Citation
Salt-water movement caused by control-dam operation in the Snake creek Canal, Miami, Florida ( FGS: Report of investigations 24, pt.4 )

Material Information

Title:
Salt-water movement caused by control-dam operation in the Snake creek Canal, Miami, Florida ( FGS: Report of investigations 24, pt.4 )
Series Title:
( FGS: Report of investigations 24, pt.4 )
Creator:
Kohout, Francis Anthony, 1924-
Leach, Stanley D. ( joint author )
Geological Survey (U.S.)
Place of Publication:
Tallahassee
Publisher:
[s.n.]
Publication Date:
Language:
English
Physical Description:
vi, 49 p. : illus., map, diagrs. ; 23 cm.

Subjects

Subjects / Keywords:
Water-supply -- Florida -- Miami Metropolitan Area ( lcsh )
Salinity -- Florida -- Miami Metropolitan Area ( lcsh )
Snake Creek Canal ( lcsh )
City of Miami ( local )
City of Tallahassee ( local )
City of Ocala ( local )
Canals ( jstor )
Saltwater ( jstor )
Aquifers ( jstor )
Dams ( jstor )
Fresh water ( jstor )
Genre:
non-fiction ( marcgt )

Notes

General Note:
"Prepared by the United States Geological Survey in cooperation with the Central and Southern Florida Flood Control District."
General Note:
"References": p. 49
Statement of Responsibility:
by F. A. Kohout and S. D. Leach.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
The author dedicated the work to the public domain by waiving all of his or her rights to the work worldwide under copyright law and all related or neighboring legal rights he or she had in the work, to the extent allowable by law.
Resource Identifier:
022573872 ( aleph )
01722744 ( oclc )
AES1338 ( notis )
a 64007369 ( lccn )

Downloads

This item has the following downloads:


Full Text
STATE OF FLORIDA
STATE BOARD OF CONSERVATION
DIVISION OF GEOLOGY
FLORIDA GEOLOGICAL SURVEY
Robert O. Vernon, Director
REPORT OF INVESTIGATIONS NO. 24 PART IV
SALT-WATER MOVEMENT CAUSED BY CONTROL-DAM OPERATION IN THE
SNAKE CREEK CANAL,
MIAMI, FLORIDA By
F. A. Kohout and S. D. Leach
U. S. Geological Survey
Prepared by the UNITED STATESGEOLOGICAL SURVEY
in cooperation with the
CENTRAL AND SOUTHERN FLORIDA FLOOD CONTROL DISTRICT Tallahassee
1964




FLORIDA STATE BOARD OF
CONSERVATION
FARRIS BRYANT Governor
TOM ADAMS RICHARD ERWIN
Secretary of State Attorney General
J. EDWIN LARSON RAY E. GREEN
Treasurer Comptroller
THOMAS D. BAILEY DOYLE CONNER
Superintendent of Public Instruction Commissioner of Agriculture
W. RANDOLPH HODGES Director
ii




LETTER OF TRANSMITTAL
Jorida geological Survey
Callakassee
October 10, 1963
Honorable Farris Bryant, Chairman Florida State Board of Conservation Tallahassee, Florida
Dear Governor Bryant:
The Florida Geological Survey has published four short papers as Parts 1, 2, 3, and 4 of Report of Investigations No. 24. This publication, Part 4, is the last of these brief papers and discusses the "Salt-Water Movement Caused by Control-Dam Operation in the Snake Creek Canal, Miami, Florida." The report was prepared by F. A. Kohout and S. D. Leach, hydrologists for the United States Geological Survey.
The principles of the relationships of fresh and salt water in the Biscayne aquifer and the movement of the salt-water, freshwater contact along the canals and in the aquifer, with use, with development of the control stations in the Central and Southern Florida Flood Control District, and in response to cyclic changes in the weather, are of great importance in understanding similar occurrences. These four reports present interesting details of these relationships.
Respectfully yours,
Robert O. Vernon
Director and State Geologist




Completed manuscript received
August 16, 1963
Published for the Florida Geological Survey
By E. O. Painter Printing Co.
DeLand, Florida
iv




CONTENTS
Abstract --- _ ----- 1 Introduction _2
History 2 Purpose of investigation ..------------- 5------------------------------ -- 5
Location and extent of area -.. .........-- -----.. 6
Climate .----...- _-----------._---- ------.-------------------- 7
Previous investigations -------------- -------- 7
Personnel and acknowledgments ------------- -------- 7
Geology and drainage 8 Methods of investigation -_-. ------------- 9
Long-term position of salt water --------------------------- 10
In the canal __-----_ ------------------------.------------ - 10
In the aquifer ---------------- 12
Short-term position of salt water 16
Salt-water movement in the canal ----- -------- 17
Natural tidal oscillation ------------17
Trapping salt water 20 Flushing of salt water from the canal 24 Ultimate loss and flushing of the remaining trapped salt water 26
Salt-water movement in the aquifer 29
Natural tidal oscillation ------------------ 29
Recharging the aquifer with trapped salt water 36 Ultimate loss and flushing of salt water in the aquifer 38 Manipulation of the control dam for maximum benefit --------------39
Factors affecting inland movement of salt water 40 Discharge related to gate opening 41 Summary 45 Conclusions as to operating criteria 47 References -------- 48
ILLUSTRATIONS
Figure Page
1 The Miami area, Florida, showing the approximate landward
extent of salt water at the base of the Biscayne aquifer in 1961 3
2 Photographs of the salinity-control dam (S-29) in the Snake
Creek Canal: (a) with right-hand gate opened 1 foot; (b)Y
with one gate fully open 4
3 Map of the northern part of the Miami area showing observation points along the Snake Creek Canal 6
4 Section along the south bank of the Snake Creek Canal
showing generalized lithologic characteristics of the Biscayne
aquifer 8
5 Graphs comparing rainfall, gate opening, and fluctuations
V




of water level and chloride content along the Snake Creek
Canal, 1960-61 .___ 11
6 Graphs comparing fluctuations of discharge, water level, and
chloride content in the Snake Creek Canal during the special
test of March 27-31, 1961 _- 12
7 Sections along the south bank of the Snake Creek Canal showing the zone of diffusion in the Biscayne aquifer, 1960-62 ... 14 & 15 8A Sections 1, 2, 3, and 4 through the Snake Creek Canal
showing movement of salt water during the rising tide, March
29, 1961 ---- ------ 18
8B Sections 5, 6, 7, and 8 through the Snake Creek Canal showing
movement of salt water during the falling tide, March
29, 1961 19
9 Sections 9, 10, and 11 through the Snake Creek Canal showing movement of salt water after closing of the control dam at
high tide, March 30, 1961 .__ .. - --... ..... .... ..... 21
10 Photograph of rock pit (at left) looking in the downstream
direction of the Snake Creek Canal (at right)_ .....- .-.--------- 22
11 Rock pit showing approximate bottom topography ..---- ------- 23
12 Sections 12, 13, 14, and 15 through the Snake Creek Canal
showing movement of salt water during and after flushing,
March 31 to April 1, 1961 25 13 Sections 16, 17, and 18 through the Snake Creek Canal
showing changes in isochlor position, April 3-11, 1961 .. 27 14 Sections 19, 20, and 21 through the Snake Creek Canal
showing changes in isochlor positions, April 18 to May 16, 1961 .-. 28 15 Sections 22, 23, and 24 through the Snake Creek Canal
showing changes in isochlor positions May 29 to July 10, 1961 30 16 Graph showing fluctuations of chloride content at well
stations 1, 2, and 3, 1960-61 ..---------- ........ ---------------31
17 Graph showing fluctuations of chloride content at well
station 4, 1960-61 .- -... .... ... ..-. ...... .... 32
18 Graph showing fluctuations of chloride content at well
station 5, 1959-61 33 19 Graph showing fluctuations of chloride content at well
station 6, 1960-61 ....... .... ....... .... 34
20 Graph showing fluctuations of chloride content at well
stations 7, 8, and 9, 1960-61 35 21 Schematic flow diagrams: (a) with gate partially submerged,
(b) with gate fully open --_42 22 Graph showing relations of head differential and discharge
for various openings of one gate of the control dam in the
Snake Creek Canal _43
vi




SALT-WATER MOVEMENT CAUSED BY
CONTROL-DAM OPERATION IN THE
SNAKE CREEK CANAL,
MIAMI, FLORIDA
By
F. A. Kohout and S. D. Leach
ABSTRACT
Movement of salt water in the Biscayne aquifer and the Snake Creek Canal was investigated to establish criteria for operation of the salinity-control dam in the canal. All four gates of the dam were opened for 3 days during a special test in March 1961. Although salt water oscillated landward and seaward a distance of 2 miles from the control in response to tide, the salinity of the ground water near the canal did not change greatly during this time. Salt water trapped in the canal by closing the dam at high tide moved landward as a density current at a rate of about 900 feet per hour and began to flow into a connected deep rock pit. A flushing operation removed salt water from the canal but the salt water in the rock pit remained. During subsequent months, wind and tide action caused upward dispersion of salt into the uppermost water of the rock pit, and the trapped salt water was gradually removed by seaward discharge through the canal.
Some of the salt water trapped in the canal moved into the aquifer. This salty water was traced by sampling fully cased wells of different depths and distances from the canal. Some of the salty water was retrieved by drainage from the aquifer when the gates were partially opened 2 months after entrapment. Because of its relatively great density some of the salt water moved downward in the aquifer and was not retrieved.
The data show the negative effects of trapping salt water, and also that the control can be closed gradually without trapping salt water. If the total opening of the control is distributed uniformly among the four partially closed gates, the increased velocity of the fresh water as it flows under the submerged sluice gates, will impede upstream movement of salt water at the bottom of the control. Calculations indicate that salt water probably will not move upstream in the canal if a head of 0.3 foot is maintained between upstream and downstream sides of the control at the time of minimum head differential, about 2 hours before high tide.
1




2 FLORIDA GEOLOGICAL SURVEY
INTRODUCTION
This report gives detailed attention to the movement of salt water in and adjacent to the Snake Creek Canal in the northern part of the Miami area, Florida (fig. 1). The movement results primarily from operations of a salinity-control dam located near the mouth of the canal.
From 1909 to 1930, canals were constructed westward from the coast through the Miami area into the Everglades to drain lowlying land for urban and agricultural use. The uncontrolled canals caused overdrainage. The water table was lowered nearly to sea level in some areas and this caused salt water to encroach into the Biscayne aquifer by two processes (Parker, 1951, p. 826):
(1) Where the water table was lowered greatly, salt water moved horizontally landward from the ocean through the highly permeable solution-riddled limestone of the Biscayne aquifer to replace the drained fresh water.
(2) Salt water moved directly inland in the canals where, because of its greater density, it tended to leak downward and laterally to contaminate the underlying fresh ground water.
Although the end result of these separate processes is the same, they attack the fresh-water resources from different directions and occur at different rates. The general aim of this report is to establish the conditions under which each process provides the greater danger of salt-water contamination.
HISTORY
In 1945, salinity-control dams were installed in most of the canals of the Miami area as barriers against further encroachment of salt water. These were emergency dams constructed by driving sheet-steel piles into the limestone. The dams were opened and closed by removing some or all of the piles during the rainy season and replacing them during the dry season. This was an inefficient procedure which frequently resulted in the dams- being left open much longer than necessary.
In recent years, the emergency dams have been replaced by permanent concrete structures with submerged sluice-gate openings. The control structure S-29 (Snake Creek Canal) is representative of some of the more recent structures. Each of the four gates, figure 2, is connected through a gear box to rollers which are driven by the rear-wheel drive of an anchored automobile. In the most recently contructed controls, in the Little River and




REPORT OF INVESTIGATIONS No. 24 3
, ic r A,
I ,: .
. : . .' : ,
CBICAYNL MA
Figure 1. 1Miami :area, Florida, showing the approximate landward extent of
salt water at the base of the Biscarne aauifer in 1961.
rANIA-I
t 7
SCALE INMILES
Figure 1. Miami area, Florida, showing the approximate landward extent of
salt- water at the baise ofthe!Biscayne aquife'r i91




4 FLORIDA GEOLOGICAL SURVEY
q 40
Figure 2. Photographs of the salinity-control dam (S-29) in the Snake Creek Canal: (a) with right-handgate opened 1 foot; (b) with one gate fully open.




REPORT OF INVESTIGATIONS NO. 24 5
Biscayne Canals, electric motors raise and lower the gates automatically by sensing the water levels upstream and downstream from the control dam. These automatically operated dams permit delicate control of water levels. However, the relation of salt-water movement to head differential needs further study before fresh-water resources can be effectively entrusted to automatic head-activated devices. As will be shown later, the relations of head and salt-water movement are very complicated and transient, and much more study and experience should be accumulated before true automatic control is attempted.
PURPOSE OF INVESTIGATION
Frequently, after heavy rainfall, headwater areas of the Snake Creek Canal remain flooded at the same time that fresh-water discharge at the control dam is not sufficient, with fully open gates, to keep the canal flushed of salt water. Under these circumstances salt water moves upstream in the lower reaches of the canal and contaminates the fresh ground water.
Questions had arisen as to the magnitude of this encroachment, the extent to which the encroachment might be tolerated during gate openings, and the feasibility of flushing the salt water out later when the gates were closed after flooding subsided. Accordingly, goals of the investigation were to establish the timerate of movement of the salt water for various circumstances of dam operation and, if possible, to develop principles that might have practical application not only in the operation of the salinitycontrol dam in the Snake Creek Canal but also in other canals of the area.
The investigation was made in cooperation with the Central and Southern Florida Flood Control District (C&SFFCD). Most of the data were collected between October 1960 and January 1962. The investigation was under the general supervision of M. I. Rorabaugh and, subsequently, C. S. Conover, Tallahassee, A. O. Patterson and K. A. Mac Kichan, Ocala, district engineers respectively of the Ground Water, Surface Water and Quality of Water Branches, U.S. Geological Survey. The investigation was under the immediate supervision of Howard Klein, geologist-incharge, Ground Water Branch, and J. H. Hartwell, engineer-incharge, Surface Water Branch, Miami.




6 FLORIDA GEOLOGICAL SURVEY
LOCATION AND EXTENT OF AREA
The Snake Creek Canal extends about 18 miles inland to the vicinity of Levee 33, on the east side of Conservation Area 3 (Inset map, fig. 3). Under the plans of the C&SFFCD (the operating agency) and the U. S. Corps of Engineers (the constructing agency), Conservation Area 3 will impound water pumped westward from urbanized areas during the rainy season. During the dry season, water stored in the conservation area will be delivered coastward to maintain water levels, thus preventing salt-water encroachment and providing replenishment water for municipal and private well systems.
This report covers the seaward reach of the Snake Creek Canal (Canal C-9) extending about 15,000 feet westward from the salinity-control dam (structure S-29). The width of the area extends 200 feet on either side of the centerline of the canal, but
ROCKJ
. OL DAMI
pll,, REL STATION NZI
Figure 3. Northern part of the Miami area showing observation points along Snake Creek Canal.




REPORT OF INVESTIGATIONS NO. 24 7
as one side should be the mirror image of the other, work was confined to the south bank of the canal. The canal widens into a lake whose dimensions are about 3,500 by 1,700 feet at the western end of the area of investigation, between State Highway 9 and Miami Gardens Drive (fig. 3). The lake, formed by removal of rock for concrete aggregate and other building materials, has a maximum depth of 40 feet below msl (mean sea level) and was included in the investigation because of its important influence on salt-water contamination.
CLIMATE
The climate of the Miami area is subtropical. The average temperature for January is about 68oF and for July about 820F. Rainfall averages about 60 inches per year, three-quarters of which falls in the May to November rainy season.
In relation to this investigation, 1960 was unusually wet with rainfall of about 75 inches, 25 inches falling in September when hurricane Donna and tropical storm Florence passed nearby. The year of 1961 was unusually dry; the rainfall amounted to about 42 inches.
PREVIOUS INVESTIGATIONS
Some phases of 'the present study coincided with, and were contemporaneous with, a study of the general hydrology of the Snake Creek Canal area by Leach and Sherwood (1963). Where data and graphs were applicable to the present report, they were used freely without acknowledgment.
Reports by Parker (1951), Parker, Ferguson, and Love (1955), and Kohout (1961) give much background information on the influence of drainage canals on salt-water encroachment in the Miami area. Many other reports cover the hydrology of the area and may be found in the "List of Publications" of the Florida Geological Survey in Tallahassee.
PERSONNEL AND ACKNOWLEDGMENTS
Thanks are extended to the Dade County Office of Water Control, especially to John Wilhelm, for furnishing rainfall and gate-opening data. Robert L. Taylor of the C&SFFCD aided by making available the facilities and personnel of his organization during the special test period of March 1961.




8 FLORIDA GEOLOGICAL SURVEY
Electrical-conductivity measurements of water in the canal during the test period of March 1961 were made under the direction of R. N. Cherry of the U. S. Geological Survey, Ocala. Other Geolog ical Survey personnel who contributed materially were B. F. Joyner of Ocala and C. B. Sherwood of Miami.
GEOLOGY AND DRAINAGE
The Biscayne aquifer of the Miami area is a highly permeable water-table aquifer consisting of solution-riddled limestone and calcareous sandstone with fairly frequent layers of unconsolidated sand. The generalized lithologic characteristics in wells drilled during the investigation are shown in figure 4.
Except for the peat and muck deposits at the surface, the sediments were deposited during the Pleistocene glacial epoch. Accumulation of snow and ice during glacial advances in the northern latitudes removed water from the oceans and lowered sea leveL Alternately, melting of the glaciers returned water to the oceans and caused sea level to rise. The alternation of limestone and sandstone beds in figure 4 relates to these fluctuations of sea level-estimated by Cooke (1939, p. 34) to have ranged from about 270 feet above to 300 feet below present sea level.
DISTANCE FROM CONTROL DAMIN FEET
3G,; 77 7 OCT 000C 900 8000 7000
0 LAND SURFACE- .0 L' -". *
/ --BOTTOM OF CANAL
DISTANCE FROM CONTROL DAM, IN FEET
5273 57co 4CC0 3000 2000 000 0
LIOESTON[.
LAND SURFACE vn r
_ C 1 ..--_ 1, .. ..... q SAND. .
-420
-IESOI -.. I -0
Figure 4. Section along the south bank of the Snake Creek Canal showing
generalized lithologic characteristics of the Biscayne aquifer.




