Caustic waste contamination in a karst aquifer : Nonconservative behavior of sodium

Caustic waste contamination in a karst aquifer : Nonconservative behavior of sodium

Material Information

Caustic waste contamination in a karst aquifer : Nonconservative behavior of sodium
Taraszki, Michael David
Place of Publication:
Tampa, Florida
University of South Florida
Publication Date:
Physical Description:
vi, 74 leaves : ill. ; 29 cm.


Subjects / Keywords:
Groundwater -- Sodium content ( lcsh )
Aluminum industry and trade -- Jamaica ( lcsh )
Hydrology, Karst -- Jamaica ( lcsh )
Dissertations, Academic -- Geology -- Masters -- USF ( FTS )


General Note:
One folded map. Thesis (M.S.)--University of South Florida, 1993. Includes bibliographical references (leaves 70-74).

Record Information

Source Institution:
University of South Florida
Holding Location:
Universtity of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
029765354 ( ALEPH )
30908894 ( OCLC )
F51-00107 ( USFLDC DOI )
f51.107 ( USFLDC Handle )

Postcard Information



This item is only available as the following downloads:

Full Text


CAUSTIC WASTE CONTAMINATION IN A KARST AQUIFER: NONCONSERVATIVE BEHAVIOR OF SODIUM by MICHAEL DAVID TARASZKI A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in the Department of Geology in the University of South Florida August 1993 Major Professor: Dr. Sam B. Upchurch


Graduate School University of South Florida Tampa, Florida CERTIFICATE OF APPROVAL MASTER'S THESIS This is to certify that the Master's Thesis of MICHAEL DAVID TARASZKI with a major in Geology has been approved by the Examining Committee on August 1993 as satisfactory for the Thesis requirement for the Master of Science. Thesis Committee: Major Dr. Sam B 'Upchurch Member: Dr. Mark T. Stewart Mernber:"Dr. H. Len Vacher


ACKNOWLEDGEMENTS I would like to extend my greatest appreciation to the members of my thesis committee, Dr. Sam B. Upchurch, Dr. Mark T Stewart, and Dr. H. Len Vacher, for their guidance and support throughout this study. I would also like to thank Mr. Morris Roberson of Aluminum Partners of Jamaica (ALPART) and Mr. Mike Gurr of Gurr & Associates, Inc. for the use of data collected at the ALPART site in Nain, Jamaica. Additionally, I would like to thank Dennis and Nancy Taraszki for their years of support, Russell Watrous and Jian Chen for all the help in the laboratory, and Jeff Burdick for repeatedly proofreading this document.


LIST OF TABLES LIST OF FIGURES ABSTRACT INTRODUCTION Objectives Geology Hydrogeology PREVIOUS WORK METHODOLOGY TABLE OF CONTENTS Construction of water-table map Chemical calculations Mixing models Calculation of deficits and surpluses DATA PRESENTATION Historical sodium concentrations Groundwater chemistry in 1990 Mineral precipitation ii iii iv 1 3 6 11 14 17 17 17 19 21 23 23 29 36 DISCUSSION 39 Historical sodium concentrations 39 Chemical trends of the Essex Valley aquifer 42 Chemical surpluses and deficit 48 Mechanisms of sodium fixation 53 Chemical modeling 59 SUMMARY 66 LIST OF REFERENCES 70 i


LIST OF TABLES Table 1. Elevations used for the construction of the water-table map 18 Table 2. Chemical analyses of waters collected in South Pond and Essex Valley aquifer, as reported by Gurr & Assoc., 1990. 30 Table 3. Log PC02 ion activities, and saturation indices of 1990 water samples calculated using PHREEQE and WATEQ4F. 35 Table 4. ALPART wells listed in the orders of decreasing specific conductivities, total dissolved solids, and chloride concentrations. 45 Table 5. Calculated surpluses and deficits for all major dissolved constituents in Essex Valley aquifer. 49 ii


LIST OF FIGURES Figure 1 Site map of the ALPART plant at Nain, Jamaica. 2 Figure 2 Comparison between chloride and sodium concentrations. . 5 Figure 3 Generalized geologic map of Jamaica. 7 Figure 4. Tertiary stratigraphy of Jamaica. 8 Figure 5. Topographic map of Essex Valley. 10 Figure 6 Contours of water-table elevations (meters) measured in September 1990. 12 Figure 7. Historical sodium concentrations in the ALPART wells from 1970 to 1990. 24 Figure 8 Chemical relationships in the Essex Valley aquifer. 31 Figure 9. Piper diagram showing the chemical trends in Essex Valley aquifer. 37 Figure 10. X-ray diffractogram of South Pond water precipitate produced in a laboratory experiment by Upchurch, 1990. 38 Figure 11. Comparison between the changes in chloride concentrations, specific conductivity, and total dissolved solids. 46 Figure 12. The sodium, carbonate, and dissolved aluminum deficits and surpluses in Essex Valley aquifer. 50 Figure 13. The calcium, magnesium, and sulfate deficits and surpluses compared to the saturation index of calcite. 51 Figure 14. Results from the PHREEQE Models 1 and 2 compared to observed data. 61 iii


CAUSTIC WASTE CONTAMINATION IN A KARST AQUIFER: NONCONSERVATIVE BEHAVIOR OF SODIUM by MICHAEL DAVID TARASZKI An Abstract of a thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in the Department of Geology in the University of South Florida August 1993 Major Professor: Dr. Sam B Upchurch iv


Aluminum Partners of Jamaica (ALPART) has been processing bauxite for the extraction of alumina since 1969 at Nain, Jamaica. Nain is located in Essex Valley on the southwest side of the island. Alumina is extracted using the Bayer process, which produces sodium-rich caustic waste. The waste slurry is piped to two nearby waste-disposal ponds called North Pond and South Pond. Caustic waste leaking from South Pond has significantly increased sodium levels in Essex Valley aquifer. The increased sodium levels pose a potential threat to other local industries, dairies, and nearby communities dependent on water from this aquifer. The ALPART plant utilizes five industrial wells to supply the plant with water for the process. The September, 1990, water-table map shows a cone of depression formed around the ALPART industrial wells. This depression captures much of the north-flowing sodium plume emanating from South Pond. Water movement in Essex Valley aquifer is influenced by fractures and karst conduits; transmissivities are estimated to range from 1,242 to 124,200 m2/day. The high transmissivities provide for a very deep water table (180 to 215 meters below ground level) and low hydraulic gradients. Groundwater velocities in the site area range from 15 to 1,200 m/day. Chloride concentrations decrease from South Pond to Essex v


Valley aquifer (at industrial well 3) by 70 percent. This change is due only to mixing with formation water and is not affected by any chemical reaction. Sodium concentrations decrease from 507 mmoles/ L (11,660 mg / L) in South Pond, to 29 mmoles/ L (667 mg / L) in well 3, which represents a reduction in sodium concentration by over 90 percent. between chloride and sodium concentrations The d iscrepancy indicates that sodium is acting nonconservatively; a sodium deficit exists between South Pond and Essex Valley aquifer. Carbonate and dissolved aluminum also act nonconservatively. Chemical reactions possibly responsible for sodium fixation include precipitation of sodium carbonates and gibbsite and sorption of sodium onto aluminum species. Trona and dawsonite are the most likely sodium carbonates to form upon evaporation in South Pond. Sodium will co-precipitate with gibbsite, and amorphous aluminum species and gibbsite crystals may also provide sorption sites for sodium. Abstract approved: Major Professor: Dr. Sam B Upchurch Date of Approval: vi


1 INTRODUCTION Aluminum Partners of Jamaica (ALPART), located in Essex Valley, Jamaica, is one of several companies in Jamaica that obtain aluminum from the large bauxite deposits on the island. Aluminum is recovered from the bauxite ore by using the Bayer process. This process involves emulsifying bauxite ore with a heated sodium-hydroxide solution and calcifying the resulting slurry to produce a calciumaluminum precipitate. The aluminum precipitate falls out of solution and is recovered for further refinement. The water needed for the Bayer process is provided by five industrial wells, located immediately north of the ALPART plant (Figure 1) After the aluminum precipitate is extracted, the remaining slurry from the ALPART plant is piped to the primary industrial pond, referred to as South Pond. North Pond is a secondary pond used primarily for rain-water runoff from the plant, but it has received industrial waste intermittently in the past. The waste ponds have been constructed with earthen dams that interconnect sinkhole depressions. The waste slurry is piped from the plant to the eastern


ALPART ll'eppetl 11 07. 8 23 01i:7 o/f; l.o o:fa ll. oa t --0 -2000 meter 5000 Figure 1 Site map of the ALPART plant at Nain, Jamaica. Solid and open circles represent industrial and monitoring wells, respectively. 2


3 edge of South Pond and is discharged in a lternating locations to evenly distribute the resulting mud delta. The slurry enters the pond at 140C. The low permeability of the slurry prevents significant infiltration through the bottom of South Pond, and liquor flows across the mud and enters limestone solution features along the western edge of the pond (Gurr, T M., 1992, pers. comm.). Although North and South Pond are unlined, both are perched above the water table due to the low permeability of the slurry. Sodium levels in Essex Valley aquifer quickly rose after the commissioning of the ALPART plant at Nain, Jamaica, in 1969. The high levels continue to cause concern about the quality of the water used by nearby dairies, communities, and private well owners. ALPART and the Jamaican Government concluded a Joint Study in 1990 to assess the risk of expanding the waste ponds. The Joint Study (Gurr & Associates, Inc. 1990) advised the construction of .larger ponds with a concomitant reduction of caustic soda in the waste. Objectives The objectives of this paper are: to determine the primary source of sodium in Essex Valley aquifer; to quantify the nonconservative behavior of sodium in Essex Valley aquifer; and to analytically assess possible mechanisms explaining the nonconservative behavior of


4 .sodium. The first objective is necessary to distinguish the impact of North Pond from South Pond on the sodium levels in Essex Valley aquifer. The second objective is accomplished by comparing the changes in chloride concentrations to changes in sodium concentrations from South Pond to Essex Valley aquifer (Figure 2). Chloride is an important indicator of advective and dispersive processes, unless it is found in extremely high concentrations (Feth, 1981) Chloride concentrations found in Essex Valley aquifer (at industrial well 3) indicate that 70 percent of the chloride found in South Pond has been removed by advective and dispersive processes. The percentage of any other chemical constituent should be similar to the percentage of chloride in a water sample if that constituent is behaving conservatively. Sodium concentrations in Essex Valley aquifer (at well 3) show that ninety-five percent of the sodium in South Pond has been removed from solution (Figure 2). The twenty-five percent difference between chloride and sodium percentages in well 3 is referred to as the sodium deficit. This deficit indicates that sodium is chemically reactive. It is necessary, then, to analytically assess possible mechanisms which may act to chemically remove sodium from Essex Valley aquifer system.


