The Floridan Aquifer

Character and distribution — The name “Floridan Aquifer" is commonly applied in Florida to the principal artesian aquifer of the southeastern United States. The aquifer consists most of limestones and dolomites, middle Eocene to middle Miocene in age, which act more or less as a hydrologic unit in most of Florida, in southeastern Georgia, and in parts of Alabama and South Carolina. The aquifer is, however, of variable porosity and permeability and consists in many places of well developed cavernous intervals separated by zones of low permeability which act as confining layers. Thus, the Floridan Aquifer may in places be thought of as a compound aquifer consisting of several subaquifers. It is one of the most extensive limestone aquifers in the United States (Stringfield, 1966, p. 95).

Parker and others (1955, p. 189) who first applied the name “Floridan,” defined the Floridan Aquifer in Florida as being limited to the following sequence: Lake City and Avon Park limestones of middle Eocene age, Ocala Limestone of late Eocene age, Suwannee Limestone of Oligocene age, Tampa Limestone of Miocene age, and permeable parts of the Hawthorn Formation of Miocene age that are in hydraulic contact with the rest of the aquifer. Although the aquifer is usually referred to as the principal artesian aquifer where it occurs in areas other than in Florida, the term “Floridan Aquifer” and the above described stratigraphic boundaries are now generally recognized in Florida, and in this (I-73) report.

Down both the north and south plunge of the Ocala Uplift the aquifer attains a thickness of as much as 1,500 feet (Stringfield, 1966, p. 97), but in most of the Barge Canal area it ranges from about 1,000 feet to 1,200 feet thick (Chen, 1965). The Suwannee and Tampa limestones are missing in the Barge Canal area, so the Floridan Aquifer in this area includes only the Lake City, Avon Park and Ocala limestones, and parts of the Hawthorn Formation. As a source of ground water for the canal, the writer considers the Ocala Limestone the most important stratigraphic unit of the aquifer.

In much of Florida the aquifer is confined by overlying poorly permeable sediments of Miocene age and younger. These sediments are thickest in structurally low areas and thin toward structurally high areas. Only scattered remnants of the once continuous Miocene cover remain on the flanks of the Ocala Uplift. Thus, in the area of the Ocala Uplift much of the Floridan Aquifer is unconfined and, therefore, under water-table conditions [Foundation's emphasis]. As discussed earlier, the aquifer is confined along the route of the Barge Canal northeast of Silver Springs by the thick sequence of poorly permeable deposits underlying the Oklawaha River valley. Near the river artesian flow occurs from wells that penetrate the Floridan Aquifer.

However, southwest of Silver Springs, along the canal route to the Gulf, these relatively impermeable materials are essentially missing, and the aquifer is under water-table conditions. These hydrogeologic factors are of fundamental importance to the design and operation of the Cross-Florida Barge Canal.

Water in the aquifer in the Barge Canal area is derived from direct recharge by local rainfall and from ground-water inflow from potentiometrically high areas immediately to the north and south. Most recharge takes place where the aquifer is unconfined, although some recharge occurs where the poorly permeable overburden is thin or has been breached by sinkholes. Major discharge from the aquifer occurs from Silver and Rainbow Springs, two of the largest fresh-water springs in the United States. A comparison of hydrographs of rainfall, ground-water level, and spring discharge in the canal area demonstrates the close relationships among these three hydrologic parameters (fig. 21 of the I-73 report).

Potentiometric surface of the aquifer — The potentiometric surfacean imaginary surface defined by the level to which water in an aquifer would rise in a well due to the natural pressure in the rocks. of the Floridan Aquifer is that imaginary surface connecting all points of equal altitude to which water will rise in tightly cased wells open to the aquifer, whether or not the aquifer is confined or unconfined [Foundation's emphasis]. Where the aquifer is unconfined, the potentiometric surface is the level at which water will stand in uncased or screened wells. Figure 22 (shown below) is a map of the potentiometric surface of the upper part of the Floridan Aquifer in north central Florida centered in the Barge Canal area. The contours (equipotential levels) are drawn at a 5-foot interval and are based on ground-water levels measured in May 1968, mostly in the upper 50 to 200 feet of the aquifer. Two major potentiometric highs appear on the map, one in Polk County at the south end of the map area, and the other in Clay and Bradford Counties at the north end of the area. These highs are separated by a saddle reflecting the large discharges at Rainbow and Silver Springs. The canal route passes through the area of this potentiometric saddle and within a few miles of both springs.

