CARBONATE EQUILIBRIA AND RADIOCARBON DISTRIBUTION RELATED TO GROUNDWATER FLOW IN THE FLORLDAN LIMESTONE AQUIFER, U.S.A.W]

Bruce B. HANSHAW, William BACK and Meyer RUBIN U. S. Geological Survey, Washington D. C., U. S. A.

SUMMARY

The principal artesian aquifer of Florida is composed predominantly of Tertiary limestone with lesser amounts of dolomite and gypsum. A north-south geohydrologic section through the piezometric high depicts the spatial and temporal changes of chemical character of water. The major area of recharge is characterized by calcium bicarbonate type water, by low concentrations of total dissolved solids and of sulfate ions, by undersaturation with respect to calcite and dolomite, and by young CI4 ages. Concentration of dissolved solids and sulfate increases markedly as a function of length of flow path and residence time in the aquifer. Radiocarbon ages indicate that within different parts of the system velocities range from 2 to 12 meters per year.

Away from the recharge area the water is supersaturated with respect to calcite and dolomite. Throughout most of the area the water has not reached equilibrium with aragonite and gypsum. Supersaturation of the water with respect to calcite tends to buffer the bicarbonate concentration. This study shows that the geochemistry of the water conforms well with the hydrologic history of the system.

RESUME

L’équilibre des carbonates et la répartition dir radiocarborie par rapport à l’e‘coulement de la nappe aquifère calcaire de Floride, aiix Etats-Unis d’dmbique

Le principal aquifère artésien de Floride se compose surtout de calcaires tertiaires auxquels se melent, en quantités moindres, des dolomies et du gypse. Une coupe géohydrologique nord-sud suivant le maximum piézométrique fait apparaître les changements dans l’espace et dans le temps du caractère chimique de l’eau. Dans l’aire principale d’alimentation de la nappe, l’eau est caractérisée par son contenu en bicarbonate de calcium, par de faibles concentrations en corps solides dissous et en ions sulfates, par une sous-saturation en dolomie et en calcite et par des carbones 14 d’âge récent. La concentration en solides dissous et en sulfate s’accroît notablement avec l’accroissement de la distance parcourue par les eaux et de la durée de séjour dans la nappe. L’âge du radiocarbone indique que, dans différentes parties du système, la vitesse d’écoulement est de l’ordre de 2 à 12 mètres par an.

Loin de l’aire d’alimentation, l’eau est sursaturée en dolomie et en calcite. Dans la plus grande partie de l’aire, l’eau n’a pas atteint son équilibre en aragonite et en gypse. La sursaturation de l’eau en calcite tend à jouer le rôle de tampon l’égard de la concentration en bicarbonate. Cette étude montre que la goéchimie de l’eau suit l’histoire hydrologique du système.

INTRODUCTION

The purpose of this paper is to describe the application of geochemical principles to a reasonably well known hydrologic environment to determine which solid phases play the most important part in controlling the chemistry of water as it moves through the aquifer system. Use of basic principles of physical chemistry show how chemical equations express solution and precipitation of common carbonate and evaporite minerals. A chemical equilibrium model aids in understanding geologic processes such as formation of secondary permeability by solution, precipitation of minerals in the aquifer, cementation, and recrystallization. Isotopic investigations may be used in the study of a hydrologic system, to determine the various origins of water, residence time

(l) Publication authorized by Director, U. S. Geological Survey.

60 1

of water, velocity of groundwater movement and rates of chemical reactions. A combined study of general water chemistry, mineral saturation of the water, and distribution of radiocarbon isotopes is helpful to identify the principal areas of recharge and to identify which geochemical processes control the chemical character of water. This approach is also useful in predicting which, if any, chemical changes may occur as a result of imposing such stresses as increased pumping or artificial recharge on the hydrologic system.