35 30 25 .... u .... t'3 '0 20 = 0 ...::: .... 15 ::J 0 Cl) .... 0 .... 10 = 0 u .... u 5 0 -5 3 4 12 6 14 20 21 7 9 16 22 18 13 Well number (--> decreasing chloride cone ) Figure 2. Comparison between chloride and sodium concentrations, presented as the percent of each constituent's concentration in South Pond, for each well. U1


6 Geology Jamaica is part of the Greater Antilles islands in the Caribbean, all of which originated from volcanic activity. Tertiary-age fault systems trend predominantly northnorthwest and subdivide the island into three principal tectonic blocks. Major fault zones include the Wagwater belt and the Newmarket-Montpelier zone; both trend northnorthwest (Figure 3). Fault zones trending east-west include the North and South Coast Fault zones and the Crawle River-Rio Minho Fault (A, B, and C, Figure 3). The early Eocene marked the beginning of an island-wide transgression and led to the deposition of thick, mineralogically pure (95-100% CaC03 ; Muhs and others, 1990) limestone units. The great thicknesses of these limestone units is a testament to the rapid subsidence during this time. Late Miocene re-emergence of the island greatly reduced the extent of carbonate deposition. The Tertiary stratigraphy of Jamaica is shown in Figure 4 The principal depositional environments through the early Tertiary were shallow-water lagoons and shelf edges (Draper, 1988; Robinson, 1990). Deep-water facies are found in areas that underwent more rapid subsidence. The White Limestone Group is the most extensive sedimentary unit in Jamaica and covers almost two-thirds of the island. The Newport Formation reaches thicknesses of over 1220 meters (Robinson, 1990) and is the only unit exposed in Essex Valley.


IS N lliiill \ ALLUVIUM Recenl 1a w 531 WHITE UMESTONE M EoceneM YELLOW UMESTONE L EoccneM Eocene D SEDIMENTARY & VOLCANIC ROCKS / FAULTS Rote Diaaram or Fault Traces nw '\ 0 ' -[) t_EOKN. ( IJA 50 km c::;> .... Figure 3. Generalized geologic map of Jamaica. (modified from Lewis and Draper, 1988) -...J




9 Essex Valley lies within the Newmarket-Montpelier zone and is bounded to the east and west by the Spur Tree and the Santa Cruz Fault Systems, respectively (labelled 1 and 3 in Figure 3). The Essex Valley Fault System lies between these two systems {labelled 2 in Figure 3). The Spur Tree Escarpment forms the Figuerero Mountains and provides over 600 meters of relief in the valley. The valley floor slopes down to the northeast from the Santa Cruz Mountains and is abruptly terminated at the Spur Tree Fault of the Figuerero Mountains (Figure 5) Karst formation began soon after the mid-Tertiary uplift. Smith {1970) estimates that present karst features of the island were formed in 3.5 to 6 million years. The most notable karst feature is called Cockpit karst and lies in the northwest section of Jamaica. Cockpit karst includes classic tower karst features, such as star-shaped sinkholes and residual limestone mounds Essex Valley lies south of the Cockpit area but has similar karst features. Smaller towers are most pronounced in the southern portion of the Valley with higher eleva-tions and become less prominent in the lower elevations to the north. Sinkholes pit the entire valley floor and create a very rugged topography. Bauxite mined at the surface consists of soil rich in gibbsite, boehmite, and diaspore {Robinson, 1990) Alluvia--tion has allowed aluminum-rich soils in solution features


D 0 I -1> .. --._ FAUUW/.ans -liiJIO""" -CIIIITIUI ,.--------Cit-0 SCAlL 0 2.500 5000 I'!ET CARIBBEAN SEA Figure 5. Topographic map of Essex valley, Jamaica. 1990). 4LUC4 TOR PONO BAY (after Gurr & Assoc., Inc., 1-' 0


and faults to be in contact with groundwater. Large vugs within the limestone near the surface contain bauxite and can be seen in roadcuts throughout the valley. Hydrogeology 11 There is a groundwater divide located approximately beneath the South Pond of the Alpart Plant (Figure 6 ) Most groundwater from Essex Valley flows to the north and drains to the Upper Morass (near Horse Savanna) which forms the headwaters of the Black River in the Black River Basin (Figure 5). Groundwater also flows south along fractures to feed offshore, freshwater springs in Alligator Pond (Figure 5; Kohout and others, 1973). An absence of surface drainage in the valley results from high infiltration rates through karst conduits found throughout valley and the unconfined conditions of the aquifer. Only during extremely heavy rains do gullies carry water for brief periods of time (Gurr & Associates, Inc., 1990). The water table north of South Pond slopes to the northeast and mimics the topography. Infiltration through the limestone along the perimeter of South Pond results in a mounded water table and causes water to flow north, south, and east. South Pond is the source of excess sodium recovered from wells north and south of the ALPART plant (Gurr & Associates, Inc., 1990). In the vicinity of the site, the depth to the water


0 2000 m.tera 20 '-.... """ 22 025:.2'\ 5000 \ 25 9 ALPART Oe-:-s 0 70 7:-t \ 0].3 I 26.6 1!. 0&.11 ---10 _15 """ 20 Figure 6. Contours of water-table elevations (meters) measured in September, 1990. 12


13 .table ranges from 180 to 215 meters, and water-table elevations range from 6 to 25 meters above sea level {Gurr & Associates, Inc. 1990). The great depth of the water table is a result of the very low hydraulic gradient produced by the extremely high transmissivities of the karstic and fractured limestone. Well-yield tests led to estimates of transmissivities in the Upper Black River Basin from 1,242 m2/day to about 124,200 m2/day {Tollenaar, 1983). Tracer studies of conduits through the White Limestone in various locations yield velocities from 15 to 1,220 m/day {Wedderburn, 1977). Groundwater flows from the Santa Cruz mountains north-eastward until reaching the valley floor. Here, the water flow continues to the north and to the south. The lack of well control or a topographic high complicates locating the exact position of a divide. Gurr & Associates, Inc. {1990) reported that a divide must separate North Pond and South Pond to explain excess sodium levels to the south. Figure 6 shows the water-table map constructed from the September, 1990, water-table elevations measured near the site area. Based on these data, the divide may lie below or slightly east of South Pond, which accounts for north and south flow of water infiltrating from South Pond.


14 PREVIOUS WORK Rock-water interaction largely regulates the evolution of water chemistry through a groundwater system. Many reactions can be responsible for the chemical changes observed along a flowpath; however, thermodynamics and massbalance modeling can eliminate many as improbable. Hem (1985) and Domenico and Schwartz (1990) list many common reactions expected in natural-water systems. Hanshaw and Back (1979) discuss reactions specific to carbonate aquifers. Mass-balance modeling of chemical change in carbonate water systems has been by groundwater chemists for many years. Plummer (1977) used a mass-balance approach to define reactions in the Floridan aquifer to simulate the observed changes in water Plummer and Back (1980) studied the chemical and isotopic evolution of the Floridan aquifer and Madison Limestone aquifer using a mass-balance approach. Although these studies primarily explained observed changes in natural chemical systems, mass-balance approaches are valid to explain any chemical changes along a flowpath. Domenico and Schwartz (1990) describe mass-transport processes as the primary controlling factors in plume


expansion in an aquifer system, while chemical reactions serve to retard the plume's spread. In a highly porous, rapid-transport system, such as Essex Valley aquifer, advective transport is the dominant physical process (Domenico and Schwartz, 1990). 15 This study utilizes the conservative nature of chloride in constructing mixing curves. Kimmel and Braids (1980) used this method to evaluate the extent and direction of a plume from a landfill in Babylon, New York. Patterson and others (1985) compared the distribution of various volatile organic constituents against chloride distribution at the Glouchester Landfill, near Ottawa, Canada. Comparing the ionic distribution of a particular element to chloride indicates whether that element is behaving conservatively or not, because changes in chloride concentrations reflect dilution and dispersion effects only. Wedderburn (1977) evaluated the potential for groundwater pollution from several aluminum industry waste ponds, including the ALPART pond at Nain, which he referred to as Site Two. In his study, Wedderburn concluded that sodium acts conservatively upon reacting with the limestone. Upchurch (1990) conducted a laboratory test to determine the mechanism for sodium fixation and concluded that precipitation of a sodium carbonate phase was at least partially responsible for removing sodium from the system. Evaporation of South Pond liquor by 50 percent yielded a


white precipitate which has been identified in this report as a mixture of trona and dawsonite. 16 PHREEQE and WATEQ4F are computer programs commonly used in groundwater chemistry studies (Henderson, 1984; Hull, 1984; Knobel and Phillips, 1988; Wood and Low, 1988; Zack and Roberts, 1988). Sprinkle (1989) used mixing curves generated from WATEQ to describe changes in groundwater chemistry within the Floridan aquifer as percent original water. Changes in water chemistry in Essex Valley aquifer are described similarly.