Ground water moves downgradient from the potentiometric highs toward the saddle and the peninsular coasts along flow paths approximately perpendicular to the equipotential lines. Ground-water drainage basins in the upper part of the aquifer may be outlined on the map by drawing lines along potentiometric divides. Thus, it is possible to demonstrate the spatial relationship of the canal route to the ground-water flow pattern and to delineate the drainage areas of Rainbow and Silver Springs. The mapped potentiometric surfacean imaginary surface defined by the level to which water in an aquifer would rise in a well due to the natural pressure in the rocks. is assumed to represent that part of the aquifer supplying the springs. A knowledge of the size and shape of the drainage areas of the two springs is essential to the quantitative analysis of aquifer characteristics along critical reaches of the canal.

In addition to showing general directions and preferential routes of ground-water flow, the maps of the potentiometric surface help to demonstrate important geohydrologic relationships in the area. For instance, the configuration of the potentiometric surface tends to confirm routes of concentrated groundwater flow suggested by the fracture trace maps, thickness of cover map, and top-of-rock map (figs. 15, 18, and 19 of the I-73 report). Conversely, the geologic maps are useful interpretive tools in estimating the potentiometric surface in areas of sparse water-level control, and for interpreting the anomalies on the potentiometric surfacean imaginary surface defined by the level to which water in an aquifer would rise in a well due to the natural pressure in the rocks.. The foregoing relationships are discussed later in this (I-73) report.

Characteristics of recharge to the aquifer — Local rainfall recharges the Floridan Aquifer in the canal area. Most direct and heaviest recharge is on the flanks of the Ocala Uplift where the Ocala Limestone of Eocene age is at or near land surface (figs. 9, 10, and 14 of the I-73 report). Here recharge is by direct infiltration through the sand and clayey sand which covers much of the limestone and which is as much as several tens of feet thick (fig. 15 of the I-73 report). Where scattered remnants of Miocene-Pliocene(?) sediments of low permeability still cover the limestone on the flanks of the uplift, such as in the Fairfield and Ocala Hills, important recharge is concentrated at sinkholes that penetrate the cover. Also, there is some minor recharge to the aquifer through the poorly permeable material separating the sinks. [See I -73 report p61].

There is direct recharge to the aquifer in those areas west and south of Rainbow Springs where Avon Park Limestone is at or near the surface on the crest of the Ocala Uplift (fig. 13 of the I-73 report). However, the comparatively low permeability of the upper part of the Avon Park, a consequence of dolomitization and presence of sand and clay fill in solution cavities, has caused local potentiometric highs, water levels near the land surface, and some rejection of recharge during very wet periods.

East of Silver Springs, where the confining layer underlies the floor of the valley of the Oklawaha River, the flood plain of the Oklawaha is a ground-water discharge area where wells flow and some upward leakage apparently occurs in association with normal faults and thin parts of the confining layer. [Foundation's emphasis]

Recharge from local rainfall occurs through out the Barge Canal area wherever the materials covering the aquifer permit infiltration. Aside from the principal river flood plains, there is no integrated surface drainage system, and rainfall must either infiltrate the land surface and percolate to the aquifer or be lost to evapotranspiration. Recharge not only occurs in the high slope and ridge areas of the potentiometric surfacean imaginary surface defined by the level to which water in an aquifer would rise in a well due to the natural pressure in the rocks., but also in the lower slope areas near the spring discharge points, provided the aquifer extends to the surface or is covered with permeable materials. As indicated earlier, some of the highest points on the potentiometric surface, such as those local highs southeast and west of Rainbow Springs, actually are areas of low permeability in the aquifer, and although they are recharge areas the rates are low.