The regional flow pattern in the hydrologic system in central Florida is controlled largely by the two piezometric highs shown in figure 1. A north-south line of wells through the major piezometric high was chosen for the study of changes in aqueous

Fig. 1 - Piezometric map of the principal artesian aquifer of central Florida showing chemistry as the water flows from a recharge area toward deeper parts of the aquifer. The principal artesian aquifer of Florida consists of a series of limestones of Tertiary age which contain minor amounts of dolomite, disseminated quartz sand, gypsum, and anhydrite. The limestones are interconnected and function as a hydrologic unit. The eight wells sampled during this study (fig. i) range in depth from approximately 80 to 300 metres and are completed in limestone that ranges in age from middle Eocene to Miocene. The principal artesian aquifer is overlain by confining beds of Miocene age, chiefly clay of the Hawthorn Formation (Pride, Meyer, and Cherry, 1961; Le Grand and Stringfield, 1965). The major piezometric high has the shape of a north- south elongated dome; water flows down gradient at right angles to the piezometric contours.

location of sample points used on cross-sections (after Stringfield, 1936).

602

This investigation is part of a continuing study of the chemistry of the principal artesian limestone aquifer of Florida. The complete study includes about 50 wells covering the centre of the State, an area of roughly a half million square kilometres. The final report will include mineral equilibrium studies of solid phases in addition to those presented here and investigations of the isotopes of hydrogen, carbon, oxygen and sulphur.

CHEMICAL CHARACTER OF WATER

The chemical character of the water reflects the combined effect of chemical activity between water and limestone and the flow pattern within the aquifer. Within the study area most of the water (Wildwood, Groveland, Polk City and Fort Meade)

CATIONS ANIONS

Fig. 2 - Chemical composition of ground water from the principal artesian aquifer of central Florida.

is of the CaHCOs type; the magnesium and sulphate content increases downgradient so that a mixed type of water is produced by the wells at Ocala and Arcadia, and a Cas04 type water is produced by the well at Wauchula. The well at Cleveland produces water of NaCI type. The chemical character of water in the area of study is shown in table 1 and is represented in figure 2 as a trilinear plot of the percentages of each major constituent to the total anions or cations in milliequivalents per litre.

The Polk City well is close to the highest part of the piezometric surface (fig. 1) and water in this part of the aquifer flows radially downgradient. The area near Polk

603

City is a principal area of recharge, and the Groveland and Wildwood areas also contribute substantial quantities of water to that portion of the aquifer system north and westward of the Polk County high (Stringfield, 1936, p. 151). Between the area of recharge and Ocala, the principal change in composition is in the anion facies, which changes from predominantly HC03 to a mixture that is predominantly S04+HC03 (table 1). A simple progressive change in type of water occurs as water moves from Polk

City toward the Cleveland area, indicating only minor recharge south of Polk City. Between Polk City and Arcadia there is a progressive increase in the Mg:Ca ratio in the cations and a similar increase of S04:HC03 ratio in the anions. A major change in aqueous chemistry occurs between Arcadia and Cleveland. At Arcadia the hydro- chemical facies is Mg+Ca, so4 whereas water produced by the Cleveland well is a NaCl type water.

Similar patterns of chemical change away from the principal recharge area are shown by the constituents which are discussed in the following paragraphs, and by the isotope measurements and mineral saturation studies which are discussed in the two following sections. The TDS (total dissolved solids) content of the water in the study area increases north and south away from the area of recharge, and in particular, from the area of recharge at Polk City (fig. 3 4 . The Polk City-Wildwood area is characterized by water with less than 200 mg/l TDS. South of the Polk City recharge area, water from the aquifer exhibits a progressive increase in TDS to a maximum of 1600 mg/l at Cleveland. Northward, the TDS increases to 420 mg/l. A similar pattern is observed for the concentration of sulphate ions in the ground-

water (fig. 3b). In the Polk City-Wildwood area the sulphate is less than 10 mg/i and increases to 148 mg/l at Ocala. South of the Polk City area the sulphate content increases progressively as far as Arcadia and decreases between Arcadia and Cleveland.

