Ground Water for Public Supply in St. Croix Virgin Islands GEOLOGICAL SURVEY WATER-SUPPLY PAPER 1663-D Prepared in cooperation with the Government of the Virgin Islands Ground Water for Public Supply in St. Croix Virgin Islands By G. E. HENDRICKSON CONTRIBUTIONS TO HYDROLOGY OF LATIN AMERICA AND THE ANTILLES GEOLOGICAL SURVEY WATER-SUPPLY PAPER 1663-D Prepared in cooperation with the Government of the Virgin Islands UNITED STATES GOVERNMENT PRINTING OFFICE, WASHINGTON : 1963 UNITED STATES DEPARTMENT OF THE INTERIOR STEW ART L. UDALL, Secretary GEOLOGICAL SURVEY Thomas B. Nolan, Director For sale by the Superintendent of Documents, U.S. Government Printing1 Office Washington, D.C., 20402 CONTENTS Page Abstract___--_--__------__-_________________--_-_---___-_-______ Dl Introduction.____________-----__________-____-___-__--_-__________ 1 Geography and climate_----__--___-_______-_-_____-_--_-____--___ 2 Geologic setting___________________________________________________ 3 Surface water_____________--_-_______-_-_-_____-_-_-_-_--_______-_ 4 Water-bearing formations__.-----_-__-______-_-_______--__--__-__- 4 Mount Eagle Volcanics_---_---____-_________--___-__-____-_--__ 6 Intrusive rocks_______________________________________________ 6 Sedimentary rocks__---_---_-____-_-___-____________--__-_-____ 7 Jealousy Formation,.______________________________________ 7 Kingshill Marl.___________________________________________ 7 Alluvium. ______-__----_-_-__________-_____-_-_________-__ 8 Public-supply well fields.-__--.-_______-___--_-__________-_-_-___-_- 9 Concordia well field_____-__-___________________-____-__-_______ 9 Mahogany Road well field.-_-_-___-_-_-_-___--______---_-_____- 13 Barren Spot well field________.__________________.______________ 16 Adventure, Golden Grove, and Manning well fields________________ 22 References cited._______________________________________.__.______. 27 ILLUSTRATIONS Page FIGURE 1. Availability of ground water in St. Croix, Virgin Islands.____ D5 2. Concordia well field______________________________________ 10 3. Mahogany Road well field_______________________-_-__---_ 14 4. Barren Spot well field.____________________________-______ 17 5. Adventure, Golden Grove, and Manning well fields---------- 23 TABLES Page TABLE 1. Concordia well field_______._________-______------_----- Dll 2. Mahogany Road well field___-__________________---_-_---- 15 3. Barren Spot well field___-__.---.________________. 18 4. Adventure, Golden Grove, and Manning well fields_--__--_--_ 24 in CONTRIBUTIONS TO HYDROLOGY OF LATIN AMERICA AND THE ANTILLES GROUND WATER FOR PUBLIC SUPPLY IN ST. GROIX, VIRGIN ISLANDS By G. E. HENDRICKSON ABSTRACT The ground-water resources of St. Croix, V.I., if properly developed should be adequate to supply the present and near-future demand for water for public supply. Ground water is obtained from weathered volcanic and intrusive rocks (bedrocks), from limestone, and from alluvium. The water obtained from weathered bedrocks and from alluvium in the valleys of the North Side Range is generally of better quality than that obtained elsewhere on the island. The greatest yields are obtained from the limestone and alluvium of the south- central coastal plain, but the water there is generally of poor to only fair quality. In the East End area no water, or salty water, is obtained from wells. INTRODUCTION The study described in this report was made during the period March 12 to June 1, 1962, at the request of the Hon. Ralph M. Paie- wonsky, Governor of the Virgin Islands. The purpose of the study was to determine the potential yield of several recently drilled test wells that are to be used for public water supply and to obtain in- formation needed to plan further studies of the water resources of the island. A field inventory of more than 130 wells was made, including most of the wells drilled on the island since the study made in 1938-39 by Cederstrom (1950). Depths to water were mQasured where possible, and field determinations of the chloride content of the wrater were made wherever samples could be obtained. P-imping tests were made on eight of the test wells intended for the public supply. An open-file report (Heiidrickson, 1962) was prepared for early use by Virgin Islands officials. This report contains an appen- dix which includes data on the wells inventoried, logs of test wells, a summary of pumping tests, and suggestions for obtaining continu- ous records of ground-water supplies. Previous reports on the ground-water resources of St. Croix include a published report by Cederstrom (1950), a brief open-file report by Dl D2 HYDROLOGY OF LATIN AMERICA AND THE ANTILLES McGuinness (1953), and consulting reports by Malcolm Pimie (1945) and by Tippets and others (1959). Cederstrom's report includes a description of the geology and a geologic map. His report sum- marizes literature on physiography and geology and contains almost all the information available on ground-water resources at the time his study was made (1938-39). The reports by Pirnie and by Tippets and others are chiefly summaries of information obtained in earlier studies, but they include specific recommendations for development of water supplies on the island. Appreciation is expressed for the cooperation of officials of the Vir- gin Islands Public Works Department in carrying out this study. Mr. Alfonso Paraliticci, Mr. Victor Gibeon, and Mr. Albert Nelth- ropp provided working space and equipment and assigned Public Works personnel to assist when needed. Mr. Nelthropp 1 °,lped de- sign the air-lift equipment used to pump the wells and took an active part in some of the tests. Mr. Alfred Ovesen collected samples of well cuttings and prepared logs of test wells. Mr. Leonard Larsen, also of the Public Works Department, worked with the author on pumping tests and inventoried some of the