REPORT OF INVESTIGATIONS No. 24 9
The peat and muck deposits were formed by the accumulation of partially decayed vegetation in swamps after the last glacial recession, about 8,000 years ago.
The Snake Creek Canal traverses a natural overflow channel from the Everglades. These numerous overflow channels are wide, topographically low areas (4 to 8 feet above sea level) that cut through relatively high land (8 to 15 feet) near the coast; they are called transverse or finger glades because of the manner in which they cut through the coastal ridge.
Prior to construction of drainage canals (fig. 1), the water level in the Everglades was high enough to cause discharge through the finger glades by poorly channelized sheet flow. In the deeper pockets, aquatic plants grew lushly and, after dying, settled to the bottoms of sloughs. In these locations, the organic material was protected from oxidation and accumulated to form the peat and black muck deposits.
In the higher parts of the finger glades, where water levels were below ground surface during the dry season, rapid oxidation prevented heavy accumulation of organic material. .In these areas, the remnant plant materials were mixed with particles of limestone and sand, eroded from the adjacent higher limestone hills, to form a deposit called marl (fig. 4).
Since construction of drainage canals, the finger glades are flooded only during extremely heavy rainfall, and the rich organic soils are cultivated- to provide table vegetables for northern markets in the winter season. With continuing improvements in the canal system and improved efficiency of water control, the agricultural activities are gradually giving way to urbanization.
METHODS OF INVESTIGATION
The investigation followed two main courses: (1) Long-term observation of the position and rate of movement of salt water under the past method of operating the control dam (S-29). The dam has been operated on an arbitrary water-level basis with no consideration being given to the movement or the position of salt water; (2) detailed inventory of the position and movement of salt water under special short-term test conditions. The short-term test served as an index of typical conditions for analysis of the long-term data. These phases required continuous inventory of changes, both in the canal and in the aquifer.




10 FLORIDA GEOLOGICAL SURVEY
The long-term study of salt-water occurrence in the canal was made primarily from fluctuations of chloride content as determined by three electrical-conductivity recorders located at S-29, at West Dixie Highway, and at the footbridge (fig. 3). The electrodes of these recorders were positioned about 1.5 feet above the bottom of the canal, and the landward movement of salt water was sensed by a sharp rise in the electrical conductance of the water. During and after the short-term test of March 27-31, 1961, the position of salt water in the canal was determined by traversing the canalwater vertically with the cell of a portable conductivity instrument. The measurements were made from a boat that moved to specific stations along the canal.
The long-term study of salt-water occurrence in the aquifer was made from fluctuations of chloride content in samples of ground water from nine stations of wells drilled along the south bank of the canal (fig. 3). The general scheme was to drill individual wells to different depths at distances of about 5 and 60 feet from the canal. Thus, a typical station might consist of wells 10, 20, 50, and 100 feet deep near the canal and wells 10 and 20 feet deep at a distance of 60 feet from the canal. Although drilling operations did not always fulfill the preferred condition of paired wells terminating at the same altitude, the rectangular coordinate positioning of the wells relative to the canal was virtually fulfilled. Each well was cased so that the water which it yielded was representative of that particular point in the aquifer. Details of the position and depth of each well terminus relative to the canal are shown in the map and section views of figures 16 to 20.
An attempt has been made to simplify the illustrations as much as possible, but as this three-dimensional problem can be shown only in two dimensions, much must be left to the imagination of the reader to keep himself oriented in three-dimensional space. Herein lies the understanding of the physical parameters being discussed.
LONG-TERM POSITION OF SALT WATER IN THE CANAL
The long-term position of salt water in the canal can be inferred by comparing figures 5 and 6. Figure 5 shows the peak daily chloride content, as indicated by the conductivity recorders at S-29, at West Dixie Highway, and at the footbridge. Also shown are the water level in the canal at West Dixie, the gate opening at S-29, and rainfall at the Opalocka gage. The maximum vertical opening




REPORT OF INVESTIGATIONS NO. 24 11
lv 1960 1961
Seac SEP OCT NOV Ea t N. i 1 APR MA NE Ae i N eC
20 0000 IA A t P 00 12 ,1 02 0 A fl .A ET O. DC 10;0M000
I 19poo z
Sw 000
; I Soo
0100 - 900O0
goo -j
....1E ,T .
Fr GGDo e jb
represents the discharge area.
All gates were fully open from early September to late October 1960, after the heavy rains accompanying hurricane Donna and tropical storm Florence produced widespread flooding (fig. 5). Only during the period September 10 to 14, 1960, was the discharge from the canal great enough to keep the canal completely flushed of salt water (chloride-content graph at S-29, fig. 5). At all other times full opening of the dam permitted salt water to move upstream on a rising tide. During the special short-term test of March 27-31, 1961 (fig. 5), salt water invaded the canal inland as far as the footbridge. The time scale for this period is expanded in figure 6 to show the detailed relations of intrusion, water-level fluctuation, and discharge. Salt water intrudes on the rising tide, first appearing at S-29, then at West Dixie Highway 2 to 3:hours




12 FLORIDA GEOLOGICAL SURVEY
AAM ryZ MARCHIII MARCM 9 9 MARtC"30 MARCH 31 APRIL I
: I I I II I
----1 1 1 ,'. ,
I F II Au
I I I I ,, , .....
4 CA rtS FEra -*Ats CLos so. AFt CL E. -IV I r
II I I- % I 7
T- / X iMi AX l I I I CAA 2I.x- '
Figure 6. Graphs comparing fluctuations of discharge, water level, and chloride content in Snake Creek Canal during the special test of March 27-31, 1961.
later, and finally at the footbridge during the subsequent falling tide, about 8 hours after first appearing at S-29. Complicated relations are involved in this time lag and they will be discussed later, but it should be noted that the length of time of salt-water stay decreases landward and that the maximum concentration also decreases landward. For example, salt water did not reach the footbridge at all on March 29, just before noon (fig. 6). In light of the movement of salt water shown in figure 6, the long-term position of salt water in the canal can be inferred from figure 5 by comparing the relative fluctuations of conductivity at the three recording stations.
IN THE AQUIFER
The interface between fresh water and salt water in the Biscayne aquifer is not sharp; a zone of diffusion is invariably present. The chloride content in the zone of diffusion ranges from about 16 ppm (parts per million), the concentration of fresh water in Miami, to about 19,000 ppm, the concentration of sea water. The concentration of chloride ion is commonly used to represent the




REPORT OF INVESTIGATIONS NO. 24 13
distribution of salt water, and isochlor lines are drawn through points of equal chloride content as determined by analyses of water from wells.
Cross sections of the zone of diffusion along the south bank of the Snake Creek Canal are shown in figure 7. The black dots represent the termini of individual fully cased wells located about 5 feet from the canal; the open circles represent the termini of wells located about 60 feet from the canal. In intercanal areas the the isochlor lines turn sharply downward at the inland extremity of the salt water, and the zone of diffusion takes the shape of a blunt-nosed wedge of salty water (Kohout, 1961, fig. 3, 4). Along the Snake Creek Canal (and presumably other canals) the blunted shape, although still present, tends to be modified by large changes in gradient and head near the canal.
When the dam is open, the lowest level of the water table coincides with the water level in the canal, and ground water flows laterally toward the canal from either side. Also, at such times, ground water beneath the canal tends to flow upward through the canal bottom as it follows the part of least resistance to the ocean.
When the control gates are closed, the water level in the canal becomes the highest level of the water table because water flows quickly downstream from inland points of higher head. Under this condition the exchange of water between the canal and the aquifer is reversed, and water, whether it be fresh or salty, moves laterally and vertically from the canal to the aquifer.
In figure 7, the general processes of salt-water encroachment in the Biscayne aquifer are recognizable:
(1) The deep salt water, below and seaward of the 1,000-ppm
isochlor, moves landward from the ocean in an essentially horizontal direction. This type of encroachment is relatively slow and insidious. It results from heads that are too low to prevent intrusion, and can be stabilized or
corrected only by raising the heads.
(2) The lenses of salt water that appear near the bottom of
the canal and then drift progressively downward with time originate as a result of inland movement of salt water directly in the canal. This type of encroachment occurs very rapidly and is the consequence of inadequate knowledge of the position of salt water in the canal during
control-dam operations.




i DISTANCE FROM CONTROL DAM,IN FEET [4p19t inn anonsa nnPn LAND *llfAC
4r' - D .
-------------------------M-j- u -- a
EXPLANATION O, A -- *4 o .n u 5l -lil -- ... "-i ;., ,, .-.0 . ..40
-0 ISO HLOR LINE IN PPM DECEMBER 8, 1960/
WELLS 5 FEET FROM CANAL
*.Ito a WELLS 60 FEET FROM CANAL
A ND URFAF- - Woi --'All T--- TA -,
U)S OTTO OF ANAl.. JANUARY 3,1961-i .n i / / ,.
2i
fLAURFAC-"
--------------- r- -P--------W7. -
-4040 -0 MARCH 24, 1961 Soo'g
- - - -", ,
LANQ 8WBFAK- f1.
1S T BOTOM 10T 6A L-l40
..a MARCH 31,1961 n 4 ,,. 120




LAND SURAC---. BOTTOM OF CANAL li" 40 .40
-80 APRIL 3, 1061 -s1M:/o
LAo suRFACF- Pg~fI4i
,j
z ...-- T '..MOF CAN40 1 -40 -1o MAY 6, 1961 ,
120 --.,-120,4'l~ LAND SURFACF-&
*' BOTTOM OF CANAL "
~ -40 Z -80 AUGUST 22, 1961
00
D~
.). .....-.-/ -,W
I
I~~.O .W .. --* e I I I i l I.
-12o a ;o -12o
Figure 7. Sections along the south bank of the Snake Creek Canal showing
the zone of diffusion in the Biscayne aquifer, 1960-62.




16 FLORIDA GEOLOGICAL SURVEY
Although the isochlor pattern in figure 7 appears to remain practically stable from December 1960 to January 1962, careful comparison of data shows that significant changes occurred at certain wells. The chloride content in the deepest well at station 4 decreased from 11,850 to 535 ppm, indicating seaward expulsion of salt water from the aquifer. During the same time, the chloride content in the deepest well at station 6 decreased from 1,220 to 490 ppm.
The reason for the expulsion of salt water may be understood by following the events shown in figure 5. From early September to late October 1960 the control dam was fully open, because of flood conditions. During this time the heads in the aquifer in the vicinity of the canal were below normal and salt water moved horizontally landward and mounded beneath the canal. The chloride content in well G 420A (the only long-term data covering the period, fig. 18) increased from 6,800 ppm in June 1960 to 9,200 ppm on October 19, 1960. The year 1960, then, was essentially a year of encroachment.
The year 1961 was abnormally dry, and even during the normally rainy months of September and October the dam remained practically closed. The infrequent opening of the control dam caused heads to remain sufficiently high so that expulsion of salt water from the aquifer occurred throughout 1961 (fig. 7). The fluctuation graphs in figures 16 to 20 confirm the seaward movement of salt water; chloride contents in most of the wells decreased generally during 1961.
It is indeed a contradictory situation that the greatest inland movement of salt water took place in a wet year and the greatest expulsion of salt water took place during a dry year. Nevertheless, this fact indicates that the means are available for very effective control of the movement of salt water in the Biscayne aquifer.
SHORT-TERM POSITION OF SALT WATER
While it is generally not feasible to observe continuously the complete distribution of salt water, short-term tests, keyed to recording instruments, provide much information with which to interpret the long-term data.
A short-term test was performed March 27-31, 1961. The control gates were completely closed for 6 weeks prior to the test, and conditions were almost stable despite rainfall of 2 inches that occurred 3 days before the test (fig. 5).




REPORT OF INVESTIGATIONS NO. 24 17
The sequence of control-dam operations is shown in the water level graph of figure 6. All four gates were opened fully at 10:00 a.m., March 27, just after high tide. The canal was in natural tidal oscillation until 9:15 a.m., March 30, when S-29 was completely closed just before high tide. The intention was to trap salt water in the canal, as might happen during normal control operations, and then trace its movement both in the canal and in the aquifer. The trapped salt water moved inland in the canal much more rapidly than expected, and at 1:30 p.m. on March 31, at high tide, two gates were opened for a flushing operation which continued until low tide, at about 5:45 p.m. The dam was then closed and remained so until May 26, 1961.
SALT-WATER MOVEMENT IN THE CANAL
The tidal movement of salt water in a canal is rapid and transient. In figures 8-9 and 12-15 this movement is depicted by a time-lapse sequence of canal sections showing the approximate position of isochlor lines along the longitudinal axis of the canal.
NATURAL TIDAL OSCILLATION
The movements of salt water under unimpeded tidal oscillation on March 29 are shown in figures 8A and 8B; figure 8A shows the inland extension on the rising tide, and figure 8B shows the seaward movement on the falling tide.
The stage of the tide cycle represented by each isochlor section is shown by an arrow on the water-level fluctuation graph (inset). The measured velocity profile for the middle vertical in the canal at West Dixie Highway is shown at the lower right side of each section. The discharge is divided into landward and seaward components as shown by the figures near the darkened arrows.
Comparison of sections 1, 2, and 3 (fig. 8A) indicates that the inland advance of salt water was slow during the early part of the rising tide. Significant landward advance of salt water did not take place until all flow became negative (landward) in section 3. This indicates that seaward-flowing water in the upper part of the canal removes the landward-flowing salt water in the bottom of the canal by a combined dispersion and interface erosion process.
Maximum inland velocity occurred at 7:37 a.m., about 11/ hours before high tide, section 4; the unshaded discharge arrows and dashed velocity curve show this condition. The solid velocity curve at 8:44 a.m., which corresponds in time with the isochlor




4toof rrff~ P ONS ASt/TUlpf,/NafrT~frmD TUISt 44jrTpj/MNFffDVAIffRR(O TO MIS AS TiTuDpINFIETfRfINN,0 F04#54
t n- 00
0
ni r o i
A LI E IN ET AYLTIDE INMfT0 9"11 -STAT/ONEs W" ""YONS'-CONTROL DAM




REPORT OF INVESTIGATIONS No. 24 19
DISTANCE FROM CNRLDMIN FEET
is 1W00 1000 8000 OO 1 Ol 60,00 4000 2000 0 :k (5) MARCH 29.1961 (9:50A.M.) JWATFER SURFAC-.
.tN- -Q, .. PARTS PERI MILIO
. . . . . .
a -. . .
SISE t EEK
....... I" LL _._ ', . .. . .- -. r,,. EAR, M. ; i I
I%-I -lLllIs
(6) MARCH 29.|961(1240 P.A.) WA -R SUR.A"%-,
-.
-10 3
MA27 MR.8 MAR29 1M02 MAR. 31 APR I SE ISF CSWA0L
i -GA G ". SP : SP 1 GA I I92' .R pa Ia-2- ... .... s 4 DX..lv
-1 IVA .. IU C,--4
1HHOR. PMTSEPE MILLION
1 (6) MARCH 2R.1961 (12: PP.)
1 WATER SUlRFAC:E9- %A9LwWWeL RIFkt1 a--pOE 'UJ2
IIL
. :' -, : .,i '- l I
1. .
(8U) MARCH o a1961i(t20P.M.)
2 WTE SURFACE-% S ISOCHLOSIR PARTS PER MILLION.SO 9-1
MAR.27 MAR.28 MAR.29 MAR30 MAR.31 APRA IN Fy ar CVYIL
-2 .p' I' I''
aI. L 1 j ; .I- -. ;'
1 ~ l f~ l - T R JI
1I I-V I-AC&
moemn (8) MARCH watr urng9hefaligtde0Mrch2M161




20 FLORIDA GEOLOGICAL SURVEY
section, indicates that relatively fresh water near the surface began to flow seaward at least 1 hour before high tide. The highsalinity water at the inland toe of the wedge, however, continued to move landward until well after high tide. This time lag in landward movement of salt water was shown previously in the chloride-fluctuation graph of figure 6, where salt water did not appear at the footbridge until about 2 hours after high tide.
These out-of-phase flow characteristics in the upper and lower parts of the canal result from the density contrast between fresh water and salt water. At, and for some time after, high tide the fresh-water head at inland points is too small to counterbalance landward movement of salt water and the inland toe of the wedge moves landward as a density current. As the tide begins to fall more rapidly after high tide, the landward hydraulic gradient associated with the density current is overcome by seaward hydraulic gradient from falling ocean level. The salt water at the bottom of the canal, then, reverses direction and flows seaward.
The salinity distribution in the canal reflects the out-of-phase flow pattern. The upper part of the canal is rapidly freshened by large positive (seaward) velocities in the fresh water as the tide falls (fig. 8B), section 6. However, even after the reversal of landward movement of salt water, seaward velocities in the salt water near the bottom of the canal are small compared to those of the overlying fresh water (velocity curves, sections 7, 8). This velocity differential causes a lens or slug of salt water to lag behind in that part of the canal westward from West Dixie Highway. Apparently, downward-directed velocity components (possibly related to turbulence near bridge pilings) develop in the fresh water near the outlet. Salt water in this part of the canal is eroded and flushed seaward before all the salt water that moved landward during the rising tide can retrace its path to the outlet. The phenomena occurred repeatedly when the control dam was fully open. This indicates that the control should not be operated exclusively on the basis of water-level measurements, for slugs of salt water could be inadvertently trapped. The position of salt water should be determined directly by conductivity measurements at the bottom of the canal.
TRAPPING SALT WATER
This section will show the dangers of inadvertently trapping salt water in a canal by shutting the gate on it. On March 30 at 9:15 a.m., just before high tide, all four gates of S-29 were closed.