Similar reversals in other chemical parametres occur between Arcadia and Cleveland. These may be due in part to mixing of waters from different geohydrologic environments, as reflected in the change in hydrochemical facies from Mg+Ca, so4 to NaCl. South of the Cleveland area, water from the principal artesian aquifer becomes increasingly more saline and approaches the concentration of ocean water. It is possible also that water produced by the well at Cleveland includes a significant amount of water from the Hawthorn Formation because the well is not cased through the Hawthorn Formation. The quality of water in the Hawthorn Formation does not affect the well at Polk City because the well is near the piezometric high and the limestone aquifer is recharged only by percolation downward through the Hawthorn Formation, whose water has essentially the same chemical character as that of the Floridan aquifer. However, at Cleveland, water from the Hawthorn is probably quite different from water in the Floridan aquifer and from water in the Hawthorn at Polk City.

The bicarbonate content of water from the aquifer is relatively uniform throughout the area of study (fig. 4a). In the Polk City-Wildwood area the bicarbonate content is somewhat lower than elsewhere, indicating that this is an area of major recharge. The uniform bicarbonate content reflects a buffering mechanism by the solid carbonate phases of the aquifer rocks.

To summarize the chemical data presented so far, an area of significant recharge between Polk City and Wildwood is identified by downgradient changes in hydro- chemical facies, and by changes in concentrations of total dissolved solids and of sulphate ion.

RADIOCARBON CONCENTRATIONS

Results of radiocarbon measurements made on the bicarbonate ion in water from each well are provided in table 2 and figure 4b. The measurements are part of a continu-

604

605

1 sa313w

606

607

ing study of the isotopic composition of water and rock samples representative of the aquifer, that will include an investigation of the isotopes of hydrogen, carbon, oxygen, and sulphur.

The results of C14 determinations have been published elsewhere (Hanshaw, Back and Rubin, 1965) for the group of wells from Polk City south to Cleveland. Because there is no evidence of significant recharge south of Polk City, differences in CI4 ages between wells are assumed to be due to time of travel and are used to calculate apparent velocities of groundwater flow. Velocities range between two and 12 metres per year (table 2). Net velocities between Polk City and wells to the south calculated from hydrologic equations are essentially in agreement with the radiocarbon-determined velocities (table 2). Although apparent radiocarbon ages will not give absolute residence time, age difference between two wells will approximate relative residence time.

MINERAL EQUILIBRIUM

Measurements of departure from chemical equilibrium between groundwater and the minerals which comprise an aquifer aid in delineating principal areas of recharge and in predicting areas subject to solution or deposition of minerals. Equilibrium studies also increase our knowledge of the minerals that control the chemistry of water and the changes that may occur with time or because of the application of a stress upon the hydrologic system.

The minerals considered in this study are those most common in the principal artesian aquifer: calcite, aragonite, dolomite, gypsum, and anhydrite. The thermo- dynamic model used in this study was developed by Back (1961) and has been applied to a study of calcium carbonate saturation of groundwater in Florida (Back, 1963). Each of the minerals is handled thermodynamically in the same manner; gypsum will be used as an example. The chemical equation describing the solution of gypsum is

CaSO, .2 H20 = Ca+’ +SO 4 ’ + 2 H20 and the equilibrium constant, Keq, for gypsum Kgyp, is

where a is activity. By definition, the activity of a solid phase is equal to one; for dilute water the activity of water may also be considered equal to one, hence equation (2) becomes

Kgyp = %a+2US04-Z (3) The equilibrium constant was calculated from the Gibbs free energy values by standard thermochemical methods (see Garrels, 1960, pp. 6-18). The Keq given in the heading of table 3 is at 25°C; the Keq of the minerals at the temperature of the groundwater was calculated by use of the Van’t Hoff equation.