wells. Mr. R. L. Kenan, of R. L. Kenan and Associates, Montgomery, Ala., provided well logs and information on pumping tests made in the Concordia and Ma- hogany Road well fields when they were installed in 1949-50. Mr. Eugene Schuster, local driller who has put down most of the drilled wells on St. Croix, donated 4 days of his time to accompany the author in a field inventory of wells. His firsthand knowledge of the wells and his friendship with the well owners made it possible to obtain much information that otherwise might not have been avrilable. GEOGRAPHY AND CLIMATE The island of St. Croix is the southernmost of the Virgin Islands, about 40 miles south of St. Thomas, site of Charlotte Amalie, the capital of the American Virgin Islands. It is about 100 miles south- east of San Juan, P.R., and about 1,100 miles southeast of Miami, Fla. The island is about 21 miles long and 6 miles wide at its widest sec- tion. The long axis of the island trends about 10° north of east. There are two highland areas, the North Side Range and the East End Range. A broad, rolling lowland makes up the central, south- central, and southwestern parts of the island. There are no perennial, through-flowing streams on the island, but a few of the strums are perennial along short reaches. The major physical features and drainage courses are shown on figure 1. The population of St. Croix in 1962 was about 16,000. AHut 5,500 live in Christiansted and 2,500 in Frederiksted. The remainder live GROUND WATER FOR ST. CROIX, VIRGIN ISLANDS D3 in small villages or in the country. Several private housing develop- ments are now being built. These will provide many homes in areas outside the present city limits of Christiansted and Frederiksted. The 'average temperature in St. Croix is 78 °F and varies only a few degrees between winter and summer. The northeast trade winds, which blow most of the time, contribute to the pleasant climate. The average annual rainfall varies from an estimated 20 inches at the east end of the island to about 54 inches at Annaly, in the mountains at the northwestern part of the island. Rain generally falls in showers lasting from a few minutes to a few hours. Continuous rainfall of a day or more is rare and occurs only during the hurricane season. Torrential rains are not uncommon, as much as 7 inches falling in 24 hours in the north-central part of the island about once every 5 years (U.S. Weather Bureau, 1961, p. 86). The rainf all is markedly seasonal, generally being concentrated in the "wet season" in May 'and August-November. Variations from the normal seasonal distribution are common, however, as are large varia- tions in total rainfall from year to year. According to Cederstrom (1950, p. 10), the average based on stations at Christiansted, Fred- eriksted, and Kings Hill showed for the period 1852-1938 a minimum of 29.10 inches in 1922 and a maximum of 71.44 inches in 1933; the average was 46.34 inches. GEOLOGIC SETTING The geology of St. Croix was described and mapped by Cederstrom (1950). A more detailed map and description were prepared by J. T. Whetten in 1961 'as a doctoral dissertation for Princeton University. The following brief summary is based on these reports. The North Side and East End Eanges are composed of dense volcanic-sedimentary rocks of Late Cretaceous age designated the Mount Eagle Volcanics. A large intrusive mass of gabbro is exposed in the central part of the North Side Range and a similar mass of diorite is exposed in the East End Kange; the intrusives, also, are probably of Late Cretaceous age. Along part of the southeast flank of the North Side Range are outcrops of the Jealousy Formation, of Oligocene age, which unconf ormably overlies the Cretaceous 1 -^Irock. In outcrop the Jealousy is a calcareous conglomerate, but in the sub- surface to the southeast the Jealousy is composed chiefly of gray, green, or "blue" clay. About 1,400 feet of this clay was penetrated in a test hole drilled near Bethlehem (Cederstrom, 1950, p. 19), and the well did not reach the bottom of the formation. The Kingshill Marl, of Oligocene and Miocene age, consisting of limestone and marl, overlies the Jealousy Formation and crops out in a broad area in the central and southwestern parts of the island. D4 HYDROLOGY OF LATIN AMERICA AND THE ANTILLES The valleys of major streams are partly filled with alluvial deposits of sand and gravel, silt, and clay of Quaternary age; the deposits are everywhere less than 120 feet in thickness, so far as is known. SURFACE WATER No records of surface runoff on St. Croix are available. Appar- ently, the Danes collected some records on streamflow durirg the late 19th century, but no one has been able to trace the locatio?i of these records. In estimating the quantity of water that could be obtained in surface reservoirs, Tippets and others (1959) assumed that the average annual runoff is 1 percent of the annual rainfall, or less than half an inch. This low value of runoff presumably is based on the high potential rate of evapotranspiration. The average anrual evap- oration from an open pan at Annas Hope is about 72 inches, and that from larger bodies of water probably is something like two-thirds as high. The estimate of runoff as 1 percent of the rainfall may be too low. Rainfall records suggest that 3 inches or more can be expected to fall in 24 hours on the average of once each year on the northwestern third of the island (U.S. Weather Bureau, 1961, p. 85). About every 5 years, on the average, 4.5 inches or more can be expected to fall on that area in 12 hours (idem., p. 83). Torrential rains such as these may be expected to produce a high rate of runoff. Floodwaters are reported to flow over the Centerline Road to depths as great as 5 feet at least once every several years. A single flood of this magnitude may discharge more water than would normally