17 METHODOLOGY Construction of water-table map The data used to describe the water table are limited to the ALPART industrial and monitoring wells. Water-table elevations collected in September, 1990 (Table 1) have been hand-contoured to produce the water-table map shown in Figure 6. The shape of the contours north of the waste ponds is constrained by chemical data indicating northward flow from South Pond. A computer contouring program would be inappropriate, as it could neither utilize chemical data nor recognize the anisotropic nature of the aquifer. The water-table map clearly shows a depression around the industrial well field. Pre-development water-table elevations are not known; however, present-day elevations distant from the well field should approximate predevelopment levels. The presence of fractures and karst solution features complicate the construction of formal hydrologic models and specific data required to construct a detailed model are not available for this study site. Chemical calculations The ALPART laboratory and PHOSLAB, Inc. each analyzed September, 1990, water samples by Standard Methods (Gurr &


18 Table 1. Elevations used for the construction of the water-table map (from ALPART, 1990) Water table and water table criteria Well Water table Well Total Casing Screen Open Diameter I. D Elevation Elevation depth length length Hole 2 10.3 175 7 219 172 24 22 6 3 10.3 183 3 218 174 24 20 7 4 9 9 163 2 204 163 24 17 7 5 12.3 164 7 206 164 24 18 7 6 13.9 158 8 212 182 0 30 7 7 7 1 77 4 7 9 6 5 55 7 7 11 7.8 146.3 183 145 15 22 2 12 12 4 170.9 212 171 15 26 2 13 26 6 229 0 232 219 12 1 2 14 20 9 177 3 212 167 12 33 2 15 15 4 196 9 212 184 15 13 2 16 8 9 165 3 273 180 15 77 2 17 8.6 182 2 220 192 12 15 2 18 21.2 191. 2 243 179 15 49 2 19 23.3 125 0 161 129 15 17 2 20 22 1 147 3 212 142 15 55 2 21 21. 8 127.9 189 137 15 37 2 22 25.2 114 9 167 107 15 45 2 23 16.7 167 7 213 164 15 34 2 values as re


19 Associates, Inc., 1990). These analyses report the unspeciated, total concentrations of chemical constituents. Speciation refers to the ion-pair complexes formed with ions under various ionic strengths, pH's, and temperatures. The programs PHREEQE and WATEQ4F are commonly used to calculate ionic speciation, ion activities, activity coefficients, and mineral-saturation indices (Parkhurst and others, 1990: Ball and Nordstrom, 1991) Both programs require pH, Eh, temperature, and constituent concentrations measured from water analyses. Mineral-saturation indices are calculated using the equation: S.I. = log IAP -log KT, where IAP refers to the ion activity product, and KT refers to the equilibrium constant for the mineral in question at environmental temperatures. Activity coefficients were calculated using the Davies equation (Parkhurst and others, 1990) Both programs check for electrical neutrality between cations and anions. Adding bicarbonate to sample analyses, as recommended by Hem (1985), corrected minor electrical imbalances in the 1990 data. An electrical imbalance of greater than 10 percent disqualified that sample from use. Mixing models In addition to the calculations mentioned above, PHREEQE also can mix two solutions in any proportion to


20 produce a series of resultant solutions. Changes in water chemistry predicted by PHREEQE are due to mixing only; rock-water interaction and mineral precipitation or dissolution reactions do not proceed. PHREEQE Model 1 mixes South Pond water and formation water (defined here as water from Well 13) in increments equal to the percentage of South Pond water found in each well. The following equation is used to calculate the percentage of South Pond water in each well, based on chloride concentrations: where: [Cl]m -[Cl]B ---___;_xlOO = % SouthPondwater = PsPw [ClJP -[ClJB [Cl]m = chloride concentration measured in a sample, (1) [CL]b = background chloride concentration in well 13, and [Cl]P = South Pond chloride concentration, = percent South Pond Water. Model 2 removes a finite amount of sodium, carbonate, and dissolved aluminum from South Pond before mixing with background water in increments defined above. Chloride concentrations are reduced only by dilution and dispersion, and progressively lower concentrations are found in wells 3, 4, 12, 6, 14, 20, 21, 7, 9, 16, 22, 18, and 13. Although this order of wells does not represent a truly evolutionary flowpath, it does represent the order of mixing between South Pond water and the formation water. All diagrams showing spatial relationships between ions are


illustrated in this order. Mixing trends are often best presented on Piper diagrams, which show the proportions of major dissolved 21 ionic constituents of a water sample in terms of the percent of the total milliequivalents per liter (%meq/L) of sample. Each vertex of the cation of anion triangles represents 100 percent of a particular ion or group of ions. The coordinates of a single point plotted on either of the two triangles add up to 100 percent (Hem, 1985). Calculation of deficits and sukPluses Once the amount of mixing between South Pond and the formation water is known for each well, changes in chemical concentrations form South Pond to the aquifer can be pre-dieted using Equations 2 and 3. The predicted concentra-tions reflect only advection and dispersion processes. where: xi -xp = deficit (-) or surplus ( +) PsPW = percent South Pond Water, XsPW = concentration in South Pond, = predicted concentration, Xi = concentration in the well, XB = background concentration. If an element is acting conservatively, as does chloride, the concentrations calculated with Equation 2 (2) (3)


22 should match closely with those found in the well. Constituents with concentrations higher that those expected from mixing calculations are described as showing a surplus, and imply an addition of that element to the system. Values less than those expected are described as deficits and imply the removal of that constituent from the system.


23 DATA PRESENTATION Historical sodium concentrations Sodium concentrations in ALPART wells are plotted against time in Figures 7a-e. With the exception of well 6, all industrial wells showed a significant increase in sodium concentrations within the six months following the commencement of ALPART plant operations. No other monitoring wells were used for water-table elevation measurements until after 1976. In addition to the five industrial wells, wells 12, 14, and 21 (located north of North Pond) historically show the highest sodium concentrations. Maximum sodium levels range from 200 to 900 mg / L (8.7 and 39.2 mmol / L ) Patterns of fluctuating sodium levels in each industrial well, although initially different, closely resemble each other by 1980, with the exception of well 6. Wells 14 and 21 both show a steady decrease from the elevated sodium levels in 1980 (400 mg/ L and 900 mg /L, respectively) Sodium levels in well 2 1 diminish to near-background concentrations by 1989 Sodium concentrations in well 12 decrease from about 800 mg / L in 1980 to about 350 mg / L in 1990. Wells 7, 8, and 9 all show a steady rise in sodium levels from 1975 to 1986. Well 10 shows a sharp rise in


1000 sooJ 800i 7001 =:::::. C) E 600 c .o -500 -c Q) 400 0 c 0 0 Industrial Wells * Well1 * We114 WellS I Plant I WellS 1974 1978 1982 1986 1990 1972 1976 1980 1984 1988 year (a) Figure 7 Historical sodium concentrations in the ALPART wells from 1970 to 1990 (Gurr & Assoc. Inc., 1990). N


ALPART (Pepper) Wells 50 45 40 C' 35 ....._ 0> .s 30 c 0 1a 25 c Q) 20 0 c 8 15 10 5 0 1970 1974 1978 1982 1972 1976 1980 year (b) Figure 7. (cont'd). 1984 I Plant I Shutdown 1 I I Well7 0 I I 1 11 WellS ......... ...... Well9 -Well10 lh I o 1986 1990 1988 N 01


Monitoring Wells 1000 sooJ IJ Plant J aool Jjl_ J:shutdown 700 :::::::::. 0) --*'E 600 Well12 -c 0 ___.__ 1ti 500 0 Well14 '-..... c 400 ... I J I Q) 0 c 0 300 0 200-i 0 ... 100 I 0 1970 1974 1978 1982 1986 1990 1972 1976 1980 1984 1988 year (c) Figure 7. (cont'd). N 0'1


Monitor i ng Wells 35 I I 30-1 1 I Plant I I I I :shutdown: I 0 II Well 11 c25 ._ CJ) Well17 E ......... c 20 0 A Well18 1ii '-..... c 15 Q) (' I II :Su22 0 c 0 0 101 II I 5 1990 1970 1974 1978 1972 1976 Figure 7 (cont'd). 1980 year (d) 1982 1986 1984 1988 N ...J


Monitoring Wells 30 -25 :::::::::. 0> E -c 20 0 c 15 Q) 0 c 0 0 10 5 I Plant I Shutdown .... 0 Well15 Well16 A Well 20 Well 23 1970 1974 1978 1982 1986 1990 1972 1976 Figure 7 (cont'd). 1980 year (e) 1984 1988 N ())


29 sodium levels in 1980 followed by a sharp decline in 1982. These four wells represent the most distant wells from South Pond addressed in this paper. They comprise a small industrial well field called the ALPART (Pepper) well field shown on Figure 1 With the exception of wells 7, 8, 9, and 15, all monitoring wells show a sharp rise in sodium levels in 1980, followed by a sharp decline in 1982 Wells 10, 12, and 21 show the largest and most distinct change in sodium concentrations during this period. The remaining monitoring wells show less distinct and smaller changes in sodium concentrations. Sodium levels in well 14 are elevated, but a distinct rise and fall from 1980 to 1982, if one was present, would not have been observed due to missing data. Groundwater Chemistry in 1990 The. chemical analysis of waters sampled from South Pond and monitoring wells in 1990 are presented in Table 2 and Figures Sa-c. Sodium is the principal cation present in the pond, and carbonate is the primary anion present. Sodium concentrations in South Pond may increase with increasing evaporation rates or decrease with the influx of rainwater. Values of up to 640 mmol/L (14,700 mg/L) and as low as 132 mmol/L (3035 mg/L) have been reported since the 1990 Joint Study and suggest both lateral and vertical variations of sodium concentrations in the pond (Gurr & Associates, Inc.


