering the well bore while water was flowing out of the well. The rock beginning at the surface in this well is typical Key Largo Limestone and is on trend with the exposed Key Largo Limestone that forms the Upper and Middle Keys (Hoffmeister and Multer, 1968; Perkins, 1977; Shinn et al., 1989). The southernmost exposure of the Key Largo Limestone forms Newfound Harbor Key. Immediately to the north of Newfound Harbor Key, the Key Largo Limestone interfingers with oolite formerly called the Key West Oolite (Sanford, 1909) but renamed the oolite facies of the Miami Limestone (Hoffmeister and Multer, 1968). The Key Largo facies of the Miami Limestone is generally believed to extend seaward to the platform margin, a distance of approximately 5 miles throughout most of the Florida Keys. Our core drilling indicates, however, that the Key Largo facies extends not more than a mile offshore, where it is replaced by grainstone/packstone facies and then reappears as a Key Largo-like coral facies near the platform margin. The Key Largo facies was also encountered in the KL and OR well transects but only near shore and at the platform margin.

Other core drilling (Kindinger, 1986), discussed in Shinn et al. (1989), shows that the Key Largo facies underlies the Lower Keys oolite as far north as the Content Keys area. The Key Largo facies, with large recrystallized massive coral heads comprising the Q3 unit of Perkins (1977), is being quarried from below pelletal and skeletal grainstone at a depth of 25 ft (7.5 m) below sea level near Key West.

Our cores indicate that the Key Largo facies encountered at the surface in SB-1A&B may dip beneath oolite to the north and merge with the coralline limestone beginning at -20 ft (6 m) in SBB-3. The coral facies in SBB-3 probably represents a backreef depositional environment that accumulated during the time the larger Key Largo reef facies was forming.

SB-2 is located in 12 ft (4 m) of water on an unnamed 20-ft-thick (6 m) Holocene coral patch (known to us as Trouble patch) that is situated on Pleistocene limestone, which, interestingly, does not form a topographic high and is not composed of Key Largo facies but contains oolite and grainstone facies (see core description in Appendix B). Trouble patch is surrounded by lime mud in 22 ft (7 m) of water. High-resolution seismic profiles and sediment thickness maps prepared by Lidz et al. (in press) show the surrounding mud is generally >6 ft (2 m) thick. Water visibility here was never observed to exceed 13 ft (4.4 m). The Q3 was not reached because drilling difficulties prevented deeper drilling. The screen is set between 31 and 35 ft (9.5-10.7 m).

SB-3 is located in 15 ft (5 m) of water on a Holocene patch reef surrounded by lime-mud sediments up to 6 ft (2 m) thick. The reef has a navigational marker and is called Nine Foot Shoal. The reef consists of a 15-ft-thick (5 m) Holocene coral accumulation that, similar to SB-2, is not situated over a Pleistocene reef. The section overlies oolitic grainstone and packstone that grade downward into pelletal grainstone. Some coral was encountered at the bottom of the hole at 34 ft (36 m). A calcareous brown soil separates Holocene from Pleistocene. Another brown soilstone crust occurs at 20 ft (6 m) and is underlain by grainstone that was cemented before deposition of the overlying crust, as evidenced by pholad borings (see core log in Appendix). Surprisingly, a layer of quartz sand at least 1 ft (30 cm) thick occurs at 23 ft (7 m), and a brown calcareous soil that may be the Q3 unconformity occurs at 25 ft (7.5 m). The upper crust could be the Q4. The soil and crust beneath the Holocene sediment caps the Q5 unit of Perkins (1977), which is the same age as the oolite in the upper part of SBB-1, 2 and 3. The well screen was placed between 31 and 35 ft (9.5-10.7 m). SB-1A&B and SB-2 and 3 will be discussed later because coliform bacteria (bacteria that responded to the standard coliform test) were consistently detected in water samples from these wells.