REPORT OF INVESTIGATIONS NO. 24 21
The salinity distribution just prior to the time of closing is shown in figure 9, section 9.
A stable density stratification exists when fresh water overlies salt water with the isochlor lines horizontal. Sloping isochlor lines in stagnant water may be generally interpreted as an indication of hydraulic instability related to a nonstable density stratification. After the gates were closed, the water (for purposes of this discussion) may be considered to have been virtually stagnant, though
. RI. IIL i i | IL R
DISTANCE FROM CONTROL DAM,IN FEET 14,000 12000 m 10,000 8000 6000 4000 2000 0 % (9) MARCH 30,1961 (8:50A.M.) Q WArER SURFAF- S ISOCHLORS,IN PARTS PER MILLION 'I0
9:
.. MAR.27 MAR.28 MAR.29 MAR.30 MAR.31 APR.I CONTR
6 S P 11 GA G P I. 1% J t2,S P 12 GA 12 GP 12 GA f2 SP OE
-2
-4 .... . .
(10) MARCH 30. 1961 (11:30A.M.)
VATER SURFACE
ISOGHLORS.IN PARTS PER MILLION 0
I l ~ ~ ~ ....... ......... l,'& 9' MR 0 MA. ...1 1 ............ /TR
9: . ... . . ..... . .... .
SMAR27 MA MA 9 MAR.30 MA A R CONTROL F e o 10, a 11 Pro th S e eC CLOSED
MARCMarch130. 1961.
ISOHL R, N ARS 03 M I O A
Ik LI' APARCR.KA L At AT(A IR EPIZ1E 2 .0y,*
-" I
1;. : i -- iI 4 I NKE CR CKAA 4
-41
V 01) MARCH 30,1961(2:40 P.M.)
' ISOCHLORS*IW PARTS PER MILLION
......
.. . . ~ .. .
Mac 30 1961




22 FLORIDA GEOLOGICAL SURVEY
some net inflow of fresh water entered the canal as the stage rose; the sloping isochlor lines in section 9, then, indicate instability.
As expected, a density current developed when the gates were closed and salt water in the seaward reach of the canal moved landward. To maintain continuity between the individual water masses, fresh water from inland points moved seaward through the upper part of the canal to replace the salt water which moved landward through the lower part.
These movements were traced primarily by making vertical conductivity traverses of the canal water from the boat. Also, vertical conductivity traverses were made continuously at the footbridge and a series of these profiles is shown at the lower right side of section 10. The leading nose of the salt-water wedge passed the footbridge at 10:44 a.m., it was about 5 feet thick and had a maximum chloride content of about 1,500 ppm.
By tracking the intruding nose of the salt-water wedge for a period of time it was determined that the rate of landward movement was about 15 feet per minute, or about 900 feet per hour. The serene water surface completely belied the rapid encroachment taking place at the bottom of the canal.
Figure 10. Photograph of rock pit (at left) looking in the downstream
direction of the Snake Creek Canal (at right).
r, 7
'V4
Figure 10. Photograph -of -rock pit (at left) looking in the dow nstream
direction of the Snake Creek Canal (at right).




REPORT OF INVESTIGATIONS No. 24 23
The salt water moved nearly 1 mile in the 41/4 hours between 9:15 a.m., the time of closing the gates, and 1:30 p.m. Careful inspection of the discharge graph at Miami Gardens Drive (fig. 6) shows a gentle increase in discharge at about :li30 p.m. This increase was sensed by a small velocity vane positioned in the fresh water above the intruding salt-water wedge, and reflects the change in velocity in the seaward flow section as the wedge passed Miami Gardens Drive.
At about 2:00 p.m. salt water began to flow into the rock pit, in a manner similar to water passing over a waterfall (fig. 3, 9, 10, 11).
By late afternoon the hydrologic significance of the rock pit became clear. The rock pit was a hydraulic sink in which the salt water would temporarily collect. The salt water would resume its upstream movement only after it filled the rock pit above the level of the bottom of the canal.
Thorough investigation of the rock pit began the following morning, March 31. As shown by figure 12, section 12, a considerable quantity of salt water had flowed into the pit. The chloride content in the rock pit was low compared to that of water in the canal because, as the salt water flowed downward into the
EXPLANATION
CONTOUR ON LAKE BOTTOM 3D 0 300 EO
IN FEET BELOW MSL
P9 /.b \\ '. \
SALT-WATER MOVEMENT, MIAMs Ap pany
MARCH 30-31,1961
Figure 11. Rock pit showing approximate bottom topography.




24 FLORIDA GEOLOGICAL SURVEY
lake, heavy mixing and dispersion of salt reduced the maximum concentration to less than 2,000 ppm chloride, though water with a concentration of 5,000 ppm was observed flowing into the pit.
Soundings, made in conjunction with numerous samplings and conductivity traverses during subsequent months, were used to develop a rough contour map of the bottom of the rock pit (fig. 11). Two deep holes nearly 40 feet below sea level are separated by a ridge having an altitude of about 25 feet below msl. In these holes most of the salt water samples were collected at depths greater than 20 feet below msl. The profile of the pit bottom in the canal sections (fig. 8, 9, 12-15) is generalized and represents no particular section line.
FLUSHING OF SALT WATER FROM THE CANAL
Once having entered deep holes such as those in the rock pit, the heavy salt water cannot easily be retrieved and flushed. As most of the salt water remaining in the canal would eventually flow into the rock pit, the canal was flushed on March 31, the day after closing of the control gates. At high tide, at 1:30 p.m., two gates of S-29 were opened fully. The distribution of salt water before flushing is shown in figure 12, section 12, and the progress of flushing is shown in sections 13 and 14.
The principle of flushing is based on the tidal fluctuation of the ocean. In Miami, the range of fluctuation from high tide to low tide amounts to about 2 feet. Assuming that the water level upstream from the control dam is at least as high as the level of the ocean at high tide, the dam can be opened at high tide or shortly thereafter. The falling tide level continually creates a head differential between inland parts of the canal and the ocean; thus, for a period of about 6 hours, from high tide to low tide, a steep seaward hydraulic gradient produces a high rate of discharge from the canal. The discharge reaches a maximum slightly before low tide (fig. 6). Referring to the velocity curve in figure 8B, section 7, the seaward velocity of the fresh water just below low tide can be estimated at about 2 fps (feet per second) while that of the salt water can be estimated at about 0.7 fps. At this velocity, the salt water will move seaward at a rate of at least half a mile per hour. During the 6 hours of falling tide, flushing of 2 to 3 miles of the downstream reach of a canal is feasible. Turbulence disperses large quantities of salt water upward into the high-velocity fresh water and the flushing is probably more effective than the calculation indicates.




REPORT OF INVESTIGATIONS No. 24 25
-. C3
I I. 4 i W DISTANCE FROM CONTROL DAM;IN FEET S 11f,000 10 0Q 6,: 40 0
'0j) MARCH 31 1961 (850A.M.)
WATER JI&FAE-,
ISOGrHLORS.LN PARTS PER MILLION.
sonaos mar.
-ac
- 4 . .. co) MAR.? MAFL. MAR.29 MAR.30 MAR.31 APR.ICO S- 6A 2 I 1R P IA P IP 12 G SA I2 SP 12 SA &P2P
1T-1 I-1 ,r IF : II ~ : -! V ]T WATER SURFACE
oc ISDCHLORS, IN PARTS PER 0: 4 (
Io
(14) MARCH 31,1961 (4:40 P.M.)
~~WATER SUIRFAGE ,,A
4 -' ........~~~~ ~~ ~~.... . . . . . .. ., .,,2 ... . .. .-. :. . .. ....
ICA 5 IQG GP1 A I1 A 12I P 2 GA P 12b 2 SP t2 612CGPOPN
.44... i L ARI 41~ P ... ...... O' R
ME I4 :
-4c
q (15) APRIL 31,1961 (:40 PA.M.) KZ WATER SURFACE-,
5.1 ISOCHLORS,IN PARTS PER MILLION 0 I, -tC _- 5
i g 2 SAI eit-n 1 S1 1 24 an A 2 hog t Sn e I sh ow
momn o i~ w a a f "1I (1E APRI IN919:0.. i I 11N2
-40 ."
(15 APRI 1. t;i (90 ..
- .:. ...i S I....... ....,.
Figure 12. Sections 12, 13, 14, and 15 through the Snake Creek Canal showing movement of salt water during and after flushing, March 31 to April 1, 1961.




26 FLORIDA GEOLOGICAL SURVEY
During the dry spring of 1962, high easterly winds on March 8 and 9 coincided with high average tides and the wind-driven waters of Biscayne Bay overtopped many control dams in the Miami area at high tide. Salt water invaded the canals at a time when dams throughout the area were held completely closed to maintain heads and to conserve fresh water. In spite of the entrapment of salt water in the canals there was reluctance to flush the canals, as it was felt that the consequent loss of fresh water could not be tolerated.
The series of isochlor sections for unimpeded tidal oscillation (fig. 8A, 8B) show that it is not necessary to have a high freshwater head behind the dam in order to perform a flushing operation. The principle of flushing depends almost entirely upon the falling stage in the ocean itself. The primary requisites are: (1) the canal should be long enough to tap a reservoir of fresh water at inland points, and (2) the control dam should be of a type that permits rapid operation so that opening and closing can be accomplished readily at opportune moments.
Flushing of salt water from a canal entails some loss of fresh water. However, wasting a small portion of the vast reservoir of fresh ground water tapped by a canal is preferable to allowing a small volume of salt water to move inland considerable distances where it may contaminate much larger volumes of fresh water.
ULTIMATE LOSS AND FLUSHING OF THE REMAINING TRAPPED SALT WATER
In the 41/2-hour flushing operation of March 31 not all salt water was removed. As shown by figure 12, section 15, a lens of salt water remained in the downstream reach of the Snake Creek Canal. Also, most of the relatively heavy salt water in the rock pit remained. These separated masses of salty water were traced until July 1961 (fig. 13-15).
The downstream lens in the canal tended to spread toward the rock pit, but a large part of the salt water in the lens disappeared between April 1 and 3 (compare sec. 15, 16). The disappearance coincided with an increase of chloride content in observation wells adjacent to the canal and will be considered in a later section.
The salt water in the rock pit appeared to remain practically stable from April 1 to 18 (sec. 15-19). The primary change during this period was a dispersion-caused widening of the contact between the 1,000- and 200-ppm isochlors.
Between April 18 and May 3, the isochlor positions changed significantly in the rock pit and also in the canal (compare sec. 19,




REPORT OF INVESTIGATIONS NO. 24 27
k R o oC! o
CAn
DISTANCE FROM CONTROL DAM,IN FEET S 14,200 12,000 l000 8000 6000 4000 2000 0
(16) APRIL 3, 1961 A RF
WATER SURFACE,1: - -
0 ISOCLORSN PARTS PER MILON
IC
c -C
" ..4
-40
-4d 10 ,
Figure 13. Sections 16, 17, and 18 through the Snake Creek Canal showing
changes in isohor positions, April 3-11, 19.) APRIL 4,1961
20). An increase in chloride content was sensed by the conductivity recorder at the footbridge between April 22 and 26 (fig. 5) and, therefore, this was probably the period when the distribution of chloride content in the rock pit changed.
In the bar graph in figure 14, lower part of section 19, the wind movement exceeds 100 miles per day for 3 consecutive days, April 21-23; this period nearly coincides with the period of increase in
WA rER SURFA CE
1u ISOCHLORS,IN PARTS PER MILLION /
CLP
wind mixed the water in the upper part of the roc it. The
C....
a.. CONTROL tu /00CLOSED
4. -0- o
50 0
S-I0
Lt;Do
2000
-2 00
Figure 13. Sections -16, 1'7, and 18 through the Snake Creek Canal showing
changes in. isochior positions, April 3-11, 1961.
20). An increase in chloride content was sensed by the conductivity recorder at the footbridge between April 22 and 26 (fig. 5) and, therefore, this was probably the period when the distribution of chloride content in the rock pit changed.
In the bar graph in figure 14, lower part of section 19, the wind movement exceeds 100, miles per day for 3 consecutive days, April 21-23; this period nearly coincides with the period of increase in chloride content at the footbridge. Apparently, the 3 days of -strong wind mixed-'the water -in the upper part of the rock pit. The




28 FLORIDA GEOLOGICAL SURVEY
DISTANCE FROM GONTROL DAM,IN FEET
. ,000oo ,2,000 so, ooo sooo sooo 4000 sooo a
APRIL 18,1961
I[( IIIARL I6 A I 1 1 11
PRLQ26WATER S URFAC4 -.
moi
I SOCMLORSIn PARTS PER MILLION
CONTROL
CLOSED
-IC r tZALEAH GAGE05
-40
1 (2 M) mAY 36.1961
0 WATER SURFACE-, SOCALotIN PATS PER MILuON
an CON t I., 00=
'A'6 _CLOSED
'-40
C
(II:
..
Figure 14. Sections 19, 20, and 21 through the Snake Creek Canal showing
changes in isochior positions, April 18 to May 16, 1961.
downward movement of isochiors in section 20 represents dispersion and upward movement of salt ions into the uppermost water of the rock pit. Small hydraulic gradients, associated with the wind and with ground-water discharge in the downstream reach of the canal, then carried the salty water downstream where it was sensed by the conductivity recorder at the footbridge.
Wave height on a given lake depends on the length of time the wind blows as well as on the wind speed. Therefore, the 3 consecutive days of high wind, April 21-23, were far more effective in




REPORT OF INVESTIGATIONS No. 24 29
stirring the waters of the rock pit than were the single days of similar wind speed in early April. The downward movement of isochlors between May 3 and 16 (sec. 20-21) undoubtedly relates to the relatively high sustained winds of May 6 to 9. It is significant that wind and waves serve as agents for removal of trapped salt water.
On May 26, after heavy rainfall, the dam was opened (fig. 5). The resulting tidal oscillation and discharge produced small currents in the bottom of the rock pit. This agitation reduced the chloride content and density more in the eastern hole (fig. 11), by upward dispersion, and the salt was gradually flushed toward the partly open control dam. Although residual traces of the original contamination remained, the partial gate openings that started March 26 were effective in reducing salinity even at the very bottom of the rock pit (fig. 15). The decrease in chloride content at the footbridge (fig. 5) at the same time indicates that the tidal impulse of the ocean can be utilized .to remove salt water that has become trapped in deep pits or lakes along the course of a canal.
SALT-WATER MOVEMENT IN THE AQUIFER
The short-term movements of salt water in the aquifer are closely influenced by happenings in the canal. Fluctuations of chloride content in wells along the south bank of the canal during 1960 and 1961 are shown in figures 16-20. During the special test of March 27-31, water samples were collected as often as four times daily and these are plotted on an expanded time scale.
NATURAL TIDAL OSCILLATION
Under unimpeded tidal oscillation, water moves from the aquifer into the canal on a falling tide and from the canal into the aquifer on a rising tide. This to-and-fro motion prevailed when the gates were open March 27 to 30 and is indicated by the two-headed arrow in the March 30 section of the sequence of insets in figure 19. During this period the chloride content in the shallow wells fluctuated only slightly at stations 1 to 5 (fig. 16-18), and practically none at the more landward stations (fig. 19-20). Although highsalinity water oscillated landward beyond station 8 in the canal (fig. 8), salt water did not appear in wells located as little as 5 feet from the canal. This is not surprising. The net flow of fresh ground water in the upper part of the aquifer is toward the canal. Any small amount of salt water that penetrates the aquifer




3so FLORIDA GEOLOGICAL SURVEY
DISTANCE FROM CONTROL DAM,IN FEET 14000 12,000 1Q000 8000 6000 4000 2000 0 :(22) MAY 29.,961
I 2 q . PI 30 I1a IILIO
-IC
WA0 29.JN 1,1961
-3.
'1 1Ju., . ;~0l;
S (22) AmY 2o.18es
0 WATER SURFACE-.
Ia C
ISaGLORSIN PARTS PER MILLION
-Ic
ISOCIIO. NS, N PANTS PEI I|LLION
CON TROL
-20
Figure 15. Sections 22, 23, and 24 through the Snake Creek Canal showing
-4C3,
changes in isochlor positions May 29 to July 10, 1961.
on a rising tide is flushed back to the canal on the subsequent facing tide. Thus, the average flow of fresh ground water toward the canal opposes and greatly minimizes movement of" salt water from the canal to the aquifer during unimpeded tidal oscillation.
A different situation exists in the deep part of the aquifer. The
~ I1- ULl0~96lWATER SURfACEISOLNIIONSIN PARTS PER MILLION
CONTROL
-3
average head in the aquifer is lowered when the gates are open.
cha4ge inL iso0.o poiinsMy29t6uy1091
This causes salt water, deep in the aquifer, to move landward. The
_i..
aveurae h5edtins the23 aufe is lhowerdg e the gateCres Caare sopen. Thi asesg std waer fuhdepain teauro mhanlovte sudwrd.qTet
movement is slow but relentless. An example is the increase in chloride content from 6,800 to 9,200 ppm in well G 420A between




20,000 II _ .._,__ _. .. "-,.ll .
Gills "11 16
0.000 -S1115
10,00 "'MAP VIEW SECTION VIEW
STATIONS 1,2,ond 3 1SoTA1 ION I OA r*I 1 *-o0 -0 (TAT N 3) 01120*
N0 IllS
zG1120 50 '00 150 Fst 7ll Q1 112 G 112s2
a.EPN DD,4GI
10 "_'1_'",_I1) I Sc 2 o 1122 . sO ot v r2 ,
.'~r ,If- .'.:.......-.- ,
F ig ure.16.G h l o h c at
1,EXPAND, a 11
0 1 : 461
Oct NOV. sib. JAN. FE.. MAR."--" MAY JUNE JULY AUG. SEP T . .. NOV. DEC.
1960 I I
Figure 16. Graph showing fluctuations of chloride content at well stations 1, 2, and 3, 1960-61.
\ J ..__ '
.. '-=. J./ / \,. ",, .....:: :':::-




000,MAP VIEW SETIO V 4
a| STATION 4 9 I *I o
LI
0 112 9 - 112 aod
0.1129-um .
. NOV. D N. MI AN. g MA e
Figure 17. Graph showing fluctuations of chloride content at well station
4. 1960-61.
100 ..... ......... !., .. .... .. ."
flit~ ~ ~ ~ ~ ~ ~~~~~~~~... iO... ",---"", ... ,,_ -,..I. .. .. .
o ~ ~ ~ ~ ~ ~ ~ ~ .. .....r" \ - 2 '"____ ''''i ... ... .. . . .......... .. ....... .
' "- "'6~ ~ 1128 -',{ 2
- EXPANDED-..TIME SCALE
0 RO I a ao- : ''' It NO''V'' .
Val N.OV. ,c. I A. I ,,,. MA-. APR. MAY JUN I ,-. SEPT. I OCT. ec.
Figure 17. Graph showing fluctuations of chloride content at well station 4p 1960-61.