flow in a period of a year or more. It appears that some of the streams in St. Croix may vary in dis- charge from nothing to perhaps thousands of cubic feet per second. To obtain significant records of surface runoff on St. Croix it will be necessary to measure flood flows as well as those confined to the stream channels. WATER-BEARING FORMATIONS The occurrence of ground water in the various geologic units was described in some detail by Cederstrom. The following discussion is based in part on his report but mainly on the results of an extensive test-drilling program earned out during the period 1959-62. Figure 1 shows the general availability of ground water in St. Croix as indi- cated by the results of drilling. The following descriptions of ground-water availability apply to the five areas shown in Figure 1. GROUND WATER FOR ST. CROIX, VIRGIN ISLANDS D5 695-967 63- D6 HYDROLOGY OF LATIN AMERICA AND THE ANTILLES Area I: Most wells in valley 'bottoms will yield 5 to 10 gpm of water of good quality that generally contains less than SCO ppm of chloride. Wells in alluvium-filled valleys draining relatively large areas (more than 1 square mile) may yield as much as 30 gpm. Water is obtained from alluvium or from weathered and fractured rock. Area II: Most wells will yield little water or salt water. Area III: Most shallow wells (less than 100 feet deep) will yield as much as 10 gpm of water of fair quality containing generally less than 700 ppm of chloride. Wells in some of the larger valleys may yield 100 gpm or more. Deeper wells and wells near the sea generally will yield saltier water, containing more than 1,000 ppm of chloride. Water is obtained from alluvium and from limestone. Water in the limestone generally is saltier than that in the alluvium. Area IV: Most wells will yield small supplies (1 to 5 gpm) of water containing less than 500 ppm of chloride. Wells near the sea generally yield saltier water. Water is obtained from alluvium and from weathered and fractured rock. Area Y: Most wells will yield salty water. Wells in the upper parts of the larger valleys may yield water of fair quality (less than 700 ppm of chloride). MOUNT EAGLE VOLCANICS The upper and weathered and fractured zone of the Monnt Eagle Volcanics yields moderate supplies, 5 to 10 gpm (gallons per minute), of water of generally good quality to a number of wells in the valleys of the North Side Range. Three- wells along Mahogany Road yield about 30 gpm each from these rocks. Because of the erratic distribu- tion of fractured zones it may be necessary to drill several test wells for every successful production well. In the east-central part of the island and in the East End Pange, the Mount Eagle Yolcanics generally yield only small amounts of water, and the water may be highly mineralized, the chloride content ranging from about 150 to 5,000 ppm or more. The chance of penetrating permeable zones in the Mount Eagle Yolcanics declines with depth. The weathered zone generally does not extend more than 100 feet below the surface, and open fractures pinch out with depth. Although the Mount Eagle includes limestone beds in places, no evidence of permeability due to solution of the limestone was noted. INTRUSIVE BOCKS Cederstrom (1950) reported a successful well in the weathered intrusive rocks in the central part of the North Side Range. This well was pumped for 6 hours at a rate of lO1/^ gpm with a drawdown of GROUND WATER FOR ST. CROIX, VIRGIN ISLANDS D7 15 feet. No additional information on the availability of ground water in this area was obtained by the present writer. It is recommended that test wells be drilled into these rocks in the area upstream from River and near Hermitage. Two drilled wells in the weathered intrusive rocks of the East End Range are reported to yield 10 gpm each. The water is of fair to good quality; the chloride content of water from the two wells is ^00 and 300 ppm. Test wells are recommended in the valley south of Southgate Farm. SEDIMENTARY ROCKS JEALOUSY FORMATION Little is known of the water-bearing capacity of the Jealousy For- mation. The gray or "blue" clay which underlies the Kingshlll Marl in the central and southwestern parts of the island yields little or no water. The calcareous conglomerate at the base of the formation appears to be moderately permeable where exposed in outcrop, but a test hole drilled at Jealousy under the supervision of Cederstrom (1950, p. 66, 68) yielded no water from the Jealousy Formation. J. T. Whetten mapped the rocks in a belt extending from just south of Judiths Fancy to Mon Bijou as possibly belonging to the Jealousy Formation. No records of wells in this outcrop area were obtained. There appears to be a general impression that no water or only salty water can be obtained from the Jealousy. Nevertheless, it may be desirable to try one or more test wells in valleys where the conglom- erate of the Jealousy lies below the alluvium. Although not an aquifer, the relatively impermeable "blue" clay of the Jealousy is of considerable hydrologic importance. It is the lower limit of fresh-water-bearing rocks wherever it occurs. Accord- ing to drillers' reports, fresh water has never been found in rocks below the "blue" clay. This same "blue" clay, however, performs an important role in restricting the upward migration of salt water. Where fresh-water aquifers are underlain directly by the clay, upward migration of salt water probably is virtually nil. Because of its important hydrologic effects, the top surface of the "blue" clay should be logged in every test hole where it is reached, and a structure map on the top of the clay should be prepared. This map would show the effective lower limit for fresh-water wolls and would also help to determine areas where salt-water encroachment is likely to occur. KINGSHILXi MARL Probably the most productive aquifer