Table 2. SamniA/wall number Date of Sample Water temperature (C) Field pH Eh Sp. Cond. (umhos) TDS(maJU Chloride Sodium Carbon (tot.) Aluminum (ctaolved) Calcium Magneelum Sulfate Iron, (tot ) Silica Chemical analyses of waters collected in South Pond and Essex Valley aquifer, as reported by Gurr & Assoc., Inc., 1990. Group I Wells Group II Wells S Pond 3 4 12 8 14 20 21 7 9 18 22 18 9/90 9/90 9/90 9/90 9/90 9/90 9/90 9/00 9/90 9/90 9/90 9/90 9/90 24. 1 28. 1 27.1 28. 0 28. 5 28. 0 27. 2 25.0 27. 1 28. 0 28. 1 28. 7 28. 1 11. 9 10. 8 10. 0 9 5 8 8 8 7 8 3 8 .64 8 .93 8 .89 8 .88 6.82 6 .72 -149 -14 141 157 158 289 278 279 285 298 305 282 320 21300 33eO 2430. 1360 1025 810 510 500 485 450 525 480 490 34480 2332 1885 1010 765 455 361 369 357 339 402 361 343 Concentrallona Df--.ted In millimole per Iller of eamole water 5 .78 1 .87 1 .38 0 .95 0 .73 0 .52 0 .49 0 .39 0 .38 0 .35 0 .28 0 .27 0 .25 507.21 29.45 9 .83 13.22 9 .83 3 .55 0 .39 0 .85 1 .37 0 .96 1 .35 0 .50 0 .24 291 .88 11. n 14. n 9 .44 7 .41 4 .93 3 .93 4.06 3 .80 3 .67 4.33 4 .13 3 .82 118.35 0 .55 N D N D N D N D N D N D N.D N .D. N D N D N D 2.20 0 .02 0 .02 0 .09 0 .21 1 .06 1 .96 1 .00 1 .49 1 .51 1 .97 2 .01 2 .14 N.D N D 0 .21 0 .02 0 .21 0 .20 0 .44 0 .48 0 .40 0 .48 0.49 0 .14 0 .21 8 .72 4 .97 2 .23 0 .87 0 .42 0 .02 0 .02 0.02 0.08 0 .06 0 .02 0 .01 0 .02 15.58 N D N D N D N D N D N D N D N D N D N D N D N D 0.54 0 .04 0 .02 0 .01 0 .01 0 .02 0 .01 0 .02 0 .01 0 .02 0.02 0 .02 0 .01 Wells listed with decreasing chloride cone to the right mo

= = 1000 100 10 1 0.1 S. P 3 4 12 6 14 20 21 7 9 16 22 18 13 Well numbers (decreasing CI -->) _sodium Carbonate __.__ Aluminum (a) Figure 8 Chemical relationships in Essex Valley aquifer. w ......


2.5 ,..------------------------..., 1 I I ./\ \ I -2 2 -.. s I \ I I \ I \ -3 0 e e '-" e 1 5 \ J -\ 1 4 ::::2 -fll v 61 cu ::E -53 "0 1 t= cu e ::::2 I 1\ I \ -6 -(.,) a o.5 I I \ I .----.-___ -'\. /'I -7 0 8 S P 3 4 12 6 14 20 21 7 9 16 2 2 1 8 13 Well numbers (d ec reasing Cl --> ) --e-Calcium --+M a gnesium __._ Log PC02 (b) Figure 8 (cont'd). w N


-1 15 -2 I \I 10 --3 I A I ::r: 0.. -4 r ...___.___..---T ....5 j -5 (.) I "{ I --6 -I I I -7 -8 S.P. 3 4 12 6 14 20 21 7 9 16 22 18 13 Well numbers (decreasing Cl -->) -log PC02 --+-pH -..Calcite S .I. (c) Figure 8. (cont'd). d 0 -5 VJ VJ


34 1992) The high levels of sodium hydroxide cause a very alkaline environment in the South Pond, with pH's ranging from 11 to 13; the high pH depletes the water of C02 (log PC02 = -7.37). The high pH values found in South Pond are maintained in the aquifer near the industrial well field but diminish to neutral levels in all other monitoring wells. Due to the high ionic strength (0.5387), speciation of most analytes in South Pond is significant, and the activity of most analytes is low relative to the concentration. The redox potential changes from a very reducing environment in South Pond to a progressively more oxidizing environment in the aquifer, downgradient from South Pond. All chemical constituents, with the exception of calcium and magnesium, decrease in concentration from South Pond to the aquifer. The concentrations of constituents presented in Table 2 are used to construct the Piper diagram shown in Figure 9. Two major clusters of wells are labelled Group I and Group II. Group I samples are described as a sodium-carbonate type water and comprise South Pond water, the five industrial wells, and monitoring well 12. The remaining monitoring samples, with the exception of the sample from well 14, comprise Group II and are described as a calciumbicarbonate water. The water chemistry at monitoring well 14 is chemically transitional between Groups I and II.


Table 3. SamDie/Well number Log PC02 Ionic strenath Sodium Carbonate Bicarbonate Alumlrum Calc ium Magnesium I HvdrOldde I Trona DaW8001te Calcite Dolomite Natron Nahcolite G i bbsite Log PC02 ion activities, and saturation indices of 1990 water samples, calculated using PHREEQE and WATEQ4F. Group I Weill Group II We ll s S Pond 3 4 12 6 1 4 20 2 1 7 9 16 22 18 -7.374 -4.297 -4.344 -3.89 -3.142 -1.362 -1.215 -1.422 1 .651 -1.842 -1.174 -1.386 -1. 498 0 .5387 0 .0439 0 0431 0 0161 0 0108 0 .0063 0 .0062 0 0067 0 0062 0 0061 0 .0067 0 0059 0 0063 -LOG 10 ION AC T IVIllES O .lfl Z 1 .845 1 ,717 1 .944 2 057 2 .488 3 444 3 .228 2 .901 3.055 2.907 3 .333 3 .651 1.588 2 .484 2 .509 2 976 3 747 6 .128 6 759 6 .291 5 .935 6 009 6 718 8 .291 8 202 3 219 2 .134 2 188 2 .188 2 .201 2 .498 2 .749 2 .802 2 .554 2 .578 2 .708 2 .597 2 803 'Z1.2n 20.449 N D N D N D N D N D N D N D N D N D N D N D 4 .301 5 .648 5 544 4 812 3 .974 3 133 2 .858 2.879 2 .988 2 974 2 .883 2 .848 2 824 e.1n N D N D 5 .058 3 .944 3 .859 3.509 3 .496 3 .554 3 473 3 .465 4 018 3 .831 2 .025 3 .968 3 .934 4 .429 5 .173 7 .289 7 621 7 .352 6.994 7 .069 7 621 7 317 7 238 SATURATION INDICIES -2.81 -8.89 -9.14 -10.16 -11.28 -15.25 -18.95 -17.07 -17.78 -16.91 -17.28 18 .02 -18.91 0.69 1 .71 ... ... .. .. .. .. ... .. 2 .60 0 .38 0 .44 0 .90 o n 0 .78 -1.12 -0.69 -0.89 0 50 -1.09 0 .65 0 54 3 .51 .. 1 .49 1 .71 -2.14 2 .74 1 .76 -1.87 -1.35 -2.62 -2.32 -1.94 -1.58 -4.49 -4.83 -5.59 -8.81 -9.83 -12.42 -11.43 -11.23 -10.85 -11.31 1 1 .71 -12. 23 -2.42 -3.24 -3.44 3 .59 -3.72 -4.45 5 .88 5 .28 -5.18 -5.10 5 .09 5 .39 -5. 72 1 39 0 .38 ... ... ... ... .. .. . . . .. Saturation Index : Wells listed with decrea sing chloride cone to the right ( -) undersaturated N D -not detected ( + ) oversaturated I nfinitely undersaturated ( o ) equilibrium 1 3 2 .881 0 0027 3 .601: 5 .9n 3 043 N D 3 4 05 3 438 6 .885 0 .92 1 n -11.78 8 .07 .. w 01


36 .Mineral Precipitation X Ray diffraction results of the filtrates collected from filtered South Pond water samples confirm the presence of calcite, gibbsite, boehmite, goethite, and hematite (Gurr & Associates, Inc., 1990). The X-ray diffractogram of the white precipitate formed from the 1991 experiment, conducted by Upchurch, is shown in Figure 10. The X-ray diffraction pattern is shown in the top block; the lower two blocks contain the X-ray diffraction patterns of two sodium carbonate minerals. Superimposing these X-ray diffraction patterns demonstrates that the white precipitate is almost completely composed of trona (NaHC03Na2C03H20) and dawsonite (NaAlC03 )


CATIONS % OF TOTAL MEO/ L Ci ANIONS 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Pioer D i L Grou Wells South Pond IW-3 IW-4 MW 12 IW-6 MW-14 G rou II Well s MW-20 MW-21 MW-7 MW-9 MW-16 MW-22 MW-18 MW-13 Figure 9 Piper diagram showing the chemical trends in Essex Valley aquifer. The legend presents the wells in order of decreasing chloride concentrations (mmol/L} Industrial and monitoring wells are listed as IW and MW, respectively. d 205 66 48 34 26 1 9 17 14 14 12 9 9 9 .4i 8 7, 5 .71 w .._J


FN:BTTEST .RO DATEIUV16/91 ID:BatCH TOP TEST 10-91 TlHEt11112 PTa 9.909 STEP I 9. 9300 SCINTAG/USA 2.976 0 SODIUM HYDROGEN CARBONATE HYDRATE / TRONA 1.823 29 19 e 0!9-1<4471 I I . J ... I ... 1 I ... J! ....... I ..... I NH3 H ( C 03 )2 2 H2 0 SODIUM ALUMINUM CARBONATE HYDROXIDE / DAWSONITE I I I I I I I II I Ill.. J 19-1175. .I I I INA AL c 03 ( 0 H l2 Figure 10. X-ray diffractogram of the South Pond water precipitate produced in a laboratory experiment by Upchurch, 1990. w (X)