Several steps are required to go from standard chemical analyses such as those in table 1 to activities of single ions. (See Garrels, 1960, pp. 3-42, Back, 1963.) The calculated activities of calcium and sulphate ions are used in equation (3) to obtain the ion activity product, Kiap. The Kisp is compared with the equilibrium constant to determine the departure from equilibrium of the water with respect to gypsum. If the ratio, Kisp: Kgyp, is equal to one (or 100%) the water is saturated with respect to gypsum; a ratio less than one (less than loo%), the water is undersaturated and if the ratio is greater than one, the water is supersaturated. From the form of equation (3) it is seen that the Kiag for gypsum and anhydrite are the same.

The principals for determining carbonate mineral saturation are identical. However,

608

a M

+

Em

a

E2

2

O

m N

m O

r. a I m

d

m

& id ;o O

z oj d

determination of saturation of carbonate minerals requires accurate determination of bothpH and bicarbonate in the field. This is necessary because the equilibrium constant for calcite is

whereas the major dissolved carbon species is commonly bicarbonate ion at the p H of most groundwater. Therefore it is necessary to calculate aCO3-2 from the relationship

Kcaic = %i+z%03-2 (4)

and G(co~-~ is sensitive to changes in a~+(pH). As shown above for gypsum and anhydrite, the Kiap for calcite and aragonite is the same.

Departure of the groundwater from equilibrium with respect to gypsum and anhydrite is shown in table 3 and figure SQ. The entire study area contains water undersaturated with respect to both gypsum and anhydrite. However, water in the Polk City-Wildwood area is most undersaturated. This is related to the low sulphate ion concentration in water of that area. The paucity of sulphate may be caused either by lack of deposition of gypsum and anhydrite in this area, or-perhaps more likely- by the great amount of recharge and the consequent removal of sulphate minerals by solution during the millennia since the rocks were elevated to where they could be dissolved by percolating waters.

It is suggested that the progressive decrease in undersaturation of water with respect to gypsum and anhydrite south of Polk City is an expression of distance of travel time from the recharge area and residence time (fig. 46) in the aquifer system.

In table 3 and on all the figures representing the results of the mineral equilibrium study (fig. 4b-6b) an apparent reversal of trend is observed in water from the Cleveland well. This reversal may be the result of mixing as mentioned above.

In the study area the groundwater is undersaturated with respect to aragonite (fig. 5b) but is most undersaturated in the principal area of recharge-the Wildwood- Polk City area.

Results of the calcite saturation investigation (fig. 6a and table 3) are generally similar to, yet significantly different from the picture developed for the aragonite study. (Compare fig. 6n and 5b.) All but one well-that at Wildwood-produce water supersaturated with respect to calcite. However, the Groveland-Polk City area contains water which is less supersaturated than elsewhere; the generally low degree of saturation at Wildwood probably indicates the influence of significant amounts of recharge on the chemistry of groundwater.

An important but unanswered question is, “How does water in a limestone aquifer remain supersaturated with respect to the major solid phase present in the aquifer?” First, perhaps the analytical results may be slightly in error and the supersaturation may be more apparent than real; only a slight excess of calcium or carbonate ion is necessary to achieve the percentages of supersaturation shown in table 3. However, water from wells used jn this study has been sampled and analysed twice at different times of the year with similar results. Thus we believe that supersaturation is real and not an implied false value resulting from analytical error.

Secondly, this study was conducted using the values of free energy of pure, ordered phases of CaC03 to calculate equilibrium constants. However, many limestones are not pure CaC03 but contain as much as several per cent magnesium in the calcite. This departure from a pure phase increases the free energy of the phase which, in turn, increases the value of the equilibrium constant. Results obtained in this study are within the limits of the possibility that the groundwater is in equilibrium with a magnesian calcite. A third and intriguing possibility (which is also compatible with the hypothesis

that the water is approximately in equilibrium with a magnesian calcite phase) is that

’

610

61 1

2 $

3 aNV13A313

viav3w

VlílH3ílVM

3a

v~

1804

A113 Nod

(INV13AOäE)

aOOMallM

VlV30

E P 4

,” 1331

612

ûNV13h313

viaviw

VlflHlflVM

wlw iüoi

111) Ilod

a00MallM

VlVIO

c 4

-8 R E B B O

o

*i' Y

*

.4

h Y

d 613

water will maintain supersaturation with respect to pure calcite, possibly for kinetic reasons, until saturation with respect to aragonite is reached. Once equilibrium is attained, aragonite may precipitate and later invert to calcite. In the ocean and in closed basin environments, aragonite, in preference to calcite, is the usual phase to precipitate out of solution. Perhaps, in groundwater also, aragonite and not calcite is the control on carbonate equilibrium.