42 G 420 :~7l0 0... 1109 a, STATION 5 ~t .0 ,
MAP VgE SECTION VIEW
G 1110
-I ., -G I2I/
co-oo..... I II2
OEN~~111 7 0/ VIP 11
MU ti-N.U G1I2 EXPANDED .
SCALE TIME SCALE
P
DEC. 4 CT. NOV DEC. JAN. FEt. MAR aAPR MAY JUNE JULY AUG. SEPT. OCT. NOV DEC.
IrS 199 to ..... gg
Figure 18. Graph showing fluctuations of chloride content at well station 5, 1959-61.




24000 -1 r MAP VIEW SIECTIOS VIEW A
Si-I..u .. U-,- , I f 1"1 "9 Il VO "9 19 9I q0 9 0
1000 -- a 9isr m
*0- 1138 aU 3 5
- r/ **r*/M to Ae .W
g a s ss umevnmauo
umCA W3AsMCA wave anae aM stase 11oo m 9ew30sesAm
ILstaunct starse sHOm c4Ann coErt (PPM) ouRNs AND AFTER
1956
G 11381114
I ...
G 1154 G 11
TIM 1SCALE 1.
z
1960 1 614
Figure 19. Graph showing fluctuations of chloride content at well station 6, 1960-61.




200 I 11 I I I I i I
STATION 9 G1156 - -G 1155
100 G IIll53 ..1 1I
1154 6. 115 / G 1154
-...,,1154 A -i
L- G 1155 G .15 EXPANDED m$ vOsw s o CI
/ G 1156 'TIME SCALE s. o
101
22000
,, o ones "eo e"o1se
, ,IIII, IIIIIIII I" I' on, e e
I I I I I I1 1 1 1 1 1, I I I I I I
STATION 8 G 15-...11 461
GII4 Ns:- G1146 G I I
G 1147'
OV -G 11G EX A P M JLJ
4eso-leel
UX NEO.TME SCALE' o,0,:2;;_;
oo aa
z a
7, 8, and 9, 1960-61.




36 FLORIDA GEOLOGICAL SURVEY
June and October 1960, when the dam was open for hurricane Donna floods (fig. 18).
RECHARGING THE AQUIFER WITH TRAPPED SALT WATER
The canal is quickly converted from a drainage canal to a recharge canal by closing the control. The water which happens to be in the canal at this time, either fresh or salty, is driven from the canal under hydraulic gradient and appears in wells adjacent to the canal.
All well stations (fig. 16-20) showed an increase of chloride content on the morning of March 31, the day after salt water was trapped. As will be noted by comparing the graphs, the degree of reaction to the recharge of salt water varied. A partial list of the factors involved is as follows: (1) The distance of the well from the canal; (2) the altitude of the terminus of the well in relation to the altitude of salt water in the canal; (3) the concentration of salt water, which varies both vertically and longitudinally along the canal; (4) the amount of dispersion of salt as the water passes through the rock; and (5) the local permeability of the aquifer near the well station.
The relations of salt-water movement at station 6 are shown in the section views of figure 19. Section a for March 30 shows that salt water had not invaded the aquifer during unimpeded tidal oscillation.
In section b (10:00 a.m., Mar. 31), approximately 24 hours after closing the control gates, the chloride content in well G 1138, 5 feet from the canal, had increased from 60 to 13,700 ppm while that in well G 1140, 65 feet from the canal, increased from 60 to 288 ppm. The sharp increase of chloride content in these wells, in contrast to the decrease in well G 1136 (about 22 feet deeper) from 125 to 102 ppm, indicates that the salt water moved nearly horizontally. This is a reasonable observation because the horizontal permeability is generally much greater than the vertical permeability in stratified aquifers.
Section c (4:00 p.m., Mar. 31) shows the flow direction while the canal was being flushed. The sharp decrease of chloride content in wells G 1138 and G 1140 in the fluctuation graph at this time indicates that part of the salt water, that had previously invaded the aquifer, was flushed back toward the canal. After the flushing operation was completed and the dam was closed, the unflushed salt water in the aquifer resumed a flow direction away from the canal (sec. d).




REPORT OF INVESTIGATIONS No0. 24 37
In addition, salt water from the salt-water lens, which remained in the downstream reach of the canal after flushing (fig. 12, sec. 15, and fig. 13, sec. 16),moved into the aquifer. The disappearance of the salt-water lens from the canal by April 3 coincided with the sharp decrease in chloride content in well G 1138 on that date; the recharge water in the canal had become relatively fresh but the increase of chloride content in well G 1140 between April 1 and 3 showed movement of the salt-water slug farther from the canal. The low maximum chloride content in well G 1140 as compared to that of well G 1138 indicates dilution of the salt water during movement between the wells.
It is interesting to note that the chloride-content curve for well G 1138 crosses over and becomes lower than that for well G 1140 after April 5 (fluctuation graph, fig. 19). As previously shown, water in the canal continued to freshen with time. Thus, as the slug of salt water passed farther into the aquifer it was followed by relatively fresh water which appeared first in well G 1138, near the canal bank, and then in well G 1140. The reversal of the chloride-content curves shows the cause-effect relation of salinity in the canal on salinity in the aquifer.
The chloride content in well G 1137 reached a maximum on April 7, more than a week after the control was closed (fluctuation graph, fig. 19). During this week, some of the salt water that remained at the bottom of the canal was dispersed upward in the canal by wind agitation (fig. 13, sec. 16, 17). As the shallow wells were recharged by water from the upper part of the canal, the time lag of salinity fluctuation in the shallow wells reflects the changes of salinity in the upper part of the canal. The lower peak chloride content in well G 1139, as compared to well G 1137 again demonstrates the reduction of salinity by dispersion of salt as ground water moves through the pore spaces of the aquifer.
Well G 1136 is 22 feet deeper than well G 1138 so that the vertical component of salt-water movement is identified by the relative fluctuations of chloride content in the two wells. The chloride content in well G 1136 did not reach its maximum until early May, about 1 month after the short-term test, whereas that of well G 1138 reached its maximum 1 day after the control was closed (fig. 19). This great difference in time results from several factors. An important driving force in the vertical direction is the difference in density between the overlying heavy salt water and the underlying light fresh water. In open water, an unstable density condition of this type is corrected rapidly by a massive discrete




38 FLORIDA GEOLOGICAL SURVEY
overturn of the two fluids. The settling of salt water in the rock pit is an example. However, in an aquifer the pores of the rock are small and the fluid movement is split into similarly small flow filaments. The fresh water rises at the same time that the salt water sinks. The overturn is reduced to slow infiltration wherein dispersion of salt plays an important part in modification of the original concentrations. The effect of trapping salt water in March 1961 is recognizable until October 1961 in the fluctuation curve for well G 1136 (fig. 19).
In the longitudinal sections of figure 7, the downward movement of salt water is depicted as a lens of salt water that appears near the bottom of the canal and progressively drifts downward with time (sections of Mar. 31 to May 16, 1961). At the same time there is a horizontal component of flow perpendicular to and away from the canal; this movement is shown by the arrows in section e of figure 19.
ULTIMATE LOSS AND FLUSHING OF SALT WATER IN THE AQUIFER
Comparison of the fluctuations of chloride content at all stations (fig. 16-20) shows that the deep wells were little affected by the short-term test but that the shallow wells were affected in a characteristic pattern. This pattern is responsive to the operations of the control dam.
The control gates were closed on March 31, and they remained closed until May 26 (fig. 5). Initially the chloride content in the shallow wells rose sharply and then tended to decrease slightly. At the downstream stations the decrease was somewhat larger because ground-water movement around the control dam and toward the ocean was more rapid there.
At the more inland stations (fig. 20), the aquifer is recharged rapidly until canal water refills that part of the aquifer that is unwatered during drainage. After this, the gradient away from the canal becomes comparatively small as the ground-water flow system approaches a steady state consistent with the boundaries of the aquifer. As the rate of recharge is proportional to the gradient, salt water at first moves rapidly into the aquifer, and subsequently at a much slower pace. The salt water, therefore, remains within reasonable proximity of the canal and is subject to retrieval.
The chloride content in most of the shallow wells decreased sharply after the control dam was opened at the end of May. The




REPORT OF INVESTIGATIONS No. 24 39
fluctuations at station 7 (fig. 20) are especially interesting. The chloride content in wells G 1145 and G 1146 near the canal decreased sharply, whereas that in G 1147 and G 1148, 60 feet from the canal, remained unchanged. Figure 5 shows that the total gate opening was decreased from 30 feet May 26 to 31 to about 5 feet thereafter. At the 30-foot gate opening, the canal caused drainage and some of the salt water flowed out of the aquifer into the canal. Water levels recovered somewhat after closing to the 5-foot opening (fig. 5), and fresh water from the canal recharged the aquifer. The movement of fresh water into the aquifer was sufficient to freshen the water in wells close to the canal (G 1145-G 1146), although water in the wells farther from the canal (G 1147-G 1148) remained relatively salty. Clearly, part of the salt water trapped in March had been removed but a part still remained as an isolated mass. During the next gate openings on August 20 .the remaining salt water was completely flushed from the aquifer and the chloride contents of all four wells became nearly the same. Thus, the low-salinity water trapped in March was recovered from the shallow part of the aquifer by August 1961.
The recovery was possible because the density of this lowsalinity water (less than 200 ppm) was not great enough to cause a large downward movement. The dense, high-salinity water that percolated downward at station 6 (shown by the graph of well G 1136, fig. 19) obviously was not recovered. It continued to move downward until the density contrast was eliminated by confluence with water of the same density in the deep part of the aquifer.
MANIPULATION OF THE CONTROL DAM FOR MAXIMUM BENEFIT
The primary function of a canal in the Miami area is to remove surplus floodwaters. As floods in Miami result from heavy rainfall, it is obvious that the floodwaters must always be fresh. If salt water penetrates landward in a canal, the drainage function of the canal is impaired because floodwaters cannot flow seaward through that part of the canal in which salt water flows landward; therefore, the optimum removal of floodwaters occurs when no salt water invades the canal. Whether the flow condition is steady state or transient due to tide, it should be recognized that intrusion of salt water in a canal is an obvious sign of reduced efficiency in the discharge of fresh floodwaters.
All four gates of S-29 were fully open for nearly 2 months after hurricane Donna passed on September 10, 1960 (fig. 3). Only




40 FLORIDA GEOLOGICAL SURVEY
during September 10 to 14 was the discharge great enough to keep the canal flushed of salt water. At all other times during the period of record shown in figure 5, full openings of the gates permitted salt water to move upstream on a rising tide. This suggests that the control was operated at full-open position much longer than necessary.
FACTORS AFFECTING INLAND MOVEMENT OF SALT WATER
The manner in which the total gate opening is manipulated has a bearing on the landward extension of salt water in the canal. For example, a gate opening of 30 feet may be represented by a 7.5-foot opening for each of the 4 gates or a 15-foot opening for each of only two gates (the other 2 remaining closed).' The schematic flow diagrams of figure 21 illustrate the hydraulic advantage that attaches to uniform distribution of the gate opening. In figure 21a fresh water is forced to flow downward around the obstruction of the gate. This downward displacement of the fresh-water discharge section impedes the upstream intrusion of salt water. Uniform opening and submersion of the four sluice gates (in contrast to fully open gates) can be used as a device to distort the flow pattern deliberately in order to prevent saltwater intrusion.
Questions arise as to whether the movement of salt water into a canal is ever justified on the basis that widest opening of the control gates will cause greatest discharge of fresh floodwater. The critical point of entrance of salt water into a canal is at the control dam, for here is the only place where intrusion can be controlled. The problem involves the total head at the dam, the amount of discharge at various gate openings under these various heads, and the influence of fresh-water head in preventing upstream movement of salt water beyond the control dam.
Using the principle of static equilibrium for fluids of different density, salt water will not move upstream as long as the head differential between the fresh-water and salt-water sides of the control gate is greater than 0.3 foot (for a canal whose bottom is 12 feet below sea level). This value is obtained by balancing the
'The values used here only demonstrate a principle; the discharge area for 2 gates opened 15 feet is not equal to the discharge area for 4 gates opened 7.5 feet because the gates at wide-open position may be above the actual water surface by as much as 3 feet.




REPORT OF INVESTIGATIONS NO. 24 41
pressure exerted by fresh water against the opposing pressure exerted by sea water at the bottom of. the canal.
Pf = pfg If PH pa g ls
pr g 1I = p, g 1s
letting 1 = h + 1. (See fig. 21a):
pf (h + 1) = p. 1
h = 1. (p. pf)
Pf
h = 12 (1.025 1.000) = 0.3 ft.
1.000
Pf = pressure exerted by fresh water
Ps = pressure exerted by sea water
g = acceleration of gravity
pf = 1.000, density of fresh water
pS = 1.025, density of sea water
1 = length of fresh-water column, in feet
1s = 12 feet, length of sea-water column
h = head differential between fresh-water and sea-water sides of control
dam, in feet.
Referring to the schematic flow diagram of figure 21b, when the gates are fully open, the head loss through the control structure itself is very small. The flood conditions would have to be very severe in order to maintain 0.3 foot of fresh-water head at the control dam near high tide, when the greatest possibility of inland movement of salt water exists. The reason for upstream penetration of salt water on September 15, 1960 (fig. 5), only 5 days after the passage of hurricane Donna, is evident. With fully open gates, the head differential and discharge decrease very rapidly until salt water is able to flow upstream during the rising tide. Submergence of the gates, however, tends to focus all the head loss at the control (fig. 21a). This is very desirable and if the four gates of S-29 are uniformly adjusted to maintain a head differential of 0.3 foot during the rising tide, no salt water will invade the canal.
DISCHARGE RELATED TO GATE OPENING
The quantity of water discharged under a sluice gate depends on the head differential across the gate and on the cross-sectional area through which the discharge occurs.
Figure 22 shows the approximate discharge related to head differential for various openings of one gate of the control dam in the -Snake Creek Canal. The plotted points represent the mean daily discharge at West Dixie Highway and the head differential,




42 FLORIDA GEOLOGICAL SURVEY
(0)
L0
WATER SURFACE
f Zt I RESHWAE
RESH WATER SALT WATER
(b)
WATER SURFACE
FRESH WATER
INTERFACE -C SALT WATER
Figure 21. Schematic flow diagrams: a, with gate partially submerged; b, with gate fully open.
as obtained from the difference in mean daily gage height at West Dixie Highway and Biscayne Bay at NE 84th Street. Comparison of past records indicated that the mean daily gage heights of Biscayne Bay at NE 84th Street and the now discontinued gaging station just below the control dam in Snake Creek Canal were comparable when the control dam was closed or the discharge was within the range shown in figure 22.
The curves were calculated by using the formula Q = ACV2 gh where Q is the discharge, A is the cross-sectional area, C is. the coefficient of friction, g is the acceleration of gravity, and h is the




REPORT OF INVESTIGATIONS No. 24 43
DISCHARGE, IN CFS
0 200 400 600 800 1000 1200 1400
on
0
o.
o
Hd 0
Z 0
o o
oo o
-2
0j 0 00
0 0
0 0 0 0 Uj 0
Woo o,
/
IL
0-
ILJ //
00 w0
-H EAD DIF FERENTIAL THAT WILL. PREVENT UPSTREAM MOVEMENT OF SALT WATER Figure 22. Graph showing relations of head differential and discharge for various openings of one gate of the control dam in the Snake Creek Canal.
head differential through the gate. As Q, A, and H were known for each of the points in figure 22, the coefficient of friction (C) was evaluated for the different gate openings. Then, using the average value of C and holding the area (A) constant for the individual gate openings, the curves were constructed through the mean value of the plotted points by solving the formula for Q while letting the head differential (h) vary.
The curves may be used to estimate low discharge for the control structure provided the upstream tidal range is less than 0.5 foot. The accuracy decreases when the tidal range of the upstream stage recorder exceeds 0.5 foot during a given day.
The following table shows the recession iof discharge following the passage of hurricane Dohna on September 10, 1960.