39 Discussion Historical Sodium Concentrations Temporal changes in sodium concentrations in the aquifer may result from either changes in sodium concentrations in South Pond or the infiltration rate of the slurry from South Pond to the aquifer. Sodium concentrations in South Pond may change in response to variations in plant efficiency, which undoubtedly occur, or from dilution or concentration from increased rainfall or evaporation, respectively. Differentiating between these mechanisms is difficult lacking plant-production records and reliable measurements of dilution or concentration of sodium levels in the pond. With the exception of well 6, sodium levels in all industrial wells rose almost immediately following the beginning of ALPART operations. From 1970 to about 1972, the slope of sodium concentrations over time in wells 1, 4, and 5 are similar to dispersion-dominated breakthrough curves. The slope of the initial sodium rise in each industrial well corresponds to its distance from South Pond; that is, well 1 shows the most rapid rise in sodium levels and is the closest well to South Pond. Wells 5, 4, and 6 all show progressively slower rates of increasing sodium


levels (in that order), and each well is increasingly further from South Pond than the previous one (Figure 1) The combination of the southernmost industrial wells 40 (wells 1, 3, 4, and 5) captures most of the sodium plume; however, wells 1 and 3 capture more than wells 4 and 5 Until 1974, well 6 was completely isolated from any sodium contamination. In 1974, earthen levees of South Pond were repaired and raised, and waste was temporarily re-routed to what became North Pond. Prior to this time, North Pond did not exist as a waste-disposal facility. Repairs to South Pond continued for nine months, after which North Pond became inactive and waste slurry was again piped to South Pond. North Pond has operated only as an emergency pond since its activity in 1974 and has primarily received rainwater runoff from the ALPART plant. The plume resulting from this nine-month period of waste disposal to North Pond approached the industrial-well field from the west. The South Pond plume approaches the same well field from the south. The sodium plume emanating from North Pond bypassed the capture zones of wells 1, 3, 4, and 5 but was partly captured by well 6. Sodium levels in well 6 have steadily decreased since 1976. Well 6 seems sensitive only to the influence of North Pond, as long as the industrial-well field continues to capture the South Pond sodium plume. The rise and fall of sodium concentrations from 1980 to


41 1982 in many of the monitoring wells may be related to further repair work done on the South Pond levees. During this time, the South Pond levees underwent a series of improvements that allowed the elevation of the pond to rise. As mentioned earlier, the waste slurry effectively seals the limestone in contact with the pond and restricts the infiltration of filtrate to the aquifer. If the elevation of the pond rises, limestone previously not in contact with the slurry becomes exposed and there is more infiltration to the aquifer. The rise of sodium levels in the aquifer in 1980 may be explained by an increase in pond elevations and renewed infiltration of slurry to the aquifer. The drop in sodium concentrations in 1982 may represent the effect of the slurry sealing the newly exposed limestone. Alternately, the slug of sodium-rich water from the nine-month period of active waste disposal to North Pond may not have reached the distant monitoring wells until 1980. This would indicate a six-year travel time from the North Pond to the affected monitoring wells. This is an unrealistic amount of time, because the industrial wells showed elevated sodium levels within six months following waste disposal to South Pond. Therefore, it is more likely that construction on the South Pond levees influences sodium levels in the aquifer.


42 Chemical trends of Essex Valley aquifer Sodium concentrations decrease sharply from South Pond to the industrial wells {Group I wells) and show minor fluctuations within Group I wells {Figure Sa) Other than in South Pond and well 3, concentrations of dissolved aluminum are below detection limits in all wells. Carbonate concentrations drop sharply similar to those of sodium from South Pond to Group I wells; however, the levels are almost constant at about 4 mmol/L within Group II wells. The abrupt change in carbonate concentrations from South Pond to the aquifer may reflect the buffering capacity of South Pond. The buffering capacity of South Pond is expected to be very large due to the high carbonate concentrations. Calcium concentrations drop sharply from 2.20 to 0 .02 mmol/L from South Pond to well 3 rise to about 2.0 mmol/L within Group II wells. Fluctuations in calcium levels in Group II wells are related to the changes in log PC02 values (Figure 8b) Magnesium concentrations increase from zero in South Pond to about 0 5 in Group II wells. The high calcium and magnesium concentrations correspond to negative saturation indices of calcite and indicate that limestone dissolution is proceeding in the aquifer. The pH drops steadily"from South Pond to the industrial wells and sharply between well 6 and Group II wells (Figure 8c) In Group II wells, pH is relatively steady between 6 and 7 Log PC02 values increase proportionally with the


43 decreasing pH values. This relationship is expected because the distribution of carbon in an aqueous system is regulated largely by the pH of the solution. The calcite saturation index drops as log PC02 values increase. Tracing the wells plotted on the Piper diagram in order of decreasing chloride concentrations shows a general trend from Group I wells to Group II wells. The trend between these groups of wells is defined primarily by decreasing sodium percentage of the total meq/L. The carbonate percent of the total meq/L remains steady from one group to the other, with the exception of a shift observed in Group I wells. The difference in behavior between cations and anions can be seen in the minor triangles of the Piper diagram. Wells plotted on the anion triangle remain clustered around the 100 percent 'carbonate + bicarbonate' vertex for both groups of wells. Group I wells on the cation triangle plot entirely within the 100 percent 'sodium' vertex, but Group II wells are spread around the 'calcium' vertex. Within Group I, the shift along the carbonate axis is explained by the sharp drop in carbonate levels from South Pond to well 3. Sodium concentrations also decrease sharply from South Pond to well 3, but no shift along the sodium axis is observed on the Piper diagram. The lack of a shift along the sodium axis is explained by the fact that, in South Pond, only iron and dissolved aluminum are found in


44 concentrations comparable to those of sodium. Neither of these two constituents is plotted on the Piper diagram, and thus the sodium percent of the total meq/ L remains near 100. The sodium percent of total meq/ L decreases from Group I wells to Group II wells because the calcium and magnesium levels are higher than sodium in Group II wells, and sodium becomes a relatively minor constituent. All samples taken from Group II wells indicate a calcium-bicarbonate type water. The relative position of individual Group II wells on the Piper diagram is determined by the changes in calcium and magnesium concentrations. Tracing a flowpath from one Group II well to another using chloride concentrations as an indicator may be misleading because the chloride concentrations range only from 0.49 to 0.16 mmol/L (17 to 6 mg/L). Because chloride concentrations are subject to analytical error and probably fluctuate naturally to some extent, Group I and II wells are presented in Table 4 and Figure 11 in order of decreasing specific conductivities and total dissolved solids for comparison. The order of Group I wells, including well 14, remains unchanged in all three cases. Listing Group II wells in order of specific conductivity approximates the relative locations of the wells in order of increasing distance from South Pond, with the exceptions of wells 7, 9, and 22. Wells 7 and 9 are industrial wells (ALPART 'Pepper' well field) and the low


45 Table 4. ALPART wells listed in the orders of decreasing specific conducti vities 1 total dissolved solids I and chloride concentrations. Chlori de Specific Cond. Total Diss Solids (mmoi/L) (umhos) Cmo/L) Group I Wells S.P. ( 205) S .P. ( 21300) S.P. ( 34460) 3 ( 66.4) 3 ( 3360) 3 ( 2332) 4 ( 48.1 ) 4 ( 2430) 4 ( 1665) 12 ( 33 .5) 12 ( 1380) 12 ( 1010) 6 ( 25.8) 6 ( 1025) 6 ( 765) 14 (18.6) 14 ( 610) 14 ( 455) Group II Wells 20 ( 17.4) 16 { 525) 16 { 402) 21 ( 13 .9) 20 ( 510) 21 ( 369) 7 { 13.6) 21 ( 500) 20 ( 361 ) 9 { 12.4) 18 ( 490) 22 ( 361 ) 16 { 9.9) 7 ( 485) 7 { 357) 22 { 9 .4) 22 ( 480) 18 ( 343) 18 ( 8 .7) 9 { 450) 9 ( 339) 13 ( 5.7) 13 { 139) 13 { 114)


4000 3500 3000 C/J 2500 "0 c:: 2000 ,; c:: 8 ci. 1500 C/J 1000 500 0 46 2 1.5 1 0.5 0 S. P 3 4 12 6 14 20 21 7 9 16 22 18 Well numbers (decreasing Cl -->) Sp. Con d (umbos) -+-TDS ( mg!L) __._ Cl ( mmoi/L) Figure 11. Comparison between the changes in chloride concentrations, specific conductivity, and total dissolved solids in Essex Valley aquifer. c= 9 ...... .... ...... c= Q) u c= 0 u 0


47 specific conductivity of the water in these wells may be influenced by factors other than advection. Well 22 lies northeast of and possibly upgradient of South Pond and, historically, sodium concentrations have not risen above background levels until 1989. It is not clear whether this slight rise in sodium levels is due to South Pond activity or not. The order of decreasing total dissolved solids also agrees well with the approximate location of Group II wells with increasing distance from South Pond if wells 7 and 9 are ignored for the same reasons stated above. Both specific conductivity and total dissolved solids will approximate the order of mixing between two waters, assuming that all dissolved ionic constituents behave conservatively. As listed in Table 3, the groundwater is undersaturated with respect to calcite in well 14 and in all Group II wells. Wells 16 and 20 show the lowest saturation of calcite at -1.09 and 1 .12, respectively (Table 3). Both of these wells are among the highest in specific conductivity and total dissolved solids within Group II wells (Table 4) Limestone dissolution occurring in the vicinity of Group II wells will influence the values for both specific conductivity and total dissolved solids of the aquifer. Chloride concentrations should remain unaffected by the dissolution of the Newport Formation because only trace levels of chloride are found in this limestone. Chloride levels, therefore, provide the best indication of


advection processes within Group II wells, despite the low concentrations. Chemical Surpluses and Deficits 4 8 Equations 2 and 3 quantify the nonconservativ e nature of several chemical constituents found in the Essex Valley aquifer system (Table 5; Figures 12 and 13). Sodium, carbonate, and dissolved aluminum are of particular concern, due to their large deficits, although calcium, magnesium, and sulfate also behave nonconservatively. The largest deficits for sodium, carbonate, and dissolved aluminum are found in well 3; however, the majority of removal of these constituents from solution occurs between South Pond and well 3. The large deficits of sodium, carbonate, and dissolved aluminum indicate that these constituents are being removed from the aquifer system in excess of that which can be attributed to advection. Here, "aquifer s ystem" refers to South Pond, the vadose zone directly beneath the South P ond, and the Essex Valley aquifer north of South Pond. It is difficult to determine exactly where the sodium, carbonate, and aluminum deficits are occurring without chemical data from the vadose zone or the aquifer below South Pond. Any chemical constituent removed from South Pond by chemical reactions (precipitation, sorption, etc.) will create a deficit, and this deficit will accumulate through