on the island is the Kingshill Marl. Four wells in this formation have been pumped at rates of D8 HYDROLOGY OF LATIN AMERICA AND THE ANTILLES 100 gpm or more. The permeability of the limestone probably is quite variable. In the Frederiksted area small caves in the lirrestone are apparent. A break in the main waterline at Frederikstei was un- noticed for several days because all the water drained away through solution channels in the limestone. Unfortunately, water in the Kings- hill generally is of poor to only fair quality, the chloride cortent rang- ing from about 400 to 1,000 ppm or more. In most wells drilled into the Kingshill within a mile of the south coast, water stands at elevations less than 10 feet above sea level. Higher water levels in a few wells probably are perched or semi- perched. North of the Centerline Road, water levels more than 100 feet above sea level also are probably perched or semiperched. These high water levels are an indication that the limestone below is prob- ably of relatively low permeability. Water in the Kingshill probably is recharged along tlie beds of streams discharging southward from the North Side Rang3 and also in the broad outcrops of limestone near the south coast. A favorable recharge area would seem to be along the southeast flank of the range where the streams first cross the Kingshill rocks. However, it is in this area that most wells in the Kingshill are unsuccessful, because little water, or only salty water, is obtained. Possibly the Jealousy Forma- tion, which crops out in places along the edge of this area, contributes salty water to the overlying Kingshill, or perhaps openings related to the fault mapped by Whetten allow upward migration of salt water. The low yield of wells in this area may be due to the shallow depth at which clay of the Jealousy Formation lies. Deposits of sand and gravel in the alluvium yield 40 gpri or more to several wells along the streambed between Adventure and Manning (fig. 5). Alluvial deposits also provide a part of the water pumped from the Concordia and Barren Spot well fields and may yield some water to the public-supply wells in the Mahogany Koad well field. Alluvial deposits along smaller streams yield small supplies to many stock and domestic wells. It is difficult to predict the yield of a well drilled in the alluvium because the character of the alluvial material varies greatly within short distances. Silt and clay mahe up the greater part of the alluvial section in most areas. Beds of sand and gravel generally are thin and are interbedded with silt and clay. Yields of some wells are greater than would be expected from examination of the drill cuttings. Drill samples that include clay, silt, sand, and gravel in varying amounts probably are obtained from thin sand and gravel beds alternating with silt and clay. GROUND WATER FOR ST. CROIX, VIRGIN -ISLAND'S D9 The quality of water obtained from the alluvium in the valleys of the North Side Range and most of the central plains area generally is fair to good, the chloride content ranging from 100 to TOO ppm. In the central plains area the water in the alluvium generally is lower in chloride than that in the underlying limestone. Water in the alluvium along the southeast flank of the North Side Range is reported to be generally salty, and water in alluvial deposits near the coast and at the east end of the island is almost invariably salty. PUBLIC-SUPPLY WELL FIELDS As of June 1962 four major well fields for public water supply were active or were being developed. In addition, there are several small public-supply systems for small communities in various partr of the island. These small public supplies generally consist of 1 to 3 wells each and a storage tank from which the local residents haul their water. Plans are now being made to provide pipelines to individual homes in some of these communities. Wells are identified in this report by field numbers that are shown on the maps (figs. 2-5) and by map coordinates that refer to the 10,000-foot grid on the 1:24,000 U.S. Geological Survey topographic quadrangle maps (1958). The grid is indicated by "tick" marlrs in the margins of the map. The coordinates are given in thousands of feet, the digit after the decimal point representing the location of the well to the nearest 100 feet. Thus, map coordinates 74.8N, 1079.6E indicate a well located 74,800 feet north and 1,079,600 feet east of the point of origin of the grid. The four major well fields are described in the following sect'ons. CONCOBDIA WELL FIELD The Concordia well field is in the Salt River valley near Concordia, about 4 miles west of Christiansted and about 1 mile southwest of Sugar Bay (fig. 2). Of the five wells now pumped in the field, four are equipped with turbine pumps, one with a jet pump. Apparently the water is obtained chiefly from limestone, although some water is obtained from sand and gravel in alluvium overlying the limestone. Records of wells in the Concordia field are listed in table 1. Pumping tests by R. L. Kenan and Associates on three of the wells indicate specific capacities ranging from 3 to 5 gpm per foot cf draw- down. Well 14 (coordinates 74.3N, 1078.7E) was pumped at a rate of 50 gpm for 30 minutes, with a drawdown of 12 feet. Of the five wells, four are now pumped at a rate of about 30 gpm each, and at times of high demand they are pumped 24 hours a day. Pitmping- test data also are summarized in table 1. D10 HTDBOLOGY OP LATIN AMERICA AND THE ANTILLES R I B B E A N Salt River Bay NORTH SIDE RANGE J Juc'ith Fancy Hill x ,rJ] Marys Fancy Rattan aiH Belvedere EXPLANATION o Well Base map from U.S. Geological Survey, 1958 Boundary of a^avium FIGURE 2. Concordia well field. 1 MILE J TABLE 1. Concordia well fleld Records of wells [Yield: M, measured; R, reported) Field No. 1... __ . _ .. 7... __ ....... 11 . 