Characteristics of water movement through the aquifer — Recharging water moves in essentially vertical paths above the capillary fringe of the unsaturated zone, but after entering the capillary fringe and thence into the zone of saturation the water assumes a horizontal component of movement in a down-gradient direction approximately normal to the contours on the potentiometric surface (fig. 25 of the I-73 report). In a highly permeable aquifer such as the cavernous limestone of the Floridan Aquifer, the horizontal component is usually dominant. As illustrated by flow line segments on Figure 25, individual flow paths converge into gently sloping troughs in the potentiometric surface as the paths approach points of concentrated discharge. The troughs of preferential or concentrated flow in a limestone aquifer, such as the Floridan, may be indicative of solution channel systems which become increasingly well developed as large springs such as Silver and Rainbow Springs are approached.

Without an increase in hydraulic gradient, such a progressive increase in permeability is essential in order that the aquifer be able to accommodate the increasing volume of flow converging on the area of discharge.

Intervening parts of the aquifer that separate zones of preferential flow are represented by gentle ridges on the potentiometric surfacean imaginary surface defined by the level to which water in an aquifer would rise in a well due to the natural pressure in the rocks. (fig. 25of the I-73 report), and transmissivities in these ridge zones may be many times less than those represented by the potentiometric troughs. Flow in the ridge areas tends towards the troughs of preferential flow rather than along independent direct routes to the area of discharge.

Maps of the potentiometric surface are valuable indicators of the two dimensional or plan-view pattern of ground-water flow. The pattern of flow in the Ocala vicinity is illustrated in the Fig. 25 by the flowpath segments plotted on the potentiometric surface. However, the two-dimensional flow-pattern is only a part of the total flow picture, and flow in the third dimension, or vertical plane of flow, may be illustrated by the vertical sections drawn along flow paths plotted on the potentiometric surface map. Figure 28 shows two such sections (X-Y and X-Z) which have been drawn leading to Silver Springs across the canal alignment. The locations of the lines of section are shown on Figures 32 and 34 of the I-73 report.

No actual measurements are available, along the lines of section, of changes in hydraulic potential with depth in the aquifer. If detailed information were available on the vertical head distribution, it would be possible to draw actual paths of flow normal to lines of equal potential. However, through knowledge of other hydrologic and geologic data it is possible to sketch presumed paths of flow based on assumptions drawn from various theoretical analyses of ground-water flow such as discussed by Freeze and Witherspoon (1966 and 1967), Thrailkill (1968), and Toth (1963). [Foundation's emphasis]

A ground-water basin is a three-dimensional closed system that contains all of the flow paths followed by all of the water recharging the basin. Thus, the Silver Springs and Rainbow Springs drainage areas or basins may be described as bounded by vertical flow lines at the potentiometric divides, by a nearly horizontal “impermeable” bottom (possibly the upper part of the Avon Park Limestone), and by an irregular top (the potentiometric surfacean imaginary surface defined by the level to which water in an aquifer would rise in a well due to the natural pressure in the rocks.).

It is the change in ratio of permeability that is needed for depth limitation of the basin rather than the presence of a truly impermeable layer. Also, the presence of comparatively shallow subregional basins, as the Silver and Rainbow Springs basins may be considered, does not preclude movement of ground water in an underlying regional basin not directly involved in the subregional or local recharge- discharge relationships active in the shallow basin area.

It is thought by the writer that upper part of the Avon Park Limestone, although not impermeable in the true sense, is sufficiently less permeable than the overlying Ocala Limestone that it may constitute the bottom to important ground-water basins in the Barge Canal area. The average thickness of the Ocala Limestone in the Ocala-Silver Springs area is estimated to be about 100 feet. Therefore, most of the flow to the Silver Springs may take place through the upper 100 feet or so of the aquifer. If such a situation does exist, it is complicated in places by the normal faulting described earlier. That is, subordinate amounts of flow probably pass through or adjacent to older rocks of the aquifer which are in fault association with the Ocala Limestone.