Table 5. Calculated surpluses a n d d e f i cits for all major dissolved constituents i n E ssex Valley aquifer -I Na/C03 Sample/well Percent Na C03 AI Ca Mg S04 Calcit e number Sout h Pond S .l. Group I Wells S .Pond 100. 0 0 .00 0 .00 0 .00 0 0 0 0 .00 0 .00 2 6 ---3 30.46 -125.03 -35.08 -34.89 -0.65 -0. 01 2.31 0.38 3.56 4 21.37 -86.81 -22.15 -24.56 -0.45 0.00 0.37 0 .44 3 .92 12 13.95 -57.53 -14.76 -16.20 0 .22 0 .02 0 .34 0 9 3 .90 6 10.09 -41.32 -10.09 -11.70 0 01 0 .20 -0.46 0 .77 4 .10 14 6 .47 -29.28 -6.30 7 .50 0 .92 0 .20 0 .54 -0.78 4 6 5 Group II Wells 20 5 .87 -29.38 6 .25 6 .80 1.83 0 4 3 0.49 -1.12 4. 7 0 21 4 1 1 -20.22 3 .07 -4.75 1. 8 1 0.46 0.34 -0.69 6 5 9 7 3 .96 -18.74 3 .08 4 .58 1 .40 0 .40 0 .26 0 .69 6 .08 9 3 .36 -16.09 -2.16 -3.88 1.4 4 0 .48 0 2 3 -0. 5 7 .45 16 2 .11 -9.34 0 .67 -2.42 1 .92 0 .49 -0.16 -1.09 -13.94 22 1.86 -8.91 0 91 -2.13 1.97 0 .14 -0.15 -0.65 9. 7 9 18 1 51 7 .39 1 .21 -1.72 2 .11 0 2 1 0 .11 0 .54 -6. 11 13 0 .00 0 .27 1 .05 0 .03 0 5 0 0 .46 0 .03 0 .92 0.26 ( -) defi cit, ( + ) surplus, ( 0 ) conservat ive **all concentrations in mmoi /L Na/A L ---3 .58 3.53 3.55 3 .53 3.90 4.32 4.26 4 .09 4. 1 5 3 .86 4 .18 4 .30 9 .00 \D


50 0 0 '-" c;l) ::s e. -50 ::s 00 M 0 ..... -u I;:: 0 Q -100 -150 3 4 12 6 14 20 21 7 9 16 22 18 13 Well number (--> decreasing chloride cone ) 888 Sodium rz.J Carbonate Aluminum Figure 12. The sodium, carbonate, and dissolved aluminum deficits and surpluses in Essex Valley aquifer. Positive values indicate surpluses and negative values indicate deficits. (.11 0


3 2 -.. s 0 8 8 1 .......... ::s e. ::s (/) .... 0 0 ... -u l+=l 0 0 -1 -2 Figure 13. 3 4 12 6 14 20 21 7 9 16 22 18 13 Well number (--> decreasing chloride cone .) 888 Calcium l2:J Magnesium Sulfate -a-Calcite S l. The calcium, magnesium, and sulfate deficits and surpluses compared to the saturation index of calcite. Ul ......


52 the entire aquifer system. The ratios of the sodium deficits to the carbonate and aluminum deficits are about 4:1 in each Group I sample (Table 5). In Group II samples, the ratios of the sodium to aluminum deficits are also about 4:1; the ratios of the sodium to carbonate deficits range from about -14: 1 to 7 : 1 All Group II wells are undersaturated with respect to calcite. Dissolution of the host limestone will add carbonate to the aquifer system and thus lead to a reduction of the carbonate deficit and an increase in the sodium to carbonate deficit ratio. A negative ratio results when carbonate shows a surplus. The consistency of the deficits in each well, with the exception of carbonate, suggests that these constituents are largely being removed prior to reaching well 3 The resulting deficits are then being maintained through the aquifer system. This consistency in deficits indicates that, although sodium and dissolved aluminum are behaving nonconservatively prior to well 3, they are relatively nonreactive chemically within the aquifer. Carbonate behaves nonconservatively throughout the aquifer system. Calcium, magnesium, and sulfate also behave nonconservatively in the aquifer system but to a much lesser extent than sodium or carbonate. The deficits and surpluses of these three constituents are shown in Figure 13 with the saturation index of calcite. Calcium shows a deficit in Group I wells and a relatively large surplus in all Group II


53 wells. Magnesium also shows a surplus in the Group II wells. Both surpluses correspond to the drop in the saturation index of calcite and indicate that dissolution of the host limestone is proceeding in the vicinity of Group II wells. Sulfur species in South Pond should be dominated by s2-, because of the very reducing conditions (Hem, 1989) Upon reaching the aquifer, s2 -should be oxidized to sulfate. The large surplus of sulfate seen in well 3 is a result of this redox reaction. The sulfate deficits in Group I and II wells may also be a product of similar sulfate reduction reactions. Mechanisms of sodium fixation Although the deficit of sodium is the largest of any chemical constituent found in South Pond, the carbonate and dissolved aluminum deficits must also be addressed. Mineral precipitation and sorption reactions which may remove sodium, carbonate, and aluminum from solution are of particular interest, because none of these constituents participate in redox reactions. Precipitation reactions in the aquifer system are governed by thermodynamic rules and can be predicted by calculating the saturation index of particular minerals. The computer programs PHREEQE and WATEQ4F were used to facilitate the calculation of saturation indices for


5 4 calcite, gibbsite, and several sodium carbonates, including nahcolite (NaHC03 ) trona (NaHC03 Na2C 03H20 ) natron (Na2C 03H20), and dawsonite ( NaAlC 03). The saturation states o f water with respect t o these minerals in the aquifer system are listed in Table 3 A p ositive saturation inde x indicates t h e p otential for precipit a t i o n Daw s onite i s the only sodium carbonate with a positive saturation index, and it may be precipitating in South Pond or near well 3 (Table 3 ) South Pond is also oversaturated with respect to calcite and gibbsite. Of the sodium carbonates listed here, trona and dawsonite are of particular interest, because the white precipitate from Upchurch's experiment consisted of trona and dawsonite almost entirely. Also, Yong and others (1989) found trona and thermonatrite (dehydrated natron) in dried slurry samples from similar waste ponds in Jamaica. An analysis of the crust that forms over South Pond during t h e dry season had not been performed at the time of this writing. Most sodium carbonates, including the ones listed here, are considered evaporites and can be very soluble. Their precipitation within the aquifer system is, therefore, likely limited to the surface of South P ond and the underlying vadose zone. Many authors have reported the behavior of various sodium carbonates. Goudarzi (1970) observed natron crystals forming in Lake Magadi at night; however, as temperatures


55 rose during the day, the crystals quickly effloresced to thermonatrite. Eugster (1980) mentions the transition of natron to trona with the addition of carbon dioxide in Lake Magadi water samples. Trona is found as lake-bottom sediments in the crystalline form or as flaky surficial deposits that are easily carried away by winds (Eugster, 1980). Thick deposits have been found in Lake Magadi, the Marada salt flats, and many playa lakes on all continents (Eugster, 1980; Goudarzi, 1970). Dawsonite is typically associated with hypersaline, evaporative environments and with hydrothermal activity (Smith and Milton, 1966) Significant deposits have been found in the Green River Formation in Colorado, the Pripyat depression of Belorussia (of the former U.S.S. R .), the Sydney Basin of Australia, and minor deposits across the globe (Smith and Milton, 1966; Dmitriev and others, 1975; Golbery and others, 1977). The precipitation of a mineral will remove its constituent elements in fixed proportions. The precipitation of calcite, for example, will remove one mole of calcium for every mole of carbonate. The reduction of carbonate in the South Pond due to calcite formation can be ignored, however, because calcium concentrations are very low. Applying a mass-balance approach to the Essex Valley aquifer system can help assess the mechanisms responsible


56 for the sodium, carbonate, and aluminum deficits. The deficits of sodium, carbonate, and dissolved aluminum found in well 3 are 125, 35, and 35 mmol /L, respectively (Table 5). If the removal of these ions occurs in South Pond, the amount needed to be removed to lead to the deficits in well 3 can be calculated by rearranging Equation 2. For example, a deficit of zero for any constituent in well 3 will yield the equation: where Xw PsPW XsPW and XB = = = = Xw [ ( 1 -PsPw ) XB ] the observed concentration in well 3, the percent South Pond water, the concentration in South Pond, the background concentration. For the sodium, carbonate, and aluminum deficits to equal zero in well 3, the required concentrations of each (4) element in South Pond (XsPW) are about 98, 57, and 2 mrnol/L, respectively. These calculated concentrations are then subtracted from the observed concentrations in South Pond to obtain the amount of each constituent needed to be removed to yield a zero deficit in well 3 These amounts are referred to as the true deficits and equal 410 mrnol/L sodium, 235 mrnol/L carbonate, and 114 mmol/L aluminum. For example, 410 mmol/L os sodium must be removed from South Pond to eliminate the deficit observed in well 3. Precipitation of trona will remove three moles of sodium for every two moles of carbonate; dawsonite will