13............. 14......... _ . Map coordi- nates 74.8N 1079.6E 75.4N 1079.6E 74.2N 1078.6E 73.8N 1077.9E 74.3N 1078.7E Altitude (feet above sea level) 40 35 40 50 40 Depth of well (feet) 82 81 86 96 85 Water-bearing material and gravel, also. .....do.. ....... ............. Water ievel (feet Date below meas- land ured surface) 24 5-25-62 25 5-25-62 28 5-25-62 51 5-25-62 20 3-14-62 Yield (gpm) 22R 30R 30R 30R 50M Chloride content (ppm) 250 240 700 600 Remarks Not being pumped when measured. Equipped with jet pump, to be replaced by turbine pump in 1962. Being pumped when measured. Equipped with turbine pump. Do. Do. with turbine pump. Pumping; tests of wells [Preliminary test by R. L. Kenan and Associates] Field No. 1... ........... 11. 13. Map co- ordinates 75. 4N 1079. 6E 74. 2N 1078. 6E 73. 3N 1077. 9E Date tested 8-10-48 5-26-49 5-23-49 Pumping rate (gpm) 20 40 45 Hours pumped 2C 9 24 Draw- down (feet) 4 13 12 Specific capacity (gpm per foot of drawdown) 5 3 3% o 50 D12 HYDROLOGY OF LATIN AMERICA AND THE ANTILLES The water obtained from the Concordia field reportedly varies in quality with rainfall and with the amount of water pumped from the field. At times of heavy rainfall and minimum pumping trs quality of water pumped from the Concordia fields is fair to good and the chloride content is less than 400 ppm. At times of drought and of heavy pumping the water reportedly becomes rather brackish. The chloride content of water also varies from one well to another. On May 3, 1962, a sample was taken from each of three wells in the Con- cordia field. Water from wells 7 and 11 had 250 ppm of chloride; that from well 13 had 550 ppm. The city supply in Christiansted on this date had 350 ppm of chloride. On May 14 water from wells Y and 11 had 300 ppm of chloride. The Concordia well field probably is developed to near its maximum capacity at the present time (June 1962). Probably additional water could be pumped from the field during the rainy season, but this is generally the time of least demand for water. Heavy pumping during the dry season probably will increase the salt content of tl Q- water. The increase of salt in the water probably is caused by inducement of greater inflow from deeper limestone beds when the water level in the well is lowered excessively. Because the amount of potable water is limited, a system of monitor- ing the supply should be. started at once. It would involve chiefly the measurement of changes in water levels and changes in chloride content of the water. Records of rainfall and of pumpage from the field also should be collected. The depth to water in producing wells should be measured by air line, electric tape, or chalked tape at least once a month. A note ac- companying each measurement should indicate whether the well was being pumped when it was measured, and which other wells were being pumped at the time. If the well is idle when measured, the note should indicate how long it has been idle. Weekly readings would be desir- able for the first year of record. If air-line measurement is used, check readings with an electric tape or chalked tape should be made occa- sionally to make sure that the air line has not sprung a leak. One or more unused wells in the Salt River valley located near the producing wells also would be useful in maintaining a record of water-level fluc- tuations in the field. Tape readings could be made at monthly inter- vals, or a continuous record could be obtained by installing a, water- level recorder on one of the unused wells. The chloride content of water from each pmnped well should be determined at the same time a water-level measurement is made. Occasional samples should be analyzed in the laboratory to check the field determinations. GROUND WATER FOR ST. CRODC, VIRGIN ISLANDS D13 Records of rainfall, pumpage, water levels, and chloride content, when they have been collected long enough to cover a period including both wet and dry years and various rates of pumping, will furnish a basis for a more reliable estimate of the future yield of the well field than could be obtained in a short study no matter how inter *dve it might be. These remarks apply to the other well fields also. MAHOGANY ROAD WELL FIELD The Mahogany Road well field is about a mile north of Frederiksted (% 3). Each of the three wells is pumped with a turbine pump at a rate of 30 gpm. The water is obtained chiefly from weathered volcanie- sedimentary rocks underlying the alluvium. Records of wells in the Mahogany Road field and adjacent areas are listed in table 2. Pumping tests by R. L. Kenan and Associates on two of thes^ wells indicated specific capacities of about 2% and 4 gpm per foot of draw- down. A brief pumping test on the third well indicated a maximum yield of about 20 gpm when pumped by air lift. This well yielded 34 gpm when tested with a turbine pump. Pumping-test data also are summarized in table 2. D14 HYDBOLOGY OP LATEST AMERICA AND THE ANTILLES "B-ooksHill FREDERIKSTEI Sandy Point Cami Little LaGrange St. Geo-ge Hill Two WP'iams Hope and Carlton La"d« A R I Base hiap from U.S. Geological Survey. 1958 B E A N .SEA EXPLANATION Well Boundary^oif alluvium ¥2 ' 1 MILE FIGCBB 3. Mahogany Road well field. TABLE 2. Mahogany Road well field Records of wells [Yield: M, measured; B, reported] Field No. 1. 2...... ........ 3 ........... 21 32 "II"... "JJ"... .. "QQ". . Map coordi- nates 63.1N 1041.0E 63.2N- 1041. 5E 63.5N 1042.0E 59.5N 1042.1E 59.0N 1042.0E 60.2N 1042.8E 61.7N 1042.4E 64.0N 1044.0E Altitude (feet above sea level) 46 56 70 20 16 30 50 100 Depth of well (feet) 106 86 104 60 31 60 64 70 Water-bearing material do......... ..... ...... do . AUuvium(?)_ .... _ ... .... . do... .... ...... ... ..... .. rock. do....... ... ... .... Water level (feet below land surface) 63 53 48 14 38 46 Date meas- ured 1947 4-9-62 4-9-62 3-15-62 4-30-62 4-27-62 Yield (gpm) 30B 30R 30M 24M 6B 3M 12B Chloride content (ppm) 100 100 300 5,300 160 300 150 Remarks turbine pump. turbine pump. Pumping: tests of wells Field No. !». ..... 