44 FLORIDA GEOLOGICAL SURVEY
Date Mean daily head Mean daily
(September) differential (feet) discharge (cfs)
11 0.48 2,430 12 .38 2,180 13 .31 1,800 14 .28 1,660 15 .23 1,510 16 .17 1,290 17 .19 1,240 18 .22 943
Salt water invaded the canal on a rising tide on September 15. A comparison of the measured discharges in the table (all gates fully open) with the plotted points in figure 22 (only one gate partially open) indicates that the four gates can be partly closed within a week after a flood peak has occurred to maintain a head differential of at least 0.3 foot. This would keep the salt water out of the canal. The small increase in head required to maintain the same discharge would be insignificant from the standpoint of flood height above the dam.
There is a distinct difference between holding salt water back, as shown in the schematic flow diagram of figure 21a, and allowing salt water to invade the canal, as in figure 21b. Inland penetration of salt water does not contribute to seaward discharge because, as previously defined, true removal of floodwaters must only mean removal of fresh floodwaters. Salt water which penetrates landward in the canal merely replaces fresh water; once it has gained entry, energy must be expended to remove it. In the velocity profiles of figures 8A and 8B, the velocities near the bottom of the canal are landward during the rising tide, and are small relative to the seaward velocities in the upper part of the canal during the falling tide. These velocities reflect two things: (1) Sea water provides back pressure which is greatest at the bottom of the canal. (2) Energy is expended in driving the intruded salt water back to the ocean on the falling tide. As the noncontributing landward flows and the reduced seaward flows of salt water must be integrated with the positive flow of fresh water to yield the mean daily discharge, the mere presence of salt water in the canal, in effect, reduces the mean daily discharge of true (fresh) floodwaters.
In some respects, salt water acts as a movable dam in the bottom of the canal; it effectively reduces the depth of the canal




REPORT OF INVESTIGATIONS No. 24 45
through which fresh water can flow seaward. For example, if salt water occupies the lower 6 feet of the canal when the control is fully open (fig. 8), the effective flow section for fresh water is reduced to a thickness of about 6 feet (total depth of canal is 12 feet). If the gates are lowered to an opening of, say, 7 feet, the fresh-water discharge will be impeded and this will cause the freshwater head to rise until salt water is driven out of the lower part of the canal. Conceivably, the head might build up sufficiently so that the fresh water would flow through the full 7-foot section below the sluice gate. However, there is no assurance that this will occur-fresh water may discharge through only the upper part of the 7-foot thick section while salt water continues to flow landward through the lower part.
This leads to an important point. The effect upon head and discharge caused by constriction from lowering the gates, as opposed to constriction and back-water effect from salt water in the bottom of the canal, is indeterminate without controlled tests. The optimum gate opening probably occurs when the salt-water wedge is held just downstream from the control and only fresh water is flowing beneath the sluice gates. This gate opening will discharge the maximum amount of water without permitting salt water to invade the canal and aquifer above the control.
Future studies should be directed toward evaluation of the relations of fresh-water to salt-water discharge with various gate openings, in the immediate vicinity of the control dam.
SUMMARY
The Snake Creek Canal cuts through highly permeable limestone of the Biscayne aquifer of southeastern Florida. The canal extends inland about 18 miles from the salinity-control dam located near the shoreline. The canal is somewhat representative of a number of canals in the area and information learned during-the study is generally applicable to other canals.
Salt water was traced in the Snake Creek Canal under various circumstances of control-dam operation during a special test in March 1961. In the first phase of this test all four gates of the control dam were opened wide. Salt water moved inland about 2 miles during the rising tide and seaward during the falling tide. Salt water penetration into the aquifer was negligible during the
3 days the gates were fully open.
In the second phase of the special test the control was completely closed at high tide, trapping salt water above the dam. A density




46 FLORIDA GEOLOGICAL SURVEY
current developed in the trapped salt water, which caused it to flow inland along the bottom of the canal at a rate of about 900 feet per hour until it reached a deep rock pit which acted as a hydraulic sink. A flushing operation removed salt water from the canal, but not that in the rock pit. Measurements of electrical conductivity during the subsequent months in the rock pit showed that the salt dispersed upward by a combination of wind, tidal mixing, and molecular diffusion. After reduction of salinity by this upward dispersion, the salt water in the rock pit was flushed seaward through the canal.
When salt water is trapped in a canal by control-dam operation or by overtopping the dam during abnormally high tides, it is not necessary to wait for nor to have high fresh-water head behind the dam in order to perform a flushing operation. The falling tide (level) of the ocean creates a seaward hydraulic gradient and a condition favorable for a high rate of discharge from the canal. Calculations and observations show that flushing of 2 to 3 miles of the downstream reach of a canal is feasible during one falling tide. The canal should tap a reasonably large reservoir of fresh water and the control dam should be the type that permits rapid operation so that the gates can be opened just after high tide and closed at low tide. Although some fresh water is lost during flushing, good water management is better served by wasting a small portion of a vast reservoir of fresh water than by allowing a small volume of salt water to move inland to contaminate a much larger volume of fresh water.
The movement of salt water in the aquifer adjacent to a canal is closely related to happenings in the canal itself. When the gates are closed, the downstream controlled reach of the canal is immediately converted from a condition of drainage to a condition of recharge. Salt water trapped behind the dam during the test entered the aquifer and quickly appeared in wells adjacent to the canal. Sections perpendicular to the canal showed that the saltwater outseepage from the canal moved much more rapidly horizontally than it did vertically. The residual effect of downward movement of salt water in the aquifer was recognizable for 6 months at one of the observation stations. The fluctuations of chloride content in shallow wells indicate that some of the lowsalinity water in the upper part of the aquifer was eventually recovered and flushed seaward through the canal during subsequent gate openings.




REPORT OF INVESTIGATIONS NO. 24 47
The movement of high salinity water deep in the aquifer responds to head changes related to long-term operations of the control dam. The chloride content in deep wells increased during the fall of 1960 when the control remained open because of flood conditions. The year of 1961 was abnormally dry; infrequent opening of the control caused heads to remain sufficiently high so that general expulsion of salt water from the aquifer occurred throughout 1961. It is a contradictory situation that the greatest inland movement of salt water took place in a wet year and the greatest expulsion of salt water took place during a dry year. Nevertheless, this fact indicates that the means are available for very effective control of the movement of salt water in the Biscayne aquifer.
CONCLUSIONS AS TO OPERATING CRITERIA
The data in this report indicate that salt water need not and should never be trapped in a canal. As the position of salt water in a canal has only a tenuous relation to head, the following conclusions are made with regard to the operation of the gates:
(1) Control gates, which operate automatically by sensing the
head differential between the upstream and downstream sides of the dam, cannot be operated effectively on an automatic basis until the relations of salt-water movement and head are properly evaluated. It is entirely possible for the automatic operations to trap salt water inadvertently.
(2) In.addition to the upstream and downstream stage recorders
which are being installed at the new automatically controlled dams, a conductivity recorder should be installed in each canal slightly upstream from the control dam. Data from these three sources will permit future evaluation of the relation of salt-water movement to total head in the
canal and to head differential through the control dam.
(3) Persons responsible for gate operation should be furnished
a portable conductivity instrument to determine the salinity at the bottom of the canal at different points. Knowledge of the presence and movement of salt water in the canal
will aid greatly in proper operation of the control.
The following suggestions are made for operation after heavy rainfall:
(1) After the first inland penetration of salt water during a
rising tide (which may occur within a week after a flood




48 FLORIDA GEOLOGICAL SURVEY
peak), the gates should be partly closed as conditions
permit.
(2) The gates should always be opened uniformly so that
maximum fresh-water pressure is applied against the intruding salt water by increased velocities of the fresh water
as it flows under the submerged sluice gates.
(3) The instantaneous dynamic head that will prevent inland
movement of salt water cannot be determined without controlled tests. However, the calculated hydrostatic head that will prevent intrusion at the bottom of S-29 is 0.3 foot. Therefore, salt water will not enter the bottom of the canal above the control structure if an instantaneous head of 0.3 foot is maintained between upstream and downstream sides of the dam at the time of minimum head
differential, about 2 hours before high tide.
Some experimentation will be necessary to establish specific operating criteria. Closing the control gates so that 0.3 foot of head will be maintained during the rising tide will tend to increase the stage in the upstream part of the canal by an amount not greater that 0.3 foot. This is a tolerable increase in stage even during floods.




REPORT OF INVESTIGATIONS NO. 24 49
REFERENCES
Cooke, C. W.
1939 Scenery of Florida interpreted by a geologist: Florida Geol.
Survey Bull. 17.
Ferguson, G. E. (see Parker, G. G.)
Kohout, F. A.
1960 Cyclic flow of salt water in the Biscayne aquifer of southeastern
Florida: Geophys. Research Jour., v. 65, no. 7, p. 2133-2141.
1961 Case history of salt water encroachment caused by a storm
sewer in Miami: Am. Water Works Assoc. Jour. v. 53, no. 11, p.
1406-1416.
Leach, S. D.
1963 (and Sherwood, C. B.) Hydrologic studies in the Snake Creek
Canal area, Dade County, Florida: Florida Geol. Survey Rept.
Inv. 24, pt. 3.
Love, S. K. (see Parker, G. G.)
Parker, G. G.
1951 Geologic and hydrologic factors in the perennial yield of the
Biscayne aquifer: Am. Water Works Assoc. Jour., v. 43, no. 10,
p. 817-835.
1955 (Ferguson, G. E., Love, S. K., and others) Water resources of
southeastern Florida, with special reference to the geology and ground water of the Miami area: U. S. Geol. Survey WaterSupply Paper 1255.
Sherwood, C. B. (see Leach, S. D.)




Full Text
xml record header identifier oai:www.uflib.ufl.edu.ufdc:UF0000121100001datestamp 2008-12-05setSpec [UFDC_OAI_SET]metadata oai_dc:dc xmlns:oai_dc http:www.openarchives.orgOAI2.0oai_dc xmlns:dc http:purl.orgdcelements1.1 xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.openarchives.orgOAI2.0oai_dc.xsd dc:title Salt water movement caused by control-dam operation in the Snake Creek Canal, Miami, Flor
(FGS: Report of Investigation 24)Report of investigations - Florida Geological Survey; no. 24, Part IVdc:creator Kohout, F. A.Leach, S. D.dc:subject Saltwater encroachment -- Florida -- Dade Countydc:description Digitized as a collaborative project with the Florida Geological Survey, Florida Department of Environmental Protection.dc:publisher Florida Geological Surveydc:type Bookdc:identifier http://www.uflib.ufl.edu/ufdc/?b=UF00001211&v=00001AAA0979 (notis)AES1338 (notis)dc:source University of Florida


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EY6GINYPA_CSH8RI INGEST_TIME 2017-05-08T18:57:42Z PACKAGE UF00001211_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES



PAGE 1

STATE OF FLORIDA STATE BOARD OF CONSERVATION DIVISION OF GEOLOGY FLORIDA GEOLOGICAL SURVEY Robert O. Vernon, Director REPORT OF INVESTIGATIONS NO. 24 PART IV SALT-WATER MOVEMENT CAUSED BY CONTROL-DAM OPERATION IN THE SNAKE CREEK CANAL, MIAMI, FLORIDA By F. A. Kohout and S. D. Leach U. S. Geological Survey Prepared by the UNITED STATESGEOLOGICAL SURVEY in cooperation with the CENTRAL AND SOUTHERN FLORIDA FLOOD CONTROL DISTRICT Tallahassee ---1964

PAGE 2

FLORIDA STATE BOARD OF CONSERVATION FARRIS BRYANT Governor TOM ADAMS RICHARD ERWIN Secretary of State Attorney General J. EDWIN LARSON RAY E. GREEN Treasurer Comptroller THOMAS D. BAILEY DOYLE CONNER Superintendent of Public Instruction Commissioner of Agriculture W. RANDOLPH HODGES Director ii

PAGE 3

LETTER OF TRANSMITTAL §Jorida ceological Survey Callahassee October 10, 1963 Honorable Farris Bryant, Chairman Florida State Board of Conservation Tallahassee, Florida Dear Governor Bryant: The Florida Geological Survey has published four short papers as Parts 1, 2, 3, and 4 of Report of Investigations No. 24. This publication, Part 4, is the last of these brief papers and discusses the "Salt-Water Movement Caused by Control-Dam Operation in the Snake Creek Canal, Miami, Florida." The report was prepared by F. A. Kohout and S. D. Leach, hydrologists for the United States Geological Survey. The principles of the relationships of fresh and salt water in the Biscayne aquifer and the movement of the salt-water, freshwater contact along the canals and in the aquifer, with use, with development of the control stations in the Central and Southern Florida Flood Control District, and in response to cyclic changes in the weather, are of great importance in understanding similar occurrences. These four reports present interesting details of these relationships. Respectfully yours, Robert O. Vernon Director and State Geologist in.

PAGE 4

Completed manuscript received August 16, 1963 Published for the Florida Geological Survey By E. O. Painter Printing Co. DeLand, Florida iv

PAGE 5

CONTENTS Abstract _-----... .___.__. .__ ____...... ......_...._.................. 1 Introduction __ _____ __.._______... 2 History ________ 2 Purpose of investigation ----------.. ........._...._----------............ --_....... 5 Location and extent of area --_____....___.._--......._.....-..-...__ 6 Climate .--_-_-_. .___. --------_---_ --....___ ____-------------------...--.-_ 7 Previous investigations __ _--.._.. ___-__.. ..-....-................._ 7 Personnel and acknowledgments ___----------._______ ..--__ .__ 7 Geology and drainage -______________ 8 Methods of investigation -____--... ___....._ __-......-------. 9 Long-term position of salt water --.......------__________ .. .............._ 10 In the canal __-----_ ---...-~__------....._.._.... .... .... .. 10 In the aquifer __.___ ... _.......___. ....................__ 12 Short-term position of salt water ___. ..___...... .._ 16 Salt-water movement in the canal -.. ..........._.......____ _.. 17 Natural tidal oscillation __.._................_______ 17 Trapping salt water _20 Flushing of salt water from the canal 24 Ultimate loss and flushing of the remaining trapped salt water 26 Salt-water movement in the aquifer 29 Natural tidal oscillation -----------_______ _ ----.. .-_....... 29 Recharging the aquifer with trapped salt water 36 Ultimate loss and flushing of salt water in the aquifer --__ 38 Manipulation of the control dam for maximum benefit ._ ----__ ----------__ 39 Factors affecting inland movement of salt water ___ 40 Discharge related to gate opening 41 Summary _ __;_____ 45 Conclusions as to operating criteria __ 47 References ------____ ___ -_. ____ 48 ILLUSTRATIONS Figure Page 1 The Miami area, Florida, showing the approximate landward extent of salt water at the base of the Biscayne aquifer in 1961 _ 3 2 Photographs of the salinity-control dam (S-29) in the Snake Creek Canal: (a) with right-hand gate opened 1 foot; (b) with one gate fully open 4 3 Map of the northern part of the Miami area showing observation points along the Snake Creek Canal 6 4 Section along the south bank of the Snake Creek Canal showing generalized lithologic characteristics of the Biscayne aquifer 8 5 Graphs comparing rainfall, gate opening, and fluctuations v

PAGE 6

of water level and chloride content along the Snake Creek Canal, 1960-61 _______.11 6 Graphs comparing fluctuations of discharge, water level, and chloride content in the Snake Creek Canal during the special test of March 27-31, 1961 ______ ____---__ 12 7 Sections along the south bank of the Snake Creek Canal showing the zone of diffusion in the Biscayne aquifer, 1960-62 --. 14 & 15 8A Sections 1, 2, 3, and 4 through the Snake Creek Canal showing movement of salt water during the rising tide, March 29, 1961 _ -___ ______________ -.18 8B Sections 5, 6, 7, and 8 through the Snake Creek Canal showing movement of salt water during the falling tide, March 29, 1961 _______ __----.._ --__. 19 9 Sections 9, 10, and 11 through the Snake Creek Canal showing movement of salt water after closing of the control dam at high tide, March 30, 1961 ___--_-____.._ -... .. ...___21 10 Photograph of rock pit (at left) looking in the downstream direction of the Snake Creek Canal (at right) _---__..-----__..._.22 11 Rock pit showing approximate bottom topography ..-----_-----23 12 Sections 12, 13, 14, and 15 through the Snake Creek Canal showing movement of salt water during and after flushing, March 31 to April 1, 1961 ______ _25 13 Sections 16, 17, and 18 through the Snake Creek Canal showing changes in isochlor position, April 3-11, 1961 _ ..--.27 14 Sections 19, 20, and 21 through the Snake Creek Canal showing changes in isochlor positions, April 18 to May 16, 1961 .---. 28 15 Sections 22, 23, and 24 through the Snake Creek Canal showing changes in isochlor positions May 29 to July 10, 1961 .---30 16 Graph showing fluctuations of chloride content at well stations 1, 2, and 3, 1960-61 ._-----._ .. ------...-31 17 Graph showing fluctuations of chloride content at well station 4, 1960-61 _ ___ ..... .._ ...__....---------... 32 18 Graph showing fluctuations of chloride content at well station 5, 1959-61 ____ ----33 19 Graph showing fluctuations of chloride content at well station 6, 1960-61 .__ ... ...-__---_...._ ...._.. 34 20 Graph showing fluctuations of chloride content at well stations 7, 8, and 9, 1960-61 _-_ .__-_ -_35 21 Schematic flow diagrams: (a) with gate partially submerged, (b) with gate fully open ------___42 22 Graph showing relations of head differential and discharge for various openings of one gate of the control dam in the Snake Creek Canal ______ 43 vi

PAGE 7

SALT-WATER MOVEMENT CAUSED BY CONTROL-DAM OPERATION IN THE SNAKE CREEK CANAL, MIAMI, FLORIDA By F. A. Kohout and S. D. Leach ABSTRACT Movement of salt water in the Biscayne aquifer and the Snake Creek Canal was investigated to establish criteria for operation of the salinity-control dam in the canal. All four gates of the dam were opened for 3 days during a special test in March 1961. Although salt water oscillated landward and seaward a distance of 2 miles from the control in response to tide, the salinity of the ground water near the canal did not change greatly during this time. Salt water trapped in the canal by closing the dam at high tide moved landward as a density current at a rate of about 900 feet per hour and began to flow into a connected deep rock pit. A flushing operation removed salt water from the canal but the salt water in the rock pit remained. During subsequent months, wind and tide action caused upward dispersion of salt into the uppermost water of the rock pit, and the trapped salt water was gradually removed by seaward discharge through the canal. Some of the salt water trapped in the canal moved into the aquifer. This salty water was traced by sampling fully cased wells of different depths and distances from the canal. Some of the salty water was retrieved by drainage from the aquifer when the gates were partially opened 2 months after entrapment. Because of its relatively great density some of the salt water moved downward in the aquifer and was not retrieved. The data show the negative effects of trapping salt water, and also that the control can be closed gradually without trapping salt water. If the total opening of the control is distributed uniformly among the four partially closed gates, the increased velocity of the fresh water as it flows under the submerged sluice gates, will impede upstream movement of salt water at the bottom of the control. Calculations indicate that salt water probably will not move upstream in the canal if a head of 0.3 foot is maintained between upstream and downstream sides of the control at the time of minimum head differential, about 2 hours before high tide. 1

PAGE 8

2 FLORIDA GEOLOGICAL SURVEY INTRODUCTION This report gives detailed attention to the movement of salt water in and adjacent to the Snake Creek Canal in the northern part of the Miami area, Florida (fig. 1). The movement results primarily from operations of a salinity-control dam located near the mouth of the canal. From 1909 to 1930, canals were constructed westward from the coast through the Miami area into the Everglades to drain lowlying land for urban and agricultural use. The uncontrolled canals caused overdrainage. The water table was lowered nearly to sea level in some areas and this caused salt water to encroach into the Biscayne aquifer by two processes (Parker, 1951, p. 826): (1) Where the water table was lowered greatly, salt water moved horizontally landward from the ocean through the highly permeable solution-riddled limestone of the Biscayne aquifer to replace the drained fresh water. (2) Salt water moved directly inland in the canals where, because of its greater density, it tended to leak downward and laterally to contaminate the underlying fresh ground water. Although the end result of these separate processes is the same, they attack the fresh-water resources from different directions and occur at different rates. The general aim of this report is to establish the conditions under which each process provides the greater danger of salt-water contamination. HISTORY In 1945, salinity-control dams were installed in most of the canals of the Miami area as barriers against further encroachment of salt water. These were emergency dams constructed by driving sheet-steel piles into the limestone. The dams were opened and closed by removing some or all of the piles during the rainy season and replacing them during the dry season. This was an inefficient procedure which frequently resulted in the damsbeing left open much longer than necessary. In recent years, the emergency dams have been replaced by permanent concrete structures with submerged sluice-gate openings. The control structure S-29 (Snake Creek Canal) is representative of some of the more recent structures. Each of the four gates, figure 2, is connected through a gear box to rollers which are driven by the rear-wheel drive of an anchored automobile. In the most recently contructed controls, in the Little River and

PAGE 9

REPORT OF INVESTIGATIONS No. 24 3 SBISCAY LE I ML I if::',:".:: '. y~ a -vi : -" .. Figure 1. Miami :area, Florida, showing the approximate landward extent of salt water at the b6ase of the Biscayne aquiferin 1961.