57 remove one mole of sodium for every mole of carbonate and aluminum. A conservative estimate of the amount of sodium removed by precipitation of sodium carbonate is obtained b y assuming that only trona will precipitate in South Pond. Assuming that this reaction will proceed until a carbonate reservoir equalling the size of the carbonate deficit (235 mmol /L) is depleted, 58 mmol/ L of the sodium deficit remains unexplained. A more realistic scenario involves a combination of the precipitation of both trona and dawsonite, as indicated by the X-ray diffractogram (Figure 10). The precipitation of dawsonite is limited by aluminum. If only dawsonite precipitated in South Pond until 114 mmol / L of carbonate and aluminum was depleted, about 296 mmol/ L of sodium and 121 mmol/L of carbonate remains unexplained. The sodium deficit cannot be explained solely by the precipitation of a sodium carbonate; therefore, a second mechanism of sodium fixation must exist within the aquifer system. The aluminum deficit, although possiblyaccounted for by dawsonite precipitation, may also be explained by the precipitation of gibbsite or the formation of amorphous aluminum tri-hydrate gels (Wefers and Bell, 1972). These tri-hydrates may lead to gibbsite formation upon aging and crystallization. Several authors indicate that sodium must be incorporate into the crystal-lattice structure of gibbsite for precipitation to proceed (Wefers and Bell,


58 1972). Therefore, the formation of an aluminum precipitate, such as gibbsite, may partially account for the sodium deficit. Gibbsite has an isoelectric point (zero point of charge) of nine; above a pH of nine, positive ions will adsorb onto a gibbsite crystal, and below this pH, negative ions will be adsorbed (Sposito, 1983). South Pond has a pH range of 11 to 13, and gibbsite has been recovered from the filtrate of samples taken in 1990 (Gurr & Assoc., Inc., 1990). Therefore, sodium may be sorbed onto gibbsite crystals forming in South Pond. Sorption of a cation onto a compound requires the desorption of another cation; therefore, any sodium removed from solution by sorption reactions will be accompanied by the addition of another cation to solution. The observed decrease in pH from South Pond to the aquifer could be partially explained by the displacement of hydrogen ions by sodium, in addition to groundwater equilibration with the host limestone. There is no indication of desorption of other cations in the aquifer system. Gibbsite may also be found in the aquifer in the form of lateritic deposits introduced along solution features, fractures, and faults. Limestone vugs exposed in roadcuts in Essex Valley are filled with terra rossa, which is an aluminum-rich soil. Sorption of sodium onto gibbsite found in terra rossa at depth is, therefore, a possible mechanism


59 for sodium fixation in the aquifer. The current theory concerning the formation of terra rossa describes this soil as a eolian deposit or limestone dissolution residual material. The remaining 58 mmol/L of sodium left unexplained by the precipitation of trona in South Pond may be reacting with aluminum species, either by co-precipitation or sorption reactions. Continuing with the conservative assumption that trona is the single sodium carbonate forming in South Pond, an aluminum deficit of 114 mmol/L remains unexplained. The sodium to aluminum deficit ratio needed to account for the remaining 58 mmol/L of sodium with reactions involving dissolved aluminum species is about 1 :2. Chemical modeling The first PHREEQE model was constructed to demonstrate the changes in chemistry from only advective processes. WATEQ4F was used to determine the saturation indices and other chemical properties of the waters in each well. A second PHREEQE model was constructed to assess the effect of removing sodium, carbonate, and dissolved aluminum from South Pond in amounts equalling their respective true deficits. Figures 14a-c show results of the two models in comparison to the observed chemical data. Model 1 does not reproduce the sodium, carbonate, or dissolved aluminum concentrations found in the aquifer


60 system. By removing the calculated true deficits from South Pond prior to mixing, the observed concentrations in the aquifer are matched very closely with Model 2 Model 2 preferentially removes the true deficits from South Pond to simulate the precipitation of sodium carbonates and sorption/precipitation reactions involving sodium and aluminum. No mineral is held at equilibrium, and the pH of each solution is not adjusted to correct any electrical charge imbalance. The decreasing sodium concentrations in the aquifer are almost completely explained by advection after removing the true deficit from South Pond (Figure 14a) The sodium concentrations of Model 2 deviate slightly from the observed concentrations in well 4 and Group II wells and may represent minor amounts of sodium fixation within the aquifer. Since the aquifer is undersaturated with respect to sodium carbonates, sorption reactions may be responsible for the small deviations. The predicted carbonate levels deviate from observed levels in wells 16, 22, and 18 of Group II and may represent carbonate surpluses (Figure 14b) Group II wells are all undersaturated with respect to calcite, and limestone dissolution is probably occurring near wells 16, 22, and 18. Aluminum shows no significant deviations in Model 2 from the observed concentrations (Figure 14c) The absence of deviations indicates that nearly all of the dissolved


1000 100 0 8 8 "-" u 10 c:: 0 0 8 ;:I .... "'0 0 CJ') 1 0.1 3 Figure 14. 4 12 6 14 20 21 7 9 16 22 18 13 Well numbers (decreasing Cl cone. -->) ----Observed --+Model 1 _.,._ Model 2 (a) Results from the PHREEQE Models 1 and 2 compared to observed data. 0'1 I-'


1000 -.. s 100 0 8 e '-"' u c::; 0 u Cl) ... c::; 0 -e a 10 1 3 4 12 6 14 20 21 7 9 16 Well numbers (decreasing Cl cone. -->) -Observed -+Model 1 --ltModel 2 (b) Figure 14. (cont'd). 22 18 13 0\ N


0 a a ...._, c.) c:: 0 0 a ::s .s a ::s < 40 30 20 10 0 -10 3 4 12 6 14 20 21 7 9 16 Well numbers (decreasing Cl cone.-->) Observed -+Model 1 ___._ Model 2 (c) Figure 14. (cont'd). 22 18 13 0\ w


12 -1 -2 .______ 11_3 ; !"=:..,_. \ 111 I .. \ I I 10 91 0.. -4 -5--. -6 -c:r lJ -7 8 . i 71 I 6 __ _. __ -8 S.P. 3 4 12 6 14 20 21 7 9 16 22 18 13 Well numbers (decreasing Cl cone. -->) observed pH -+Model 2 pH _..,._ observed log PC02 -aModel 2 log PC02 (d) Figure 14. (cont'd). m


65 aluminum found in South Pond is removed from solution before entering the aquifer. Aluminum may be removed by the precipitation of dawsonite or the precipitation of gibbsite and amorphous aluminum species. Observed and modeled log PC02 and pH values are shown in Figure 14d. Predicted pH and PC02 values do not change upon mixing South Pond water and aquifer water in either PHREEQE model. This discrepency is a result of the very large buffering capacity of South Pond, despite the removal of 235 mmol/L of carbonate (80 percent) from South Pond prior to mixing. The observed PC02 levels may decrease more rapidly as a result of microbial respiration, although no data concerning the biota of South Pond are available. Microbial respiration would add dissolved C02 to the solution and the pH would decrease as a result. This process may also be occurring within the vadose zone or the aquifer below South Pond.


66 Swmnary Sodium concentrations have been monitored in the industrial and monitoring wells at the Aluminum Partners of Jamaica (ALPART) plant since 1969, when plant production began. The ALPART plant extracts aluminum from bauxite ore and produces a caustic slurry as a by-product. The slurry is disposed the primary waste pond, called South Pond. South Pond is perched about 200 meters above Essex Valley aquifer, and caustic waste infiltrating through the porous limestone has significantly increased sodium concentrations in the aquifer. North Pond is a secondary waste pond and is located west of the ALPART plant and northof South Pond. Historical records of sodium concentrations in four industrial wells and 15 monitoring wells indicate that the cone of depression around the active industrial wells captures a large portion of the plume emanating from the South Pond. North Pond was active for a nine-month period in 1974, and the resulting plume approached the well field from the west and caused sodium concentrations to rise sharply in industrial well 6 Prior to this time, well 6 had been protected by the other four industrial wells and had showed only background-level sodium concentrations. Sodium levels in well 6 have steadily dropped since 1976 and indicate that North Pond does not contribute a significant


67 amount of sodium to the aquifer. The chemical analyses of all well samples and South Pond define two chemically distinct types of water within the aquifer system. These two groups are identified on a Piper diagram and are referred to as Groups I and II. The water in Group I wells is a sodium-carbonate type. This group includes South Pond, the industrial wells, and wells 12 and 14. Water i n Group II wells is a calcium-bicarbonate water; this group includes all remaining monitoring wells. A flowpath can be traced from each well on the Piper diagram using chloride concentrations, which decrease from Group I to Group II in a succession approximating the progressive increase in distance from South Pond. Chloride concentrations decrease from South Pond to the aquifer by 70 percent; however, chloride concentrations change in response to advection only. The concentrations of any constituent acting conservatively should decrease with chloride concentrations. Sodium levels drop from South Pond to the aquifer by over 90 percent and indicate that sodium is being removed from solution. The discrepancy between changes in sodium and chloride concentration indicate that a sodium deficit exists between South Pond and the aquifer, and that sodium is chemically reactive between South Pond and the aquifer. Ratios of sodium deficits to carbonate and aluminum deficits for each well are consistent at about 4:1. The


consistency of these ratios indicates that sodium, carbonate, and aluminum are being removed from South Pond prior to mixing with the formation water and that the resulting deficits are maintained through the aquifer. Sodium and aluminum, therefore, do not appear to be significantly chemically reactive beyond South Pond. Carbonate is added to the aquifer in Group II wells from limestone dissolution. A mass-balance approach can help assess the mechanism responsible for the removal of sodium, aluminum, and carbonate from the aquifer system. The precipitation of sodium carbonates cannot account for the entire sodium deficit, and thus other reactions must be considered. Sodium may coprecipitate with or adsorb onto gibbsite crystals, which have been recovered from the filtrate of samples taken from South Pond. Amorphous aluminum species found in South Pond may also act as sorption sites for sodium. 68 Chemical modeling indicates that the removal of the true deficits of sodium, carbonate, and aluminum exclusively from South Pond largely eliminates the deficits calculated in the aquifer. The lack of a deficit in the aquifer after removal from South Pond indicates that sodium and aluminum act conservatively within the aquifer. Carbonate is chemically active within Group II wells and represents limestone dissolution.