2 » ....- Map co- ordinates 63.1N... 1041.0E 63.2N 1041. 5E Date tested 4-25-49 4-26-49 5-23-49 6-24-49 Pumping rate (gpm) 30 38 Hours pumped 24 24 Draw- down (feet) 7 14 Specific capacity (gpm per foot of drawdown) 4 zy* i Betest by K. L. Kenan and Associates. » Final test by R. L. Kenan and Associates. CO d I 1 Oi Dl,6 HYDROLOGY OF LATIN AMERICA AND THE ANTILLES The water from the Mahogany Eoad well field is of excellent quality, containing only about 100 ppm of chloride. Wells near the mountains south of the well field (wells JJ, II, and 21) produce water of fair to good quality. However, well 32 was abandoned because it yielded only a trickle of very salty water. Salty water is1 obtained also from wells within the city limits of Frederiksted and along the coast north of the city. If the wells in the Mahogany Road field are pumped continously at a rate greater than the natural discharge to the sea, salt-water encroachment from the sea could occur. However, it appears that such encroachment is unlikely unless the present rate of pumping is greatly exceeded. If salt water is present in the rocks below the depths penetrated by the wells, its upward migration must be slow because of the low permeability of the rocks. The Mahogany Eoad well field is recharged by the underf ow in the valley of a westward-flowing stream which drains about 3 square miles in the wettest part of the island. Ideally, the field should 'be designed to intercept as much of the underflow as possible without lowering water levels to the point where salt-water encroachment might occur. Possibly some additional development could be attempted in this field. The three existing wells, which lie along the north side of the alluvium-filled valley, penetrate volcanic-sedimentary bedrock. If a well is drilled about 200 feet south of Mahogany Road well 2 (coordinates; 63.2N", 1041.5E) it is possible that a greater thickness of alluvium and weathered rock may be penetrated and a greater yield of water obtained. Additional supplies may be obtainable in the valley upstream from the present well field. Two or three wells might have to be drilled for each new production well, because the water-bearing crevices in the rock are erratic. At the time the well field was installed in 1949, only three of nine wells were successful; these were the discovery well and the 2 original production wells. BARREN SPOT WELL FIELD The Barren Spot well field is about 4 miles southwest of Christian- sted (fig. 4). There are seven wells in this field; as of May 1962 none had been equipped with pumps. Records of wells in the Barren Spot field are listed in table 3. Plans are underway to install well pumps and a pumping station in the Barren Spot field and to run a pipeline from the field to Christiansted. Each of the wells in the Barren Spot field penetrates two aquifers, alluvial material, including1 sand and gravel, and limestone or sandy limestone underlying the alluvium. In at least Some of the wells the first water obtained was under artesian pressure and rose in the hole GROUND :WATER FOR ST. CROIX, VIRGIN ISLANDS D17 as much as 20 feet. There was no indication of a change in water level when the underlying.liinestone was reached. Four of the wells '^ere test pumped in April and May 1962. Results of the tests are summarized in table 3. EXPLANATION Well Base map from U.S. Geological Survey/1958 Boundary of alluvium 1 MILE FIGURE 4. Barren Spot well field. TABLE 3. Barren Spot well field Records of wells [Yield: M, measured; R, reported] Field No. 3 __ -. ..... 4 ----- 5... _ .. 6. . __ - 7 ----- 8 - 9 ______ ... 28-...., ....... 29............. 31............. "L" ........ "0" _ . "R" .......... "YY" ...... Map coordi- nates 62.6N 1079.4E 63.0N 1080.4E 63.3N 1081.4E 63.5N 1082.1E 61.9N 1081.9E 61.6N 1081.3E 84.8N 1086.9E 62.5N 1076.0E 61.1N 1075.6E 61.7N 1081.6E 59.4N 1085.5E 60.1N 1075.6E 59.7N 1083.0E 64.2N 1075.3E Altitude (feet above sea level) on 68 75 82 55 52 108 150 125 53 80? 100 30 10 Depth of well (feet) 202 114 131 121 140 130 179 180 165 131 95 135 40 60 Water-bearing material ..do ..do .. . .... ... . .do _. Limestone and alluvium .... .....do......... .... ....... . .do......... ........... . _ .do . .. .....do....... .. .....do .................... Sand _.__ - ...-.-_ "Water level (feet below land surface) 80 60 68 75 52 J.Q 91 121 106 51 77 87 26 7 Date meas- ured 3-14-62 3-14-62 3-14-62 3-14-62 3-19-62 3_1Q_fiO 3-14-62 5-3-62 5-9-62 5-18-62 3-28-62 3-29-62 3-29-62 11-23-60 Yield fgpm) 4m? 100M 35R 38R 145M 100M 25R 10R 10R 64M 5R 15R 10R 10R Chloride content (ppm) 1,100 i nnn 1,000 150 400 850 900 600 300 1,500 Remarks 109 to 204 feet. to 115 feet. marl from 19 to 131 feet. marl from 95 to 121 feet. to 140 feet. to 130 feet. Sand and gravel from 119 to 121 feet; limestone from 177 to 180 feet. representative. 106 to 131 feet. Had been pumping 15 minutes before water level measured. d I-* 00 I§ Pumping tests Field No. 4 7- .. 8 31. . 31. Map coor- dinates 63.0N 10S0.4E 61. 9N 1081 .9F 61.6N 10S1.3E 61. 