PAGE 10

4 FLORIDA GEOLOGICAL SURVEY r7 me. ._ _;. T O Figure 2. Photographs of the salinity-control dam (S-29) in the Snake Creek Canal: (a) with right-handgate opened 1 foot; (b) with one gate fully open.

PAGE 11

REPORT OF INVESTIGATIONS NO. 24 5 Biscayne Canals, electric motors raise and lower the gates automatically by sensing the water levels upstream and downstream from the control dam. These automatically operated dams permit delicate control of water levels. However, the relation of salt-water movement to head differential needs further study before fresh-water resources can be effectively entrusted to automatic head-activated devices. As will be shown later, the relations of head and salt-water movement are very complicated and transient, and much more study and experience should be accumulated before true automatic control is attempted. PURPOSE OF INVESTIGATION Frequently, after heavy rainfall, headwater areas of the Snake Creek Canal remain flooded at the same time that fresh-water discharge at the control dam is not sufficient, with fully open gates, to keep the canal flushed of salt water. Under these circumstances salt water moves upstream in the lower reaches of the canal and contaminates the fresh ground water. Questions had arisen as to the magnitude of this encroachment, the extent to which the encroachment might be tolerated during gate openings, and the feasibility of flushing the salt water out later when the gates were closed after flooding subsided. Accordingly, goals of the investigation were to establish the timerate of movement of the salt water for various circumstances of dam operation and, if possible, to develop principles that might have practical application not only in the operation of the salinitycontrol dam in the Snake Creek Canal but also in other canals of the area. The investigation was made in cooperation with the Central and Southern Florida Flood Control District (C&SFFCD). Most of the data were collected between October 1960 and January 1962. The investigation was under the general supervision of M. I. Rorabaugh and, subsequently, C. S. Conover, Tallahassee, A. O. Patterson and K. A. Mac Kichan, Ocala, district engineers respectively of the Ground Water, Surface Water and Quality of Water Branches, U.S. Geological Survey. The investigation was under the immediate supervision of Howard Klein, geologist-incharge, Ground Water Branch, and J. H. Hartwell, engineer-incharge, Surface Water Branch, Miami.

PAGE 12

6 FLORIDA GEOLOGICAL SURVEY LOCATION AND EXTENT OF AREA The Snake Creek Canal extends about 18 miles inland to the vicinity of Levee 33, on the east side of Conservation Area 3 (Inset map, fig. 3). Under the plans of the C&SFFCD (the operating agency) and the U. S. Corps of Engineers (the constructing agency), Conservation Area 3 will impound water pumped westward from urbanized areas during the rainy season. During the dry season, water stored in the conservation area will be delivered coastward to maintain water levels, thus preventing salt-water encroachment and providing replenishment water for municipal and private well systems. This report covers the seaward reach of the Snake Creek Canal (Canal C-9) extending about 15,000 feet westward from the salinity-control dam (structure S-29). The width of the area extends 200 feet on either side of the centerline of the canal, but ROCK FOOTBRIDG .p f t i i a i i i l Figure 3. Northern part of the Miami area showing observation points along Snake Creek Canal.

PAGE 13

REPORT OF INVESTIGATIONS NO. 24 7 as one side should be the mirror image of the other, work was confined to the south bank of the canal. The canal widens into a lake whose dimensions are about 3,500 by 1,700 feet at the western end of the area of investigation, between State Highway 9 and Miami Gardens Drive (fig. 3). The lake, formed by removal of rock for concrete aggregate and other building materials, has a maximum depth of 40 feet below msl (mean sea level) and was included in the investigation because of its important influence on salt-water contamination. CLIMATE The climate of the Miami area is subtropical. The average temperature for January is about 68°F and for July about 820F. Rainfall averages about 60 inches per year, three-quarters of which falls in the May to November rainy season. In relation to this investigation, 1960 was unusually wet with rainfall of about 75 inches, 25 inches falling in September when hurricane Donna and tropical storm Florence passed nearby. The year of 1961 was unusually dry; the rainfall amounted to about 42 inches. PREVIOUS INVESTIGATIONS Some phases of the present study coincided with, and were contemporaneous with, a study of the general hydrology of the Snake Creek Canal area by Leach and Sherwood (1963). Where data and graphs were applicable to the present report, they were used freely without acknowledgment. Reports by Parker (1951), Parker, Ferguson, and Love (1955), and Kohout (1961) give much background information on the influence of drainage canals on salt-water encroachment in the Miami area. Many other reports cover the hydrology of the area and may be found in the "List of Publications" of the Florida Geological Survey in Tallahassee. PERSONNEL AND ACKNOWLEDGMENTS Thanks are extended to the Dade County Office of Water Control, especially to John Wilhelm, for furnishing rainfall and gate-opening data. Robert L. Taylor of the C&SFFCD aided by making available the facilities and personnel of his organization during the special test period of March 1961.

PAGE 14

8 FLORIDA GEOLOGICAL SURVEY Electrical-conductivity measurements of water in the canal during the test period of March 1961 were made under the direction of R. N. Cherry of the U. S. Geological Survey, Ocala. Other Geological Survey personnel who contributed materially were B. F. Joyner of Ocala and C. B. Sherwood of Miami. GEOLOGY AND DRAINAGE The Biscayne aquifer of the Miami area is a highly permeable water-table aquifer consisting of solution-riddled limestone and calcareous sandstone with fairly frequent layers of unconsolidated sand. The generalized lithologic characteristics in wells drilled during the investigation are shown in figure 4. Except for the peat and muck deposits at the surface, the sediments were deposited during the Pleistocene glacial epoch. Accumulation of snow and ice during glacial advances in the northern latitudes removed water from the oceans and lowered sea leveL Alternately, melting of the glaciers returned water to the oceans and caused sea level to rise. The alternation of limestone and sandstone beds in figure 4 relates to these fluctuations of sea level-estimated by Cooke (1939, p. 34) to have ranged from about 270 feet above to 300 feet below present sea level. DISTANCE FROM CONTROL DAM.IN FEET 39G,; Q 0 7 10 C0 0_00C 9000 8000 7000 01 _LAND SURFACE--.40 '.0L, -". " ' S/ -BOTTOM OF CANAL7 L: -" _ D'STANCE FROM CONTROL DAM,IN FEET 5?03 53Co 4CC0 3000 2000 I000 0 -LAND SURFACE -n .r BOTTOM CF CAMNAI S AND TAN » -1 '= te -c ----ESTONE * ITE d I. . -120 Figure 4. Section along the south bank of the Snake Creek Canal showing generalized lithologic characteristics of the Biscayne aquifer.

PAGE 15

REPORT OF INVESTIGATIONS No. 24 9 The peat and muck deposits were formed by the accumulation of partially decayed vegetation in swamps after the last glacial recession, about 8,000 years ago. The Snake Creek Canal traverses a natural overflow channel from the Everglades. These numerous overflow channels are wide, topographically low areas (4 to 8 feet above sea. level) that cut through relatively high land (8 to 15 feet) near the coast; they are called transverse or finger glades because of the manner in which they cut through the coastal ridge. Prior to construction of drainage canals (fig. 1), the water level in the Everglades was high enough to cause discharge through the finger glades by poorly channelized sheet flow. In the deeper pockets, aquatic plants grew lushly and, after dying, settled to the bottoms of sloughs. In these locations, the organic material was protected from oxidation and accumulated to form the peat and black muck deposits. In the higher parts of the finger glades, where water levels were below ground surface during the dry season, rapid oxidation prevented heavy accumulation of organic material. .In these areas, the remnant plant materials were mixed with particles of limestone and sand, eroded from the adjacent higher limestone hills, to form a deposit called marl (fig. 4). Since construction of drainage canals, the finger glades are flooded only during extremely heavy rainfall, and the rich organic soils are cultivatedto provide table vegetables for northern markets in the winter season. With continuing improvements in the canal system and improved efficiency of water control, the agricultural activities are gradually giving way to urbanization. METHODS OF INVESTIGATION The investigation followed two main courses: (1) Long-term observation of the position and rate of movement of salt water under the past method of operating the control dam (S-29). The dam has been operated on an arbitrary water-level basis with no consideration being given to the movement or the position of salt water; (2) detailed inventory of the position and movement of salt water under special short-term test conditions. The short-term test served as an index of typical conditions for analysis of the long-term data. These phases required continuous inventory of changes, both in the canal and in the aquifer.

PAGE 16

10 FLORIDA GEOLOGICAL SURVEY The long-term study of salt-water occurrence in the canal was made primarily from fluctuations of chloride content as determined by three electrical-conductivity recorders located at S-29, at West Dixie Highway, and at the footbridge (fig. 3). The electrodes of these recorders were positioned about 1.5 feet above the bottom of the canal, and the landward movement of salt water was sensed by a sharp rise in the electrical conductance of the water. During and after the short-term test of March 27-31, 1961, the position of salt water in the canal was determined by traversing the canalwater vertically with the cell of a portable conductivity instrument. The measurements were made from a boat that moved to specific stations along the canal. The long-term study of salt-water occurrence in the aquifer was made from fluctuations of chloride content in samples of ground water from nine stations of wells drilled along the south bank of the canal (fig. 3). The general scheme was to drill individual wells to different depths at distances of about 5 and 60 feet from the canal. Thus, a typical station might consist of wells 10, 20, 50, and 100 feet deep near the canal and wells 10 and 20 feet deep at a distance of 60 feet from the canal. Although drilling operations did not always fulfill the preferred condition of paired wells terminating at the same altitude, the rectangular coordinate positioning of the wells relative to the canal was virtually fulfilled. Each well was cased so that the water which it yielded was representative of that particular point in the aquifer. Details of the position and depth of each well terminus relative to the canal are shown in the map and section views of figures 16 to 20. An attempt has been made to simplify the illustrations as much as possible, but as this three-dimensional problem can be shown only in two dimensions, much must be left to the imagination of the reader to keep himself oriented in three-dimensional space. Herein lies the understanding of the physical parameters being discussed. LONG-TERM POSITION OF SALT WATER IN THE CANAL The long-term position of salt water in the canal can be inferred by comparing figures 5 and 6. Figure 5 shows the peak daily chloride content, as indicated by the conductivity recorders at S-29, at West Dixie Highway, and at the footbridge. Also shown are the water level in the canal at West Dixie, the gate opening at S-29, and rainfall at the Opalocka gage. The maximum vertical opening

PAGE 17

REPORT OF INVESTIGATIONS NO. 24 11 1960 1961 SSEP OCT NOV E N. APR MA NE EC a 0000 E PaA AIUA" S ] DEC, )E I I 19oo -i i 9000 3100 --go -------------O 1960, I the. h v rai.. acompan in....*,---, -urr -.. I c-, D-| ..-on n. ,atropical to rm Florence produced wide Ipread | ong | fg 5). | during.! the ero tem r 1 t 14 1 theGE icarger sal water ( lide ten ra t S29, l fig.° 5) *L I I -I ..I .. ..I . OP,-LOCKA GAGE Iu r mo ar t e. a t Wi l h Figure 5. Graphs comparing rainfall, gate opening, and fluctuations of water level and chloride content along the Snake Creek Canal, 1960-61. of each of the four gates is 15 feet (fig. 2); therefore, if all gates are fully open, the total gate opening is 60 feet. As the gates of S-29 are of equal width, the amount of gate opening in figure 5 represents the discharge area. All gates were fully open from early September to late October 1960, after the heavy rains accompanying hurricane Donna and tropical storm Florence produced widespread flooding (fig. 5). Only during the period September 10 to 14, 1960, was the discharge from the canal great enough to keep the canal completely flushed of salt water (chloride-content graph at S-29, fig. 5). At all other times full opening of the dam permitted salt water to move upstream on a rising tide. During the special short-term test of March 27-31, 1961 (fig. 5), salt water invaded the canal inland as far as the footbridge. The time scale for this period is expanded in figure 6 to show the detailed relations of intrusion, water-level fluctuation, and discharge. Salt water intrudes on the rising tide, first appearing at S-29, then at West Dixie Highway 2 to 3fhours

PAGE 18

12 FLORIDA GEOLOGICAL SURVEY AAM , y MARCHI MARCN 1 MARC"30 MARCHi 3 APRIL I j«__._4.__^--s|--Jub-I---.I4.-----~ol : I I I i III S-Ti l. FE a I ....L I _ o 3 i D1 LM vils f t later, and fally at the footbridge during the subsequent falling ,, , , , ....... Srelations are involved in this time lag and they will be discussed te o nt o sat atr son n f re , t on-trm poston of salt water in the canal can be inferred from figure 5 by comrIa/I!INII I-.I-I % I 5 /AXII I , 2 'I , E-roo NII \ Figure 6. Graphs comparing fluctuations of discharge, water level, and chloride content in Snake Creek Canal during the special test of March 27-31, 1961. later, and finally at the footbridge during the subsequent falling tide, about 8 hours after first appearing at S-29. Complicated relations are involved in this time lag and they will be discussed later, but it should be noted that the length of time of salt-water stay decreases landward and that the maximum concentration also decreases landward. For example, salt water did not reach the footbridge at all on March 29, just before noon (fig. 6). In light of the movement of salt water shown in figure 6, the long-term position of salt water in the canal can be inferred from figure 5 by comparing the relative fluctuations of conductivity at the three recording stations. IN THE AQUIFER The interface between fresh water and salt water in the Biscayne aquifer is not sharp; a zone of diffusion is invariably present. The chloride content in the zone of diffusion ranges from about 16 ppm (parts per million), the concentration of fresh water in Miami, to about 19,000 ppm, the concentration of sea water. The concentration of chloride ion is commonly used to represent the

PAGE 19

REPORT OF INVESTIGATIONS NO. 24 13 distribution of salt water, and isochlor lines are drawn through points of equal chloride content as determined by analyses of water from wells. Cross sections of the zone of diffusion along the south bank of the Snake Creek Canal are shown in figure 7. The black dots represent the termini of individual fully cased wells located about 5 feet from the canal; the open circles represent the termini of wells located about 60 feet from the canal. In intercanal areas the the isochlor lines turn sharply downward at the inland extremity of the salt water, and the zone of diffusion takes the shape of a blunt-nosed wedge of salty water (Kohout, 1961, fig. 3, 4). Along the Snake Creek Canal (and presumably other canals) the blunted shape, although still present, tends to be modified by large changes in gradient and head near the canal. When the dam is open, the lowest level of the water table coincides with the water level in the canal, and ground water flows laterally toward the canal from either side. Also, at such times, ground water beneath the canal tends to flow upward through the canal bottom as it follows the part of least resistance to the ocean. When the control gates are closed, the water level in the canal becomes the highest level of the water table because water flows quickly downstream from inland points of higher head. Under this condition the exchange of water between the canal and the aquifer is reversed, and water, whether it be fresh or salty, moves laterally and vertically from the canal to the aquifer. In figure 7, the general processes of salt-water encroachment in the Biscayne aquifer are recognizable: (1) The deep salt water, below and seaward of the 1,000-ppm isochlor, moves landward from the ocean in an essentially horizontal direction. This type of encroachment is relatively slow and insidious. It results from heads that are too low to prevent intrusion, and can be stabilized or corrected only by raising the heads. (2) The lenses of salt water that appear near the bottom of the canal and then drift progressively downward with time originate as a result of inland movement of salt water directly in the canal. This type of encroachment occurs very rapidly and is the consequence of inadequate knowledge of the position of salt water in the canal during control-dam operations.