The middle Key Largo transect is located off Port Largo and the community of Key Largo (Fig. 6). This, the most populated area on Key Largo, contains numerous Class V disposal wells that range from 40 to 90 ft (12-27 m) in depth. Many disposal wells are located within 3,200 ft (1,000 m) of three of our monitoring wells. Hundreds of septic tank systems are located within 1,600 ft (500 m) of these wells. KLI-2A&B are situated less than 65 ft (20 m) from at least two septic-tank systems. Two new 90-ft disposal wells were installed nearby while this study was in progress.

The Port Largo housing development centers on an elaborate system of artificial canals cut to a depth of approximately 25 ft (7.5 m). The canals incise classic (type section) extremely porous and permeable Key Largo Limestone. The canal-based port is protected on the ocean side by a linear artificial strip of land that in the 1970s served as a landing strip for small planes. The strip, now called Ocean Key, is a growing, upscale development with individual septic-tank systems. A monitoring well (KL-1) was installed immediately seaward of Ocean Key.

The Key Largo well transect consists of two onshore locations (two wells at one location and a multi-completion well at the other) and five sites offshore. Two of the wells (KLI-1 A& B) were drilled near the center (highest elevation) of the island on private property. The KLI-1A well penetrated 45 ft and KLI-1B, 2 ft (60 cm) away, is only 20 ft (6.5 m) deep. The screen in the shallow well spans between 16 and 20 ft (5-6 m) of depth. The bottom of the screen is approximately 6 ft (2 m) below the top of the water table, which was 14 ft (4.2 m) below the surface when the well was drilled. The rock consisted of Key Largo Limestone. The Q3 unconformity in KLI-1 occurs at 32.5 ft (10 m). The screen was placed below the unconformity with the top of the screen at 35 ft (10.5 m).

The other onshore well, KLI 2, is located at the NOAA/NURP facility situated on the south side of the main commercial canal that runs along the north side of the Port Largo canal system. KLI-2 is a multi-completion well located 60 ft (18 m) from the edge of the canal. The deep part of the well is KLI-2A and the shallow part is KLI-2B. The well penetrated 45 ft (14 m) and the Q3 unconformity was located at 35 ft (10.7 m). The bottom of the hole was backfilled with quartz sand in order to place the screen so its top was just below the Q3 unconformity. KLI-2B, completed in the same well bore, penetrates the upper 5 ft (1.5 m) of the water table. Water level in this well fluctuates synchronously with tide levels in the Port Largo canal. The upper 3 ft (1 m) is artificial limestone fill and the remainder of the well penetrates typical Key Largo Limestone down to 16 ft (5 m), where it changes to a chalky skeletal grainstone that continues to 32 ft (10 m) and which then merges with cemented quartz sandstone immediately above the Q3 unconformity.

The remaining five wells in the transect are under water. The landwardmost well is situated in 3 ft (1 m) of water on exposed Key Largo Limestone approximately 100 ft (35 m) off the artificial island of Ocean Key. The well (KL-1) is 40 ft (12 m) deep and penetrates typical Key Largo Limestone down to 17 ft (5 m), where it changes to skeletal grainstone with scattered Acropora cervicornis rubble. The Q3 unconformity is 35 ft (10.7 m) below the surface and is overlain by 6 ft (2 m) of calcite-cemented cross-bedded quartz and carbonate sandstone. The sandstone contains mollusks, Halimeda grains and scattered fragments of Acropora cervicornis. Similar sandstone was encountered above the Q3 unconformity in wells drilled on Key Largo by Harrison et al. (1984). The top of the well screen was set below the unconformity in chalky skeletal grainstone. Tidal pumping was strong in this well.