7N 1081 .6E 61.7N 1081 .6E Date tested 4-13-62 4-26-62 4-27-62 4-6-62 5-17-62 5-18-62 Pumping rate (gpm) 100 145 100 36 54 Hours pumped 01 , 24 3 3 Draw- down (feet) 21 15 5 4 Specific capacity (gpm per foot of drawdown) 7 7 7 13 Remarks Part of drawdown in well may be caused by entrance loss. Do. D20 HYDROLOGY OF LATIN AMERICA AND THE ANTILLES Most of the tests were too short to be subject to mathematical anal- ysis by conventional "aquifer test" methods. Furthermore, the as- sumption of a homegeneous extensive aquifer made in the calculations is not valid for this area. Nevertheless, the tests provided informa- tion that should be useful in planning the development of this field. The transmissibility of the limestone aquifer in the well field is es- timated to be about 50,000 gpd per ft. In an aquifer having such a transmissibility, 50,000 gallons will move each day through each section of the aquifer 1 mile wide under a hydraulic gradient of 1 foot per mile. The transmissibility of the limestone may be less in the upland area to the east and west of the field than in the well- field area, where the limestone underlies water-bearing alluvium. Wells 28, 29 and "O," in the uplands west, of the Barren Spot well field, were reported to yield 10 to 15 gpm when bailed. These bail- ing tests did not provide information adequate to justify an estimate of the transmissibility of the limestone in this upland area. The transmissibility of the alluvial material in the Barren Spot field is estimated to be about 10,000 gpd per ft. The short pumping tests were not adequate to determine the co- efficients of storage. When well 7 was pumped, the drawdown in well 8, which is 600 feet distant, though small, was almost immediate. It appears that the coefficient of storage in at least one of the aquifers, especially in the limestone, is small perhaps in the range of 0.001 to 0.0001. The coefficient of storage of the alluvium in places where the water is under artesian pressure may be in the same range. Where water in the alluvium is under water-table conditions the coefficient of storage is much greater, perhaps 0.1 to 0.3. Measurements of water levels indicate that the hydraulic gradient of the ground water is about 3 feet per mile toward the coast. Thus, if the average transmissibility is 50,000 gpd per ft., about 150,000 gpd is moving toward the coast in each mile width of the limestone. It is possible, however, that the permeable section of the limestone is not even as much as 1 mile wide. The significance of the small coefficient of storage is that the cone of depression expands rapidly. Even so, before there could be an ap- preciable lowering of the water level at the coast, saline water would tend to rise from below if wells were pumped heavily. The chloride content of water from well 31 (coordinates: 61.71SF, 1081.6E) when it was 108 feet deep was about 700 parts pe^ million at a pumping rate of 36 gpm. When the well was deepened to 131 feet, the chloride content increased to 850 ppm at a pumping rate of 54 gpm. The chloride content of water from wells 7 (coordinates: 61.9N, 1081.9E) and 8 (coordinates: 61.6N, 1081.3E) was 1,000 ppm GROUND WATER FOR ST. CROIX, VIRGIN ISLANDS D21 at a pumping rate of more than 100 gpm. It appears that water of better quality will be obtained from the Barren Spot field when the wells are pumped at a lower rate than when they are pumped at or near their maximum capacity. The chloride content of water from well "E," about half a mile south of the Barren Spot well field and about a mile north of Krause Lagoon, was only 300 ppm; but well "L," at about the same distance from the coast, yielded water con- taining 900 ppm of chloride. Wells 28, 29 and "O," in the uplands west of Barren Spot, yielded water containing 150, 400, and 600 ppm of chloride, respectively. The occurrence of chloride in the water from the limestone appears to be quite erratic and probably is re- lated, among other things, to depth of the well, distance from the sea, and the rate and duration of pumping. The Barren Spot well field is recharged by rainfall on the limestone and alluvial area to the north. No surface runoff occurs in the valley of Barren Spot even during the wet season in the average year. Con- sequently, it is assumed that water normally discharges from tris area entirely as ground-water flow. The amount of water that can be pumped from the Barren Spot well field depends on the amount of chloride that will be acceptable to the users. If each of the seven wells were pumped constantly at a rate of 30 gpm the field would yield about 300,000 gallons p°r day. At this rate of pumping it is estimated that the initial chloride con- tent of the water would be less than 1,000 ppm, possibly as low as 800 ppm. However, it is probable that the chloride content of the water would increase as the water stored in the alluvium was used up and a relatively larger proportion of the water was drawn from the limestone. If each of the wells is pumped at a rate of 100 gpm the field would yield about 1 million gallons of water per day. At this rate of pump- ing the initial chloride content of the water probably would be more than 1,000 ppm; at the end of a year of pumping the chloride content of the water might exceed 2,000 ppm. It is unlikely that drastic changes in water levels or quality will take place overnight. In all probability a trend of lowering water levels, increasing chloride, or both will be noticeable long before the problem is critical. If such a trend were detected, pumping- would have to be reduced. If the water levels, chloride content of water, pumpage from the field, and local precipitation are all plotted on a single graph, the relation of each factor to the others will be readily apparent. The ground-water supply at Barren Spot should be monitored be- fore and after pumping begins. Records obtained before water is pumped from the field will show the fluctuations in ground-water D22 HYDROLOGY OF LATIN AMERICA AND THE ANTILLES levels and in quality of water that occur under natural conditions. Records obtained after pumping begins will show changes in ground- water levels and in quality of water induced by withdrawing water from the field. As in the Concordia field and elsewhere, such records will be the best guides to future operations. ADVENTURE, GOLDEN GROVE, AND MANNING WELL FIELDS The Adventure Well field is just south of the Center-line Poad about 4 miles east of Frederiksted (about 5y2 miles by road) (fig. 5). There are four public-supply