69 The observed PC02 and pH values are not approximated by removing 235 mmol / L of carbonate from South Pond. The discrepancy is a result of the large buffering capacity of the South Pond. The addition of an acid, possibly from microbial respiration, may remove the buffering capacity to allow PC02 and pH to change upon mixing with the background water. The exact location of where this reaction occurs cannot be determined but is limited to within South Pond, or the vadose zone, or the aquifer beneath South Pond.


LIST OF REFERENCES Ball, J., and D K Nordstrom, 1991. User's manual for WATEQ4F, with revised thermodynamic database and test cases for calculating speciation of major, trace, and redox elements in natural waters, U.S. G S Open-File Report 91-183. Boggs, Sam, Jr., 1972. Petrology of Sedimentary Rocks, Macmillian Publishing Company, New York, pg. 569-570. 70 Da vis, J C 1986. Statistics and Data Analysis in Geology, John Wiley and Sons, New York, 2nd ed. 646 p. Domenico, P and W. Schwartz, 1990. Physical and Chemical H ydrogeology, John Wiley and Sons, New York, 824 p. Eugster, H.P., 1980. Lake Magadi, Kenya, and its precursors, in Hypersaline Brines and Evaporitic En vironments, in Sedimentology 28, A. Nissenbaum (ed), Elsevier Scientific Publishing Company, Amsterdam, pp. 195-2. 30 Feth, J 1981. Chloride in natural continental water-a review, U.S.G.S. Water-Supply Paper 2176. Freeze, A and J. Cherry, 1979. Groundwater, Prentice-Hall, Inc., Englewood Cliffs, 'New Jersey, 604 p. Golbery, R. and F.C. Loughnum 1977. Dawsonite, alumohydrocalcite, nordstrandite, and gorceixite in Permian marine strata of the Sydney Basin, Australia, Sedimentology, vol. 24 (4), pp. 565-79. Goudarzi, G 1970. Nonmetallic mineral resources: Saline deposits, silica sand, sulfur, and trona, in Geology and mineral resources of Libya -A reconnaissance, U.S. G .S. Professional Paper 660. Gurr & Associates, Inc. 1990. ALPART North Mud Lake Risk Assessment, vols. I-III. Gurr, T.M., 1992, personal communication.


Hanshaw, W and William Back, 1979. Major geochemical processes in the evolution of carbonate-aquifer systems, Journal of Hydrology, vol. 43, pp. 287-312. 71 Henderson, T., 1984. Geochemistry of ground-water in two sandstone aquifer systems in the Northern Great Plains in parts of Montana, Wyoming, North Dakota, and South Dakota, U.S.G.S. Professional Paper 1402-C. Hem, John, 1985. Study and interpretation of the chemical characteristics of natural water, U S .G.S. Water-Supply Paper 2254, third edition. Hingston, F., A. Posner, and J. Quirk, 1972. Anion adsorption by goethite and gibbsite, I The role of the proton in determining adsorption envelopes, Journal of Soil Science, vol. 23, no. 2, pp. 177-192. Hull, L., 1984. Geochemistry of ground water in the Sacramento Valley, California, U .S.G.S. Professional Paper 1401-B. Kimmel, G., and 0. Braids, 1980. Leachate plumes in ground water from Babylon and Islip Landfills, Long Island, New York, U .S.G. S Professional Paper 1085. Knobel, L., and S. Phillips, 1988. Aqueous geochemistry of the Magothy Aquifer, Maryland, U.S.G.S. Water-Supply Paper 2323. Kohout, F .A., M.C. Kolipinski, and A.L. Higer, 1973. Remote sensing of submarine springs: Floridan Plateau and Jamaica, West Indies, (1). Proc. 2nd Int'l Symp on Ground-water, Int. Assoc. Hydrol. pp. 571-578. Lewis, J., and G Draper, 1988. Geology and tectonic evolution of the Northern Caribbean margin, in The Geology of North America, vol. H, The Caribbean Region, pg. 77-140, G.S.A., 1990. Monnin, C and J. Schott, 1984. Determination of the solubility products of sodium carbonate minerals and an application to trona deposition in Lake Magadi, Kenya, Geochimica et Cosmochimica Acta, vol. 48, pp. 571-581. Morse, J., and F Mackenzie, 1990, Geochemistry of Sedimentary Carbonates, Developments in Sedimentology 48, Elsevier, Amsterdam, pp. 1-131.


72 Muhs, D., C. Bush, K Stewart, and R Crittenden, 1990. Gechemical evidence of Saharan dust parent material for soils developed on Quaternary limestones of Caribbean and Western Atlantic Islands, Quaternary Research, vol. 33, pp. 157-177. Parkhurst, D D. Thorstenson, and L .N. Plummer, 1990. PHREEQE A computer program for geochemical calculations, U S .G.S. Water Resources Investigations 80-96, fifth edition. Patterson, R., R. Jackson, B Graham, D Chaput, and M. Priddle, 1985. Retardation of toxic chemicals in a contaminated outwash aquifer, Water Science Technology, vol. 17, pp. 57-69. Plummer, L.N., 1977. Defining reactions and mass transfer in parts of the Floridan Aquifer, Water-Resources Research, vol. 13, no. 5, pp. 801-812. Plummer, L.N., and W Back, 1980. The mass balance approach: Applications to interpreting the chemical evolution of hydrologic systems, American Journal of Science, vol. 280, pg. 130-142. Plummer, L.N., E. Prestemon, and D Parkhurst, 1991. An interactive code (NETPATH} for modeling NET geochemical reactions along a flowPATH, U.S.G. S Water-Resources Investigations Report 91-4078. Robinson, E., 1990. Report on the geology of the Essex Valley Region, St. Elizabeth, Jamaica, with particular reference to structure and groundwater movement, in Gurr & Associates, Inc., ALPART North Mud Lake Risk Assessment, vol. !!-report, 1990. Snoeyink, V., and D. Jenkins, 1980, Water Chemistry, John Wiley and Sons, Inc., New York, pp. 197-315. Smit, w and C. Holten, 1980. Zeta-potential and radiotracer adsorption measurements on EFG a-Al203 single crystals in NaBr solutions, Journal of Colloid and Interface Science, vol. 78, no.1, pp. 1 -14. Smith, D.L., 1970. The residual hypothesis for the formation of Jamaican bauxite a consideration of the rate of limestone erosion, Geological Society of Jamaica, vol. 11, pp. 3-12.


7 3 Smith, J. and C Milton, 1966 Dawsonite in t h e Green River Formation of C olorado, Economic Geology, vol. 61, pp. 1029-1042. Sposito, G., 1989, The Environmental Chemistry o f Soils, CRC Press, Inc., 317 pgs. Sprinkle, C., 1989. Geochemistry of the Floridan Aquifer System in Florida and in parts of Georgia, South Carolina, and Alabama, U S .G.S. Professional Paper 1403-I. Sprycha, R., 1983. Attempt to estimate the charge components on oxides from anion and cation adsorption measurements, Journal of Colloid and Interface Science, vol. 96, no. 2, pp. 551-554. Stumm, W., and J Morgan, 1962. Chemical aspects of coagulation, Journal of American Water Works Association. Tollenaar, P., 1990. Water Resources Survey of the Upper Black River Basin, St. Elizabeth, Inventory of Groundwater Resources, in Gurr & Associates, Inc. ALPART North Mud Lake Risk Assessment, vol. III, 1990. Upchurch, S.B., 1992, personal communication. Upchurch, S.B., Grain-surface chemistry in ground-water systems, in Ground-Water Geochemistry, Chapter 3, in prep. Upchurch, S B 1990. Hydrogeochemical processes in the karst ground-water system at the ALPART Plant, Essex Valley, Jamaica, in Gurr & Associates, Inc. ALPART North Mud Lake Risk Assessment, vol. III, 1990. van Straten, H., B Holtkamp, and P. DeBruyn, 1984. Precipitation from supersaturated aluminum solutions, I Nucleation and growth of solid phases at room temperature, Journal of Colloid and Interface Science, vol. 98, no. 2, pp. 342-362 Wedderburn, L., 197 7 Groundwater pollution of a limestone aquifer by caustic waste, in Hydrologic Problems in Karst Regions, R Dilamarter (ed), Western Kentucky University, pp. 303-313. Wefers, K and G Bell, 1972. Oxides and Hydroxides of Aluminum, Technical Paper no. 19, ALCOA Research Laboratories.


Wood, W and W. Low, 1988. Solute geochemistry of the Snake River Plain Regional Aquifer System, Idaho and Eastern Oregon, U .S.G. S Professional Paper 1408-D. 74 Yong, R.N., R.D. Ludwig, and A.S. Wagh, 1986. Problems in predication of crustal properties of disposed red mud, Journal of Geological Society of Jamaica, Proc. Bauxite Symp. VI. March 17-22, 1986, pp. 177-187. Zack, A., and I. Roberts, 1988. The geochemical evolution of aqueous sodium in the Black Creek Aquifer, Horry and Georgetown Counties, South Carolina, U.S.G.S. Water Supply Paper 2324


Download Options

No images are available for this item.
Cite this item close


Cras ut cursus ante, a fringilla nunc. Mauris lorem nunc, cursus sit amet enim ac, vehicula vestibulum mi. Mauris viverra nisl vel enim faucibus porta. Praesent sit amet ornare diam, non finibus nulla.


Cras efficitur magna et sapien varius, luctus ullamcorper dolor convallis. Orci varius natoque penatibus et magnis dis parturient montes, nascetur ridiculus mus. Fusce sit amet justo ut erat laoreet congue sed a ante.


Phasellus ornare in augue eu imperdiet. Donec malesuada sapien ante, at vehicula orci tempor molestie. Proin vitae urna elit. Pellentesque vitae nisi et diam euismod malesuada aliquet non erat.


Nunc fringilla dolor ut dictum placerat. Proin ac neque rutrum, consectetur ligula id, laoreet ligula. Nulla lorem massa, consectetur vitae consequat in, lobortis at dolor. Nunc sed leo odio.