The next well seaward (KL-2) forms a south-bending dogleg in the transect (Fig. 6). The well is located off the northeast side of Rodriguez Key, a Holocene sediment bank (Turmel and Swanson, 1977) on a broad Pleistocene high of Key Largo Limestone surrounded by muddy lime sediment. The rock high is populated by scattered small coral heads and alcynarians (sea whips) but has numerous areas of bare rock. The well location is in 6 ft (2 m) of water and penetrates 40 ft (12 m) of Key Largo Limestone. The entire section down to 35 ft (10.7 m) consists predominantly of coral. Because of chalkiness and high porosity of the rock, there was no core recovered between 35 and 40 ft. If corals had been present in this interval, they would probably have been recovered. The Q3 unconformity was not recovered so the screen was set in the bottom of the hole where it receives water from the 36- to 40-ft interval. Again, tidal pumping was pronounced and the well could only be completed when the tide was rising and water was flowing into the wellbore.

KL-3 was drilled on a soft-sediment bottom. The Holocene sediment consisted of packstone to wackstone facies and was 14 ft (4.5 m) thick). The location is in 14 ft of water approximately 50 ft (15 m) seaward of a small coral patch. The well is located in Hawk Channel, a backreef lagoonal area that parallels the Florida Keys from Miami to Key West. Its position on the shelf is similar to that of SB-3 discussed earlier and OR-3 discussed later. This was the deepest well in the study (65 ft, or 22 m). Typical Key Largo Limestone was not encountered in this well. The entire section consisted of chalky, poorly cemented, skeletal packstone similar to the overlying uncemented Holocene sediment. The Q3 unconformity was not present. The well screen was placed at the bottom of the well between 61 and 65 ft (21-22 m). Tidal pumping was detected but was weak compared to that encountered at KL-1 and KL-2.

KL-4 was drilled on carbonate-sand bottom adjacent to a low-relief coral patch on a sediment facies belt called White Bank. Water depth at KL-4 was 15 ft (5 m). White Bank is a Holocene backreef sediment accumulation composed mainly of grainstone facies. White bank, in places, is as shallow as 4 ft (1.2 m) and its topography forms the seaward margin of the relatively deeper Hawk Channel. White Bank is accreting landward; thus, much of the grainstone facies overlies muddy packstone and wackestone facies like that presently being deposited in Hawk Channel (Enos, 1977). Holocene sediment on White Bank is as much as 25 ft (8 m) thick (Enos, 1977). At this well site, 18 ft (6 m) of Holocene sediment were penetrated before the top of the Pleistocene was encountered. The well penetrated 51 ft (15.5 m, including Holocene sediment). The top of the Pleistocene consists of chalky skeletal grainstone. Corals become abundant at 34 ft (10 m) and continue to the bottom of the well. The Q3 unconformity was not encountered. The well screen was placed between 46 and 50 ft (14-15 m). Tidal pumping was not detected during drilling.

KL-5 was drilled on an unnamed dead reef area between Molasses and Pickles Reef to the south. The location is in 16 ft (5 m) of water on coral rubble. The drill penetrated 16 ft (5 m) of Holocene coral rubble and submarine-cemented grainstone. The top of the Pleistocene is capped with a brown soilstone crust and, after 1 to 2 ft (0.3-0.6 m) of grainstone, changes to coralline Key Largo Limestone. The Q3 unconformity was not recognized in this well. The well screen was placed at the bottom of the well between 56 and 60 ft (17-18m). Tidal pumping was not detected during drilling. The site is bathed in clear oceanic water, often with visibility exceeding 75 ft (24 m).

The upper Key Largo transect was the last transect to be drilled. The wells were drilled one month after passage of Hurricane Andrew, which interrupted the study. Well-emplacement strategy here was modified for two reasons: 1) because of knowledge gained during drilling of the previous wells, and 2) because the disposal wells in this area are old and do not penetrate the Q3 unconformity. The offshore wells, except for OR-1A, were completed in the top 5 to 6 ft (1.5-1.8 m) of the Pleistocene limestone.