wells in this field. The water-bearing material, logged as yellow marl, sand, and gravel, is chiefly alluvium, although the lower marl may belong to the Kingshill Marl, Each of the wells is bottomed in "blue" clay, probably of the Jealousy For- mation. Downstream from the Adventure well field are the Golden Grove well field and the Manning well field. There are four indus- trial wells in the Golden Grove field and five in the Manning field. Records of wells in the Adventure, Golden Grove, and Manning well fields are listed in table 4. Pumping tests were made on each of the wells in the Adventure field. Results of the tests are summarized in table 4. GROUND WATEB FOB ST. CEOIX, VERGEST ISIiANDa D23 Kingshill \ EXPLAN/TION Well Base map from U.S. Geological Survey, Boundary of alluvium I MILE FIGDEH 5. Adventure, Golden Grove, and Manning well fields. u TAKLE 4. Adventure, Golden Grove, and Manning well Records of wells [Yield: M, measured; K, reported] Field No. 1 2 3 4 5 6- _ ...... .... 7 _ .... 8 9 ....... ... 15.......... 16.. .... .... 17 18 19 Map coordi- nates 56.9N 1072.7E 57.3N 1072.2E 57.4N 1072.0E 57.7N 1070.9E 57.7N 1070.3E 59.2N 1068.1B 59.1N 1067.7B 59.3N 1067.7E 57.2N 1067.3E 60.8N 1064.1E 60.9N 1064.7E 59.3N 1064.9E 60.7N 1064.5E 60.6N 1063.7E Altitude (feet above sea level) 15 15 20 35 M 50 50 50 50- - 80 90 108 80 80 Depth of well (feet) 106 107 164 103 127 Water-bearing material do. . .....do . _ __do _ _....do__ . _.. do. ...-_..... . .....do .'.do... .._.... Alluvium AlliivinTn do................ ._ Water level (feet below land surface) 29 8 15 39 14 10 Date meas- ured 4-29-62 3-15-62 3-15-62 3-15-62 3-15-62 3-15-62 Yield (gpm) 40K 30K 30R 30K 30K 40B 40R 40B 40R 150M 40M 10R 65M 65M Chloride content (ppm) 600 550 inn 550 500 300 inn 150 200 inn Remarks turbine pump. Do. Do. Do. Do. Do. Do.; O he] I Pumping tests of wells Field No. 15...... ...... 16...... ... 18.. ____ .. 19............ Map co- ordinates 60.8N 1064.1E 60.9N 1064.7E 60.7N 1064.5E 60.6N 1063.7E Date tested 4-16-S2 4-18-62 5-8-62 4-17-62 Pumping rate (gpm) 150 40 65 65 Hours pumped Ui o 5 1 Draw- down (feet) 11 43 Specific capacity (gpm per foot of draw- down) 1^ Remarks pacity. casing. O to D26 HYDROLOGY OF LATIN AMERICA AND THE ANTILLES In all but one of the tests there was evidence that the ^ater dis- charged from the well into the gut (streambed) was returning to the well. None of these tests was continued long enough to lower the water levels in nearby wells. The quality of water yielded by the Adventure wells is gooi. Chlo- ride ranged from 150 to 300 ppm. No change in chloride content was noted during the short pumping tests. The aquifer apparently is recharged by infiltration of rainfall and runoff in the alluvial area at the well field itself and to the north and northwest. Probably most of the recharge comes from water running in the gut after heavy rains. During unusually heavy rains in Oc- tober 1960, water reportedly was running over the Center]ine Road at the Adventure field, covering the road with more than 5 feet of water. During such floods it is probable that the alluviurr becomes saturated to the land surface. The short time needed for water in the gut to reach the aquifer was apparent when well 15 was test pumped on April 16 and again on April 17. The drawdown on April 16 was about 10 feet after 150 gpm had been pumped into the gut for 1 hour. On April 17 the drawdown was less than 2 feet after 2^ hours of pumping at the same rate. The potential yield of the Adventure field could not be determined from the results of the short-term pumping tests. Probably each of the wells could be pumped at a rate of not less than 30 gprr. There appears to be little danger of encroachment of salt water from below, as each of the wells is bottomed in impermeable material. Spread of the cone of depression to the limestone bordering the alluvial ma- terial may bring in water higher in chloride content. It is possible that the cone of depression formed by pumping the wells may induce recharge of fresh water during floods that otherwise would b^ lost to the sea. Additional wells could be drilled along the gut between the Adventure and Golden Grove fields and also perhaps along the east branch of the gut south of Centerline Road. Salt-water encroachment from the sea could result from heavy pumping in the Manning well field and possibly in the Golden Grove field. There appes.rs to be little danger of salt-water encroachment from the sea to the Adventure field because the static water level there is about 70 feet above sea level. The potential yield of the alluvial materials in the entire drain- age area can be determined only by additional test drilling and test pumping and by keeping records of rainfall, pumpage, water levels, and chloride content for at least several years. GROUND WATER FOR ST. CROIX, VIRGIN ISLANDS D27 REFERENCES CITED Cederstrom, D. J., 1950, Geology and ground-water resources of St. Croix, Virgin Islands: U.S. Geol. Survey Water-Supply Paper 1067,117 p. Hendrickson, G. E., 1962, Suggestions for the development of the ground-water resources of St. Croix, Virgin Islands: U.S. Geol. Survey open-file rep*., 83 p. McGuinness, C. L., 1953, Summary of the water resources of the Virgin Islands: U.S. Geol. Survey open-file rept, 8 p. Pirnie, Malcolm, 1945, Report on water supply and sanitary facilities for St Croix, Virgin Islands: New York, Malcolm Pirnie Associates, 27 p. Tippets, E. F., Abbet, R. W., McCarthy, G. T., and Stratton, J. H., 1959, Potable water supply for St. Croix, Virgin Islands: New York, Tippets, Abbet, McCarthy, and Stratton, 61 p. U.S. Weather Bureau, 1961, Generalized estimates of probable maximum pre- cipitation and rainfall-frequency data for Puerto Rico and Virgin Islands: Hydrol. Services Div., Coop. Studies Sec., U.S. Weather Bur. Tech, Paper 42, 94 p. o