The data used to map the potentiometric surface in this study are mostly from wells penetrating only the upper part of the aquifer, that is, in most wells to depths less than 75 feet below the top of the aquifer.

Water-level data from deeper zones of the aquifer in the Silver Springs-Ocala vicinity are limited to several deep wells in the city of Ocala. These data indicate no appreciable change in head with depth. Three of the wells, all city of Ocala public-supply wells in the south-central part of town, bottom in the Avon Park Limestone, and are open to the aquifer between about 120 and 350 feet below land surface. A storm drainage well in east Ocala bottoms at 500 feet below land surface at a depth near the top of the Lake City Limestone. An industrial well at the Libby-McNeil citrus plant on the north edge of the city is cased to 850 feet and is completed open hole to 1,083 feet near the bottom of the aquifer. During the drilling of the well in late 1951 and early 1952, static water-level measurements were made and posted on the driller’s log at different depth intervals in the cased zone as well as in the present open hole. No appreciable changes in head with depth were indicated.

Various hydrogeologic factors suggest the regional presence of an appreciable reduction in permeability from the Ocala Limestone to the Avon Park Limestone. Potentiometric-surface maps indicate pronounced highs in areas where the Avon Park Limestone crops out or where the Avon Park is close to the surface, southeast and west of the Rainbow Springs area (figs. 13, 23, and 24 of the I-73 report). The highs are the result of the relatively low permeability of the Avon Park Limestone. The primary permeability of the Avon Park was reduced later by dolomitization and it is thought, still further by sand-filling of solution cavities, some of which cavities may have been developed prior to dolomitization. Not only were effective inter-granular porosity and permeability reduced in the Avon Park Limestone in north-central Florida by recrystallization during dolomitization but, because the dolomite is much less soluble than limestone, development of solution channel porosity and permeability from ground-water circulation also were reduced.

Dewatering operations at the Inglis Lock construction site, pumping tests at the future site of the Dunnellon Lock, and data from observation wells in the vicinity of the Dunnellon Lock site all indicate relatively low permeability in the upper part of the Avon Park Limestone, as compared with the high permeability of the more or less undolomitized Ocala Limestone.

A two-dimensional section through a ground-water basin is representative of the three-dimensional basin if it is taken parallel to the direction of dip of the water-table slope. Freeze and Witherspoon (1966 and 1967) investigated details of steady flow in regional ground-water basins by using digital computer solutions of numerous mathematical models. By machine plotting of head distributions, they were able to determine ground-water flow patterns resulting from the effects of many different types and combinations of types of geologic and hydrologic conditions.

By using their results and that of others in conjunction with a knowledge of certain geologic and hydrologic factors in the study area, sections X-Y and X-Z (fig. 29 of the I-73 report) were prepared as conceptual models of the third dimension of flow in that part of the Silver Springs drainage basin through which the Barge Canal will pass. Flow lines are sketched in the sections in accordance with the kind of head distribution the writer would expect from the set of recharge, discharge, and geologic conditions illustrated in the sections. It was necessary to use a highly exaggerated vertical scale; therefore, the reader is cautioned to take cognizance of some avoidable distortions of the true scale model.

Head distribution is dependent on the locations of recharge and discharge areas. In a recharge area the flow is downward, so that the head decreases with depth. Conversely, in a discharge area flow is upward so the head increases with depth. Between a recharge area and a discharge area the flow is nearly horizontal so that the change in head with depth is slight. Changes in head with depth may be represented in vertical section by equipotential lines. In a recharge area the water table in the upgradient direction of the water table. In a discharge area they slope in the downgradient direction. In intermediate areas of no recharge or discharge, the equipotential lines are approximately perpendicular to the water table. If the water table is perfectly flat, no horizontal component of flow exists, any flow is vertical, equipotential lines are horizontal, and the water table is an equipotential line.