A single multi-completion well was drilled onshore within approximately 1,500 ft (457 m) of the 50-well disposal-well field and near the south end of the single runway belonging to the Ocean Reef Club airport. The deep well (ORO-1A) is 40 ft (12 m) deep and consists of typical Key Largo Limestone down to 22 ft (6.7 m). From 22 ft (6.7 m) down to the Q3 unconformity at 36 ft (11 m), the Pleistocene consists mainly of skeletal packstone with scattered fragments of the massive coral Montastrea sp. Some quartz sand was recovered above the unconformity. The top of the screen was placed just below the unconformity. ORO-1B was completed in the same well bore but without a screen. The bottom of the PVC pipe is 6 ft (1.8 m) below the surface. The water table fluctuates with tides and rainfall and lies about 3 ft (1 m) below ground level.

The first offshore wells in the transect, OR-1A&B, are located approximately 500 ft (150 m) offshore on exposed Key Largo Limestone in about 2 ft (60 cm) of water. The 2 wells are 6 ft (2 m) apart. The deep well, A, is 40 ft (12 m) deep and was drilled to determine the stratigraphy and location of the Q3 unconformity. The shallow well, B, penetrates 10 ft (5 m) and the screen is between 6 and 10 ft (1.8-3.0 m) below the surface of the rock. The Q3 unconformity was located in OR-1A at 35.5 ft (10.7 m) and the screen was placed below the unconformity. The entire section consists of typical Key Largo Limestone. Quartz sand was not detected above the unconformity. Tidal pumping was observed in both wells but no attempt was made to quantify differences.

OR-2 is located in 15 ft (5 m) of water near the center of Hawk Channel where there is 8 ft (2.8 m) of silty lime mud (packstone/wackstone facies) overlying approximately 3 ft (1 m) of sticky lime mud, which in turn overlies chalky, skeletal Pleistocene limestone. The limestone has a thin soilstone crust. The well penetrated 7 ft (2.1 m) into the Pleistocene. The screen was placed between 11 and 15 ft (3.4 - 4.6 m) and quartz sand was used to pack the annulus around the screen. The overlying lime mud was allowed to slump in and seal around the PVC pipe. Portland cement was not poured into the annulus above the sand pack. Tidal pumping was strong and the quartz sand could not be poured down the well bore until the tide changed.

OR-3 was completed in the eastern side of Hawk Channel in 16 ft (5 m) of water. The sediment at this site is similar to that at OR-2 and is 7 ft thick (2.1 m). The core was drilled 6 ft (1.8 m) into the underlying limestone. The Pleistocene consists of chalky grainstone and has a soilstone cap. The rock is the same as at OR-2. The well was completed in the same manner as OR-2. Tidal pumping was readily apparent.

OR-4 was drilled in 16 ft (5 m) of water on medium to coarse rippled carbonate sand. The location, like that of KL-4 to the south, is on White Bank. The Holocene sediment was 26 ft (8 m) thick and the lower 10 to 15 ft (3.0 - 4.8 m) consisted of lime mud. Peat was encountered just above the Pleistocene limestone. Although the core barrel did not recover the sediment or peat, fragments of black peat and gray lime mud circulated to the surface and blanketed the sandy bottom during drilling. The well was drilled into the underlying limestone to a depth of 35 ft (10.7 m). The screened interval of the well is between 31 and 35 ft (9.5 and 10.7 m). The Pleistocene consists entirely of chalky skeletal grainstone similar to the sediment on the surface at this site. Tidal pumping was not noticed during drilling.

OR-5 was drilled in 17 ft (6 m) of water on a small reef north of Carysfort Reef informally known as Carysfort North. There are two permanent Sanctuary mooring buoys (CN1 and 2) on this reef. The well was drilled on hard coral-rock bottom approximately 60 ft (20 m) northwest of buoy CN2. The well penetrated 26 ft (8 m) of partially cemented Holocene coral reef. A brown soilstone crust separates the Holocene rock from the Pleistocene. The top 5 ft (1.5 m) of Pleistocene consists of skeletal grainstone The last 2 ft (0.6 m) contain coral. Total depth of the well is 35 ft (10.7 m). The screened interval is between 31 and 35 ft (9.5-10.7 m). The well was completed with quartz sand and Portland cement. The completed well is shown in Figure 2C. Tidal pumping was not noticed during drilling.