Although the water between Ocala and Polk City is undersaturated with respect to dolomite, the water from the wells at Ocala and Polk City is nearly at equilibrium. The departure from equilibrium at Wildwood and Groveland suggests this area is one of major recharge to the aquifer. In this study, a commonly accepted value 1 x 10-17 was used as the Keg for dolomite. The relatively high degree of supersaturation may be the result of choosing a Keg which is too high; perhaps a value of 2 or 3 x 10-17, as suggested by Hsu (1963) and by Barnes and Back (1964), is more nearly correct.

Lack of equilibrium with respect to dolomite in the Ocala-Polk City area is controlled by the low magnesium content of groundwater in this region. Away from the area of major recharge the calcium content increases by a factor of two or, at most, three. The magnesium content, however, increases by a factor of from three to 30 (table 1 and fig. 2). The increase in magnesium content probably is caused by t ~ e greater solubility of a magnesium calcite compared to pure calcite. The concentration of calcium ion is maintained at a rather constant value because the groundwater in nearly all the study area is supersaturated with respect to calcite.

CONCLUSIONS

This investigation confirms previous studies (Stringfield, 1936, p. 151) that major recharge to the Floridan aquifer may occur not only in high but also in some lower areas of the piezometric surface. Principal areas of recharge in the Floridan aquifer are indicated by low total dissolved solids, low sulphate and magnesium ion content, high C14 concentrations (low apparent ages), and by degree of undersaturation with respect to various solid phases. As water moves away from the recharge area it increases in total dissolved solids, sulphate and Mg:Ca ratio. In addition, CI4 decreases in a systematic manner allowing calculation of apparent groundwater velocities. A study of mineral equilibrium indicates that groundwater increases in saturation away from recharge areas and may become supersaturated with respect to some solid phases.

REFERENCES BACK, William, 1961, Calcium carbonate saturation in ground water, from routine

analyses: U.S. Geol. Survey Water -SuppìyPaper 1535-D, 14 p. -, 1963, Preliminary results of a study of calcium carbonate saturation of ground

water in central Florida: Internat. Assoc. Sic. Hydrol.: 8, 43-51. BARNES, Ivan and BACK, William, 1964, Dolomite solubility in ground water: Art. 160

in U.S. Geol. Siircey Prof. Paper 475-D, D179-180. GARRELS, R. M. (1960) Mineral Equilibria: Harper and Bros., New York, 254 p,. HANSHAW, B. B., BACK, William, and RUBIN, Meyer, 1965, Radiocarbon determinations

for estimating ground-water flow velocities in central Florida: Science, Apr., 23, 148, no. 3669, 494-495.

HEALY, H.G., 1962, Piezometric surface of the Florida aquifer in Florida, July 6-17, 1961: Fla. Geol. Szwuey, Map Series No. 1.

Hsu. K. 3.. 1963. Solubilitv of dolomite and comDosition of Florida ground waters: J. Hydrol., 1,’288-310.

LEGGRAND, H.E., and STRINGFIELD, V.T., 1965, Development of permeability in the Tertiary limestones of the southeastern states, U.S.A.: (in press) International Assoc.

I

Sci. Hydrology. PRIDE, R.W., MEYER, F.W., and CHERRY, R.N., 1961, Interim report on the hydrologic

features of the Green Swamp area in central Florida: Fla. Geol. Survey Cire., no, 26, 968.

STRINGFIELD, V.T., 1936, Artesian water in the Florida peninsula: US. Geol. Survey Water-Supply Paper 7734, pp. 115-195.

614