Recharge moves downward through the unsaturated zone along approximately vertical lines, but when it passes into the capillary fringe of the zone of saturation, a horizontal component of movement is applied parallel to the down-gradient direction of the water table. As long as there is additional recharge in the downstream direction there will remain a vertical component of flow resulting in a gradual deepening of flow paths, so that recharge entering at the divide will have to travel deepest in the aquifer to reach the discharge area. Where there is little or no recharge, the equipotential lines will tend to become vertical, and the flow paths horizontal. However, if ground water then flows from under a covered area and enters an area downgradient where recharge occurs, a downward component of flow will again result because of the higher head at shallower depths resulting from the recharge. As flow reaches the discharge area, the flow paths turn upward and the hydraulic head lessens toward the water table.

Section X-Y (fig. 28 of the I-73 report) represents the flow situation just described. Under natural conditions, because of low topography and a very gently sloping potentiometric surfacean imaginary surface defined by the level to which water in an aquifer would rise in a well due to the natural pressure in the rocks., no small subbasins or local basins, as described by Toth (1963), exist along the line of section. However, as shown on the section, a subbasin will exist when the Barge Canal is completed, because water will discharge from the aquifer to the canal. Consequently the canal will be subordinate discharge area within the Silver Springs drainage basin. Along section X-Z (fig. 28) no subbasin exists now, nor will one when the canal is completed. In this area water will discharge from the canal to the aquifer.

As discussed by Toth (1963), theoretically three types of flow systems may occur in a ground-water basin: local, intermediate, and regional. The different types of systems are separated by subhorizontal boundaries, with local systems at the top, underlain in turn by intermediate and regional systems. The local systems are separated from each other by subvertical boundaries resulting mostly from topographic relief. The higher the topographic relief, the more important are the local systems. Since topographic relief is so subdued, and permeability of the aquifer is so great, virtually none of the so-called local basins exist in the Floridan Aquifer in the canal area. However, the Rainbow and Silver Springs ground-water drainage areas may be categorized as intermediate flow systems, the two basins being separated from the deeper regional flow system of the Floridan Aquifer in Peninsular Florida by the upper part of the Avon Park Limestone, which may be treated as the sub-horizontal boundary.

Certain water-quality factors support the premise that most of the flow of both Rainbow and Silver Springs is from near the top of the aquifer, although the more highly mineralized character of Silver Springs water indicates a longer residence time and a somewhat deeper source than that of water issuing from Rainbow Springs. Dissolved solids concentrations in the spring waters compare favorably with ranges of concentrations found in water from wells in the shallow part of the aquifer, and sulfate (SO4) concentrations in the spring waters indicate the upper part of the aquifer is the major source of spring discharge. Like dissolved solids, sulfate increases markedly with depth, probably to a large extent the result of the more common occurrence of gypsum and anhydrite (calcium sulfate) in rocks deeper in the aquifer. In Ocala a concentration of about 260 mg/L of sulfate occurs in water from a well open to the aquifer from 850 to 1,083 feet. Concentrations of about 150 mg/L are found in the city of Ocala public supply wells open to the interval 120 to 350 feet, and concentrations ranging between 30 and 50 mg/L occur in Silver Springs water.

In the Ocala-Silver Springs vicinity an average sulfate concentration of 22 mg/L has been calculated for water samples collected from 18 wells in the upper part of the aquifer. The well waters range in sulfate concentrations from 0.0 to 92 mg/L and the wells range in depth from 40 to 200 feet.

On page 93 of the I-73 report, Mr Faulkner records, "In a heterogeneous and isotropic aquifer the equipotential and stream surfaces are orthogonal (that is, equipotential lines and flow lines intersecting at right angles). In the analyses made the aquifer is assumed artesian and flow steady. It is further assumed that the blah of equipotential and stream surfaces holds true when the aquifer receives recharge vertically."