PAGE 20

i DISTANCE FROM CONTROL DAM,IN FEET LAND *URfAC -. 14rtTum -li'D . --------------,-------, -w--a..EXPLANATION O, , Ac---" *04 .n-500-, y f X " ;.-.0 ......u 0 -0 ISOCHLOR LINE IN PPM DECEMBER 8, 1960 / * S WELLS 5 FEET FROM CANAL WELLS 60 FEET FROM CANAL
PAGE 21

LAND SURAC---. S-BOTTOM OF CANAL A' L '' 40 ..-40 -80 APRIL 3, 1961 .411M: /o S-.... ----tM -.F ' -o M 1AY 16, 1961 '80 1,1 2 4 SLAND SURFACE-& '*S BOTTOM OF CANAL -' U-3 Z -80 AUGUST 22, 19612 20 .) ..........--.:pj/ -W TI -120 5a 1;oe -120 Figure 7. Sections along the south bank of the Snake Creek Canal showing the zone of diffusion in the Biscayne aquifer, 1960-62. Qn

PAGE 22

16 FLORIDA GEOLOGICAL SURVEY Although the isochlor pattern in figure 7 appears to remain practically stable from December 1960 to January 1962, careful comparison of data shows that significant changes occurred at certain wells. The chloride content in the deepest well at station 4 decreased from 11,850 to 535 ppm, indicating seaward expulsion of salt water from the aquifer. During the same time, the chloride content in the deepest well at station 6 decreased from 1,220 to 490 ppm. The reason for the expulsion of salt water may be understood by following the events shown in figure 5. From early September to late October 1960 the control dam was fully open, because of flood conditions. During this time the heads in the aquifer in the vicinity of the canal were below normal and salt water moved horizontally landward and mounded beneath the canal. The chloride content in well G 420A (the only long-term data covering the period, fig. 18) increased from 6,800 ppm in June 1960 to 9,200 ppm on October 19, 1960. The year 1960, then, was essentially a year of encroachment. The year 1961 was abnormally dry, and even during the normally rainy months of September and October the dam remained practically closed. The infrequent opening of the control dam caused heads to remain sufficiently high so that expulsion of salt water from the aquifer occurred throughout 1961 (fig. 7). The fluctuation graphs in figures 16 to 20 confirm the seaward movement of salt water; chloride contents in most of the wells decreased generally during 1961. It is indeed a contradictory situation that the greatest inland movement of salt water took place in a wet year and the greatest expulsion of salt water took place during a dry year. Nevertheless, this fact indicates that the means are available for very effective control of the movement of salt water in the Biscayne aquifer. SHORT-TERM POSITION OF SALT WATER While it is generally not feasible to observe continuously the complete distribution of salt water, short-term tests, keyed to recording instruments, provide much information with which to interpret the long-term data. A short-term test was performed March 27-31, 1961. The control gates were completely closed for 6 weeks prior to the test, and conditions were almost stable despite rainfall of 2 inches that occurred 3 days before the test (fig. 5).

PAGE 23

REPORT OF INVESTIGATIONS NO. 24 17 The sequence of control-dam operations is shown in the water level graph of figure 6. All four gates were opened fully at 10:00 a.m., March 27, just after high tide. The canal was in natural tidal oscillation until 9:15 a.m., March 30, when S-29 was completely closed just before high tide. The intention was to trap salt water in the canal, as might happen during normal control operations, and then trace its movement both in the canal and in the aquifer. The trapped salt water moved inland in the canal much more rapidly than expected, and at 1:30 p.m. on March 31, at high tide, two gates were opened for a flushing operation which continued until low tide, at about 5:45 p.m. The dam was then closed and remained so until May 26, 1961. SALT-WATER MOVEMENT IN THE CANAL The tidal movement of salt water in a canal is rapid and transient. In figures 8-9 and 12-15 this movement is depicted by a time-lapse sequence of canal sections showing the approximate position of isochlor lines along the longitudinal axis of the canal. NATURAL TIDAL OSCILLATION The movements of salt water under unimpeded tidal oscillation on March 29 are shown in figures 8A and 8B; figure 8A shows the inland extension on the rising tide, and figure 8B shows the seaward movement on the falling tide. The stage of the tide cycle represented by each isochlor section is shown by an arrow on the water-level fluctuation graph (inset). The measured velocity profile for the middle vertical in the canal at West Dixie Highway is shown at the lower right side of each section. The discharge is divided into landward and seaward components as shown by the figures near the darkened arrows. Comparison of sections 1, 2, and 3 (fig. 8A) indicates that the inland advance of salt water was slow during the early part of the rising tide. Significant landward advance of salt water did not take place until all flow became negative (landward) in section 3. This indicates that seaward-flowing water in the upper part of the canal removes the landward-flowing salt water in the bottom of the canal by a combined dispersion and interface erosion process. Maximum inland velocity occurred at 7:37 a.m., about 11/ hours before high tide, section 4; the unshaded discharge arrows and dashed velocity curve show this condition. The solid velocity curve at 8:44 a.m., which corresponds in time with the isochlor

PAGE 24

A4rrUp f,I /NFrfffRIERfRFOTOMOL ATruo,rPlTNrRflr rouSt ijriruofNff erNrRtFTnOIr ros sI ArlvTUDpfI IFTRFfRff roFs0 t n -00 ;i rl n R |OOK fITi §I *.t.8, |, o,, , I .I F-STATION 9 a i"Ro " "i . = -.Or D A S .1 8 1 -SA/ ,Ol I -CrIONTROL A -U.S. HWY.

PAGE 25

ALTITUOE,INFE£TREFERRED TOMSL ALTITUD£E,INFETREFERREDrO TOSL ALTIUDEINFEETREPERRED TOMSL ALTIrUDE,INPETREFERRED TO.MSL -1 100 a o I , T S. .5( 0 wp1 T 1S, -a .. ..o.. O Pl ' ' ,, -C ._ ,^f; .-,>,^^; X , Si -Foo .-r "1 0 M-$ N" N I 'l o tT -' ., •,. .I I -Srz m e .,i I C a U1'c) -N I .' ., Z '' |2 'T .t .....H -" 'ol t izti .... F 1 ir Q & -US H. I_ _ IN I iS I _T._-8 Of _ A --S U-O 6 Sr -$ rL ON 4 ---------0CONNrOL DAM

PAGE 26

20 FLORIDA GEOLOGICAL SURVEY section, indicates that relatively fresh water near the surface began to flow seaward at least 1 hour before high tide. The highsalinity water at the inland toe of the wedge, however, continued to move landward until well after high tide. This time lag in landward movement of salt water was shown previously in the chloride-fluctuation graph of figure 6, where salt water did not appear at the footbridge until about 2 hours after high tide. These out-of-phase flow characteristics in the upper and lower parts of the canal result from the density contrast between fresh water and salt water. At, and for some time after, high tide the fresh-water head at inland points is too small to counterbalance landward movement of salt water and the inland toe of the wedge moves landward as a density current. As the tide begins to fall more rapidly after high tide, the landward hydraulic gradient associated with the density current is overcome by seaward hydraulic gradient from falling ocean level. The salt water at the bottom of the canal, then, reverses direction and flows seaward. The salinity distribution in the canal reflects the out-of-phase flow pattern. The upper part of the canal is rapidly freshened by large positive (seaward) velocities in the fresh water as the tide falls (fig. 8B), section 6. However, even after the reversal of landward movement of salt water, seaward velocities in the salt water near the bottom of the canal are small compared to those of the overlying fresh water (velocity curves, sections 7, 8). This velocity differential causes a lens or slug of salt water to lag behind in that part of the canal westward from West Dixie Highway. Apparently, downward-directed velocity components (possibly related to turbulence near bridge pilings) develop in the fresh water near the outlet. Salt water in this part of the canal is eroded and flushed seaward before all the salt water that moved landward during the rising tide can retrace its path to the outlet. The phenomena occurred repeatedly when the control dam was fully open. This indicates that the control should not be operated exclusively on the basis of water-level measurements, for slugs of salt water could be inadvertently trapped. The position of salt water should be determined directly by conductivity measurements at the bottom of the canal. TRAPPING SALT WATER This section will show the dangers of inadvertently trapping salt water in a canal by shutting the gate on it. On March 30 at 9:15 a.m., just before high tide, all four gates of S-29 were closed.

PAGE 27

REPORT OF INVESTIGATIONS NO. 24 21 The salinity distribution just prior to the time of closing is shown in figure 9, section 9. A stable density stratification exists when fresh water overlies salt water with the isochlor lines horizontal. Sloping isochlor lines in stagnant water may be generally interpreted as an indication of hydraulic instability related to a nonstable density stratification. After the gates were closed, the water (for purposes of this discussion) may be considered to have been virtually stagnant, though 0 R DISTANCE FROM CONTROL DAM,IN FEET 14,000 2000 12000 80000 6000 4000 2000 0 SO --" ---------------" ----------------I--1 1 ' :' % (9) MARCH 30,1961 (8:50A.M.) Q WATER SURFACF__ -ISOCHLORSIN PARTS PER MILLION t.. -MAR.27 MAR.28 MAR.9 MAR.O3 MAR.31 APR.I CONNTR 6A SP 11; , G Z A 1G 1% J t 6,P 12 GA 12 GP 12 GA f2 SP OPEN -2 i -4 ; --|' (10) MARCH 30.1961(11:30A.M.) VATER SURFACE S ISOCHLORS.IN PARTS PER MILLION -l movemn.ti "o water ft er, 'f *T 'A" Ai of the 'E2 0y ,t hghti Mc > "11 "' P. IS OCHLORSIN PARTS lPER ILLION Ik AI1R CAf A AL 1 PR. -c " ' § , --, .I ' 'I '. ' ' -" ' , 1s i i........ ...... ..L R.EEKLA.L . Figure 9. Sections 9, 10, and 11 through the Snake Creek Canal showing movement of salt water after closing of the control dam at high tide, .1 March 30, 1961. March 30, 1961.

PAGE 28

22 FLORIDA GEOLOGICAL SURVEY some net inflow of fresh water entered the canal as the stage rose; the sloping isochlor lines in section 9, then, indicate instability. As expected, a density current developed when the gates were closed and salt water in the seaward reach of the canal moved landward. To maintain continuity between the individual water masses, fresh water from inland points moved seaward through the upper part of the canal to replace the salt water which moved landward through the lower part. These movements were traced primarily by making vertical conductivity traverses of the canal water from the boat. Also, vertical conductivity traverses were made continuously at the footbridge and a series of these profiles is shown at the lower right side of section 10. The leading nose of the salt-water wedge passed the footbridge at 10:44 a.m., it was about 5 feet thick and had a maximum chloride content of about 1,500 ppm. By tracking the intruding nose of the salt-water wedge for a period of time it was determined that the rate of landward movement was about 15 feet per minute, or about 900 feet per hour. The serene water surface completely belied the rapid encroachment taking place at the bottom of the canal. Figure 10. Photograph of rock pit (at left) looking in the downstream direction of the Snake Creek Canal (at right).

PAGE 29

REPORT OF INVESTIGATIONS No. 24 23 The salt water moved nearly 1 mile in the 41/4 hours between 9:15 a.m., the time of closing the gates, and 1:30 p.m. Careful inspection of the discharge graph at Miami Gardens Drive (fig. 6) shows a gentle increase in discharge at about 1:30 p.m. This increase was sensed by a small velocity vane positioned in the fresh water above the intruding salt-water wedge, and reflects the change in velocity in the seaward flow section as the wedge passed Miami Gardens Drive. At about 2:00 p.m. salt water began to flow into the rock pit, in a manner similar to water passing over a waterfall (fig. 3, 9, 10, 11). By late afternoon the hydrologic significance of the rock pit became clear. The rock pit was a hydraulic sink in which the salt water would temporarily collect. The salt water would resume its upstream movement only after it filled the rock pit above the level of the bottom of the canal. Thorough investigation of the rock pit began the following morning, March 31. As shown by figure 12, section 12, a considerable quantity of salt water had flowed into the pit. The chloride content in the rock pit was low compared to that of water in the canal because, as the salt water flowed downward into the EXPLANATION -SCALE IN FEET CONTOUR ON LAKE BOTTOM 300 0 300 6 IN FEET BELOW MSL "S N'--' I / \ /I / --·\I -15 / ,\/i " \\ \ \E // s J---^ r N SALT-WATER MOVEMENT, a' lM pARIf MARCH 30-31,1961 Figure 11. Rock pit showing approximate bottom topography.

PAGE 30

24 FLORIDA GEOLOGICAL SURVEY lake, heavy mixing and dispersion of salt reduced the maximum concentration to less than 2,000 ppm chloride, though water with a concentration of 5,000 ppm was observed flowing into the pit. Soundings, made in conjunction with numerous samplings and conductivity traverses during subsequent months, were used to develop a rough contour map of the bottom of the rock pit (fig. 11). Two deep holes nearly 40 feet below sea level are separated by a ridge having an altitude of about 25 feet below msl. In these holes most of the salt water samples were collected at depths greater than 20 feet below msl. The profile of the pit bottom in the canal sections (fig. 8, 9, 12-15) is generalized and represents no particular section line. FLUSHING OF SALT WATER FROM THE CANAL Once having entered deep holes such as those in the rock pit, the heavy salt water cannot easily be retrieved and flushed. As most of the salt water remaining in the canal would eventually flow into the rock pit, the canal was flushed on March 31, the day after closing of the control gates. At high tide, at 1:30 p.m., two gates of S-29 were opened fully. The distribution of salt water before flushing is shown in figure 12, section 12, and the progress of flushing is shown in sections 13 and 14. The principle of flushing is based on the tidal fluctuation of the ocean. In Miami, the range of fluctuation from high tide to low tide amounts to about 2 feet. Assuming that the water level upstream from the control dam is at least as high as the level of the ocean at high tide, the dam can be opened at high tide or shortly thereafter. The falling tide level continually creates a head differential between inland parts of the canal and the ocean; thus, for a period of about 6 hours, from high tide to low tide, a steep seaward hydraulic gradient produces a high rate of discharge from the canal. The discharge reaches a maximum slightly before low tide (fig. 6). Referring to the velocity curve in figure 8B, section 7, the seaward velocity of the fresh water just below low tide can be estimated at about 2 fps (feet per second) while that of the salt water can be estimated at about 0.7 fps. At this velocity, the salt water will move seaward at a rate of at least half a mile per hour. During the 6 hours of falling tide, flushing of 2 to 3 miles of the downstream reach of a canal is feasible. Turbulence disperses large quantities of salt water upward into the high-velocity fresh water and the flushing is probably more effective than the calculation indicates.

PAGE 31

REPORT OF INVESTIGATIONS No. 24 25 .A .q ·t 0 II i S I I DISTANCE FROM CONTROL DAM;IN FE'ET , 114 000 1 _ -0 ._ 4000 . WATER SI&FAE-,_ i ISOCHLORS.LN PARTS PER MILLIO. -I MA.e? M .2 MA 2 MAR.0 MAR.I3 APR.I ,R (13) MARCH 31,1961(240P.M.) 5 WATER SURFACE' 0 A ISOCHLORS, I PARTS PER MILLIO -MAR.7 MAR. MAR.29 MAR.30 MAR.31 APR.I CONHO IQ GA5 GP I 2 4P A It 6P 12 GA 12 6P2 GA 2 6P 12 GA b 2 SP 2 6A 12 GP OPEN -z , .,[ .. " 1A(A1 " ,-. " -,G--5 AR-c ..s _-I J-I I -r 1 .. ME .0 1---A .I -C -I Fig_ 1 ... .ec,"I -1 1 an .15 -m (14) MARCH 31,1961 (49:OA.M.) K WATER SURFACE-*| S.1 ISOCHLORS,IN PARTS PER MILLION * 5, -t-. _I5 I, MAR.? MAR28 MAR9 MAR.30 MAR.31 APR. I CONTROL F6A6 AP 12. GecRtP 14, adI P oipn moe WE of N ted .a .. , I -4 0" --.Ve (15) APRIL I, 1961 (9:00A.M.) uing ty ISOCHLORS,IN PARTS PER MILLION K MAR.27 MAR.28 MARR31 APR.I CCARTRIN. IESS 5. 1 ; I Figure 12. Sections 12, 13, 14, and 15 through the Snake Creek Canal showing movement of salt water during and after flushing, March 31 to April 1, 1961.

PAGE 32

26 FLORIDA GEOLOGICAL SURVEY During the dry spring of 1962, high easterly winds on March 8 and 9 coincided with high average tides and the wind-driven waters of Biscayne Bay overtopped many control dams in the Miami area at high tide. Salt water invaded the canals at a time when dams throughout the area were held completely closed to maintain heads and to conserve fresh water. In spite of the entrapment of salt water in the canals there was reluctance to flush the canals, as it was felt that the consequent loss of fresh water could not be tolerated. The series of isochlor sections for unimpeded tidal oscillation (fig. 8A, 8B) show that it is not necessary to have a high freshwater head behind the dam in order to perform a flushing operation. The principle of flushing depends almost entirely upon the falling stage in the ocean itself. The primary requisites are: (1) the canal should be long enough to tap a reservoir of fresh water at inland points, and (2) the control dam should be of a type that permits rapid operation so that opening and closing can be accomplished readily at opportune moments. Flushing of salt water from a canal entails some loss of fresh water. However, wasting a small portion of the vast reservoir of fresh ground water tapped by a canal is preferable to allowing a small volume of salt water to move inland considerable distances where it may contaminate much larger volumes of fresh water. ULTIMATE LOSS AND FLUSHING OF THE REMAINING TRAPPED SALT WATER In the 41/-hour flushing operation of March 31 not all salt water was removed. As shown by figure 12, section 15, a lens of salt water remained in the downstream reach of the Snake Creek Canal. Also, most of the relatively heavy salt water in the rock pit remained. These separated masses of salty water were traced until July 1961 (fig. 13-15). The downstream lens in the canal tended to spread toward the rock pit, but a large part of the salt water in the lens disappeared between April 1 and 3 (compare sec. 15, 16). The disappearance coincided with an increase of chloride content in observation wells adjacent to the canal and will be considered in a later section. The salt water in the rock pit appeared to remain practically stable from April 1 to 18 (sec. 15-19). The primary change during this period was a dispersion-caused widening of the contact between the 1,000and 200-ppm isochlors. Between April 18 and May 3, the isochlor positions changed significantly in the rock pit and also in the canal (compare sec. 19,

PAGE 33

REPORT OF INVESTIGATIONS NO. 24 27 k o § o DISTANCE FROM CONTROL DAM,IN FEET 14.000 12, 0 00 1000 8000 6000 4000 2000 0 -4 10 *' I I-------~ ---1 -----------(16) APRIL 3, 1961 -3/0 C' 4WATER SURFACESOCHLORSN PARTS PER MIL ,u -CONTROL -2 _ CLOSED -40I 2 (17) APRIL 4,1961 WATER SURFACE-* SISOCHLORIN PARTS PER MILLION3 , 10therefore, this was probably the period when the distribution of chloride content in the rock pit changed. In the bar graph in figure 14, lower part of section 19, the wind movement exceeds 100 miles per day for 3 consecutive days, April 21-23; this period nearly coincides with the period of increase in chloride content at the footbridge. Apparently, the 3 days of strong wind mixedthe waterin the upper part of the roc it. TheLLION 1-000a-o changes in isochlor positions, April 3-11, 1961. 20). An increase in chloride content was sensed by the conductivity recorder at the footbridge between April 22 and 26 (fig. 5) and, therefore, this was probably the period when the distribution of chloride content in the rock pit changed. In the bar graph in figure 14, lower part of section 19, the wind movement exceeds 100 miles per day for 3 consecutive days, April 21-23; this period nearly coincides with the period of increase in chloride content at the footbridge. Apparently, the 3 days of strong wind mixed the water in the upper part of the rock pit. The

PAGE 34

28 FLORIDA GEOLOGICAL SURVEY Qt , t 1 .DISTANCE FROM CONTROL DAM,IN FEET .4,000 ,2000 10,000 ooo 6000o 4000 o200 0 APRIL 18,1961 --------WATER SURFACE-4. * ISOCHLORnsI PARTS PER MILLION CONTROL 0. HLEH GAGE CLOSED -40 Fa ,1DO l (20) maY 3.1961 --WATER SURFACE-. ' socIaLtoRslm PATS PER MILLuO ' _»f SCLOSED .. . Figure 14. Sections 19, 20, and 21 through the Snake Creek Canal showing changes in isochlor positions, April 18 to May 16, 1961. downward movement of isochlors in section 20 represents dispersion and upward movement of salt ions into the uppermost water of the rock pit. Small hydraulic gradients, associated with the wind and with ground-water discharge in the downstream reach of the canal, then carried the salty water downstream where it was sensed by the conductivity recorder at the footbridge. Wave height on a given lake depends on the length of time the wind blows as well as on the wind speed. Therefore, the 3 consecutive days of high wind, April 21-23, were far more effective in

PAGE 35

REPORT OF INVESTIGATIONS No. 24 29 stirring the waters of the rock pit than were the single days of similar wind speed in early April. The downward movement of isochlors between May 3 and 16 (sec. 20-21) undoubtedly relates to the relatively high sustained winds of May 6 to 9. It is significant that wind and waves serve as agents for removal of trapped salt water. On May 26, after heavy rainfall, the dam was opened (fig. 5). The resulting tidal oscillation and discharge produced small currents in the bottom of the rock pit. This agitation reduced the chloride content and density more in the eastern hole (fig. 11), by upward dispersion, and the salt was gradually flushed toward the partly open control dam. Although residual traces of the original contamination remained, the partial gate openings that started March 26 were effective in reducing salinity even at the very bottom of the rock pit (fig. 15). The decrease in chloride content at the footbridge (fig. 5) at the same time indicates that the tidal impulse of the ocean can be utilized .to remove salt water that has become trapped in deep pits or lakes along the course of a canal. SALT-WATER MOVEMENT IN THE AQUIFER The short-term movements of salt water in the aquifer are closely influenced by happenings in the canal. Fluctuations of chloride content in wells along the south bank of the canal during 1960 and 1961 are shown in figures 16-20. During the special test of March 27-31, water samples were collected as often as four times daily and these are plotted on an expanded time scale. NATURAL TIDAL OSCILLATION Under unimpeded tidal oscillation, water moves from the aquifer into the canal on a falling tide and from the canal into the aquifer on a rising tide. This to-and-fro motion prevailed when the gates were open March 27 to 30 and is indicated by the two-headed arrow in the March 30 section of the sequence of insets in figure 19. During this period the chloride content in the shallow wells fluctuated only slightly at stations 1 to 5 (fig. 16-18), and practically none at the more landward stations (fig. 19-20). Although highsalinity water oscillated landward beyond station 8 in the canal (fig. 8), salt water did not appear in wells located as little as 5 feet from the canal. This is not surprising. The net flow of fresh ground water in the upper part of the aquifer is toward the canal. Any small amount of salt water that penetrates the aquifer

PAGE 36

30 FLORIDA GEOLOGICAL SURVEY C.. I, ,. T, DISTANCE FROM CONTROL DAM,IN FEET -4000_,. 1 ,2.000 _ o0.000o oo8000 6000 4000 2000 0 (22) nAY 29.1961 0 WATER SURFACE-. ISaMLOSiN PAMTS PER MILLION -IC 'c (23) JUNE 13. 1p96 o WATER SURFACE-| ISOCILOUIS.IN PAR'S PER MILLION PARTLrLPEN Thiscases salt water, deep inthe aquifer, to move landward The (24) JULY 10.1961 __IC1__JUL? __0___6__WATER SURFACE-. ISO.IILOS. N PARTS PER MILLION -tl t_.CLOSED -: | Figure 15. Sections 22, 23, and 24 through the Snake Creek Canal showing changes in isochlor positions May 29 to July 10, 1961. on a rising tide is flushed back to the canal on the subsequent falling tide. Thus, the average flow of fresh ground water toward the canal opposes and greatly minimizes movement -of* salt water from the canal to the aquifer during unimpeded tidal oscillation. A different situation exists in the deep part of the aquifer. The average head in the aquifer is lowered when the gates are open. This causes salt water, deep in the aquifer, to move landward. The movement is slow but relentless. An example is the increase in chloride content from 6,800 to 9,200 ppm in well G 420A between

PAGE 37

620,0001 _ ..,-... , "__ _-GIS ______ GIll11 10.000 -S 1115 ,00 MAP VIEW SECTION VIEW Of DOAM 0 STATIONS 1,2,ond 3 ISTA116 IONI) A , 011 '0 (r j 3) 0 Qw •1120 1121'G 1118 :. ' a 0 IIIS 1 "20 : 10 " 0 1120 0 .112... OCT 1NOV st 5 > -0. AN. Fr M MAY JUNE JULY AU. se P 1 0 FOCt NOV. 1 -. .0II' , .... ..-... Figure 16. Graph showing fluctuations of chloride content at well stations : ... .... 11.0 ...G... .. -EXPANDED, ,G 1117 TIME SCALE Oct OCT vW. Dtk. JAN. FEB. MAR. APR. __ MAY JUNE JULY AUG. SEPT. OCT, NOV. DEC. Figure 16. Graph showing fluctuations of chloride content at well stations 1, 2, and 3, 1960-61. _: oo .. ,' --"..Y "'"-",,.. /-, ,=....... ....",, \._ '

PAGE 38

..oo II I ll 111 1 .I !I I I I I I I 0 \ MAP VIEW «TlON V 4Y 0 1125 / f, o 89\0 10 >ooB| SSTATION 4 V * * 0 11 1 -1125 ...... ...... S29 -g27 11 g -"-31120 " -'"~ -6«128 -| ---EXPANDED-.._ TIME SCALE I IM. ' ' ' ' I I Figure 17. Graph showing fluctuations of chloride content at well station 4, 1960-61. L

PAGE 39

7l1 ,09 -0---A-------1109a -STATION 5 ---. MAP ViEg SECTION VIW , _ \__ ___. _ ___. I o_,,,__ I 0*ow j " z I 0.,. , , -,,IR1/ a-g I -9 "'""% .5 _ Figure 18. Graph show Ina l stat A 5,G 1110 1010 J too G G I I12 MU I112 EXPANDED scALE TIME SCALE DEC. 1 CT. NOV. DEC. JAN. Ft&B. MA a e *APR' MAY I UNE ULY AUO. SEPT. OCt. L NV DEC. Ira19 5 1919 to a ag Figure 18, Graph showing fluctuations of chloride content at well station 5, 1959-61. ci3

PAGE 40

24000 1 -1 1rmr 100000-9.-. If 19I Vie0 10 9 ** -/G 1138 U1 oCC., **JAN* .e *s A *R .JUNE * Ja n one* a 0** b * c *OM d ' e 19056 Figure 19. Graph showing fluctuations of chloride content at well station 6, 1960-61. G 11736-4 G 1138 9D 1136 I 0 -M 114 IA-6 1140 *1139 G1139 EXPANDEDP G 113M ~ TIME SCALE~ STATION 6 OCT. I mm KEC. JAN. I FE&. _MA R.'. * *. I MAY I JUNE I JULY I AM. SEPT OCT. NOV. M £C, 16 1961 Figure 19. Graph showing fluctuations of chloride content at well station 6,1960-61.

PAGE 41

200 I I i 111 11 I I I I STATION 9 1156/-G 1155 100, -G I 53 .. -,I S1154 ./ -" -G1154 ~-. ,,s1154 A. P i ' ,. ( STATION 8 G 51152 S ITIME SCALE G o , ,E ,/ ,_ ,,,,I ,,,, ^ ij" ,II -I , , ,2 g S G1149w 0ss, as,... 0 ''**-'"| _ 00 7 G 11504 Iz G I s 1151"''' G 1149"-* -* G 1150 1 " S300 1 5 STATION 7 116 , 61148 "Z G D S1001 J --"e "^ ^ ae..^ ^ --1_<_G 1148 .....'.... ..... // 1 7" " ---G 1145 / G 1147:f MAP VIEw SECTION VIEW TIME SCALE all. , aiml NOV. :PEC. JAN. FEB. MAR. I AP. MAY JUIE JULY U. SEPT. OCt. 1 Hw. DEC. o1960 _ 1961 Figure 20. Graph showing fluctuations of chloride content at well stations 7, 8, and 9, 1960-61.

PAGE 42

36 FLORIDA GEOLOGICAL SURVEY June and October 1960, when the dam was open for hurricane Donna floods (fig. 18). RECHARGING THE AQUIFER WITH TRAPPED SALT WATER The canal is quickly converted from a drainage canal to a recharge canal by closing the control. The water which happens to be in the canal at this time, either fresh or salty, is driven from the canal under hydraulic gradient and appears in wells adjacent to the canal. All well stations (fig. 16-20) showed an increase of chloride content on the morning of March 31, the day after salt water was trapped. As will be noted by comparing the graphs, the degree of reaction to the recharge of salt water varied. A partial list of the factors involved is as follows: (1) The distance of the well from the canal; (2) the altitude of the terminus of the well in relation to the altitude of salt water in the canal; (3) the concentration of salt water, which varies both vertically and longitudinally along the canal; (4) the amount of dispersion of salt as the water passes through the rock; and (5) the local permeability of the aquifer near the well station. The relations of salt-water movement at station 6 are shown in the section views of figure 19. Section a for March 30 shows that salt water had not invaded the aquifer during unimpeded tidal oscillation. In section b (10:00 a.m., Mar. 31), approximately 24 hours after closing the control gates, the chloride content in well G 1138, 5 feet from the canal, had increased from 60 to 13,700 ppm while that in well G 1140, 65 feet from the canal, increased from 60 to 288 ppm. The sharp increase of chloride content in these wells, in contrast to the decrease in well G 1136 (about 22 feet deeper) from 125 to 102 ppm, indicates that the salt water moved nearly horizontally. This is a reasonable observation because the horizontal permeability is generally much greater than the vertical permeability in stratified aquifers. Section c (4:00 p.m., Mar. 31) shows the flow direction while the canal was being flushed. The sharp decrease of chloride content in wells G 1138 and G 1140 in the fluctuation graph at this time indicates that part of the salt water, that had previously invaded the aquifer, was flushed back toward the canal. After the flushing operation was completed and the dam was closed, the unflushed salt water in the aquifer resumed a flow direction away from the canal (sec. d).

PAGE 43

REPORT OF INVESTIGATIONS No. 24 37 In addition, salt water from the salt-water lens, which remained in the downstream reach of the canal after flushing (fig. 12, sec. 15, and fig. 13, sec. 16),moved into the aquifer. The disappearance of the salt-water lens from the canal by April 3 coincided with the sharp decrease in chloride content in well G 1138 on that date; the recharge water in the canal had become relatively fresh but the increase of chloride content in well G 1140 between April 1 and 3 showed movement of the salt-water slug farther from the canal. The low maximum chloride content in well G 1140 as compared to that of well G 1138 indicates dilution of the salt water during movement between the wells. It is interesting to note that the chloride-content curve for well G 1138 crosses over and becomes lower than that for well G 1140 after April 5 (fluctuation graph, fig. 19). As previously shown, water in the canal continued to freshen with time. Thus, as the slug of salt water passed farther into the aquifer it was followed by relatively fresh water which appeared first in well G 1138, near the canal bank, and then in well G 1140. The reversal of the chloride-content curves shows the cause-effect relation of salinity in the canal on salinity in the aquifer. The chloride content in well G 1137 reached a maximum on April 7, more than a week after the control was closed (fluctuation graph, fig. 19). During this week, some of the salt water that remained at the bottom of the canal was dispersed upward in the canal by wind agitation (fig. 13, sec. 16, 17). As the shallow wells were recharged by water from the upper part of the canal, the time lag of salinity fluctuation in the shallow wells reflects the changes of salinity in the upper part of the canal. The lower peak chloride content in well G 1139, as compared to well G 1137 again demonstrates the reduction of salinity by dispersion of salt as ground water moves through the pore spaces of the aquifer. Well G 1136 is 22 feet deeper than well G 1138 so that the vertical component of salt-water movement is identified by the relative fluctuations of chloride content in the two wells. The chloride content in well G 1136 did not reach its maximum until early May, about 1 month after the short-term test, whereas that of well G 1138 reached its maximum 1 day after the control was closed (fig. 19). This great difference in time results from several factors. An important driving force in the vertical direction is the difference in density between the overlying heavy salt water and the underlying light fresh water. In open water, an unstable density condition of this type is corrected rapidly by a massive discrete

PAGE 44

38 FLORIDA GEOLOGICAL SURVEY overturn of the two fluids. The settling of salt water in the rock pit is an example. However, in an aquifer the pores of the rock are small and the fluid movement is split into similarly small flow filaments. The fresh water rises at the same time that the salt water sinks. The overturn is reduced to slow infiltration wherein dispersion of salt plays an important part in modification of the original concentrations. The effect of trapping salt water in March 1961 is recognizable until October 1961 in the fluctuation curve for well G 1136 (fig. 19). In the longitudinal sections of figure 7, the downward movement of salt water is depicted as a lens of salt water that appears near the bottom of the canal and progressively drifts downward with time (sections of Mar. 31 to May 16, 1961). At the same time there is a horizontal component of flow perpendicular to and away from the canal; this movement is shown by the arrows in section e of figure 19. ULTIMATE LOSS AND FLUSHING OF SALT WATER IN THE AQUIFER Comparison of the fluctuations of chloride content at all stations (fig. 16-20) shows that the deep wells were little affected by the short-term test but that the shallow wells were affected in a characteristic pattern. This pattern is responsive to the operations of the control dam. The control gates were closed on March 31, and they remained closed until May 26 (fig. 5). Initially the chloride content in the shallow wells rose sharply and then tended to decrease slightly. At the downstream stations the decrease was somewhat larger because ground-water movement around the control dam and toward the ocean was more rapid there. At the more inland stations (fig. 20), the aquifer is recharged rapidly until canal water refills that part of the aquifer that is unwatered during drainage. After this, the gradient away from the canal becomes comparatively small as the ground-water flow system approaches a steady state consistent with the boundaries of the aquifer. As the rate of recharge is proportional to the gradient, salt water at first moves rapidly into the aquifer, and subsequently at a much slower pace. The salt water, therefore, remains within reasonable proximity of the canal and is subject to retrieval. The chloride content in most of the shallow wells decreased sharply after the control dam was opened at the end of May. The

PAGE 45

REPORT OF INVESTIGATIONS NO. 24 39 fluctuations at station 7 (fig. 20) are especially interesting. The chloride content in wells G 1145 and G 1146 near the canal decreased sharply, whereas that in G 1147 and G 1148, 60 feet from the canal, remained unchanged. Figure 5 shows that the total gate opening was decreased from 30 feet May 26 to 31 to about 5 feet thereafter. At the 30-foot gate opening, the canal caused drainage and some of the salt water flowed out of the aquifer into the canal. Water levels recovered somewhat after closing to the 5-foot opening (fig. 5), and fresh water from the canal recharged the aquifer. The movement of fresh water into the aquifer was sufficient to freshen the water in wells close to the canal (G 1145-G 1146), although water in the wells farther from the canal (G 1147-G 1148) remained relatively salty. Clearly, part of the salt water trapped in March had been removed but a part still remained as an isolated mass. During the next gate openings on August 20 .the remaining salt water was completely flushed from the aquifer and the chloride contents of all four wells became nearly the same. Thus, the low-salinity water trapped in March was recovered from the shallow part of the aquifer by August 1961. The recovery was possible because the density of this lowsalinity water (less than 200 ppm) was not great enough to cause a large downward movement. The dense, high-salinity water that percolated downward at station 6 (shown by the graph of well G 1136, fig. 19) obviously was not recovered. It continued to move downward until the density contrast was eliminated by confluence with water of the same density in the deep part of the aquifer. MANIPULATION OF THE CONTROL DAM FOR MAXIMUM BENEFIT The primary function of a canal in the Miami area is to remove surplus floodwaters. As floods in Miami result from heavy rainfall, it is obvious that the floodwaters must always be fresh. If salt water penetrates landward in a canal, the drainage function of the canal is impaired because floodwaters cannot flow seaward through that part of the canal in which salt water flows landward; therefore, the optimum removal of floodwaters occurs when no salt water invades the canal. Whether the flow condition is steady state or transient due to tide, it should be recognized that intrusion of salt water in a canal is an obvious sign of reduced efficiency in the discharge of fresh floodwaters. All four gates of S-29 were fully open for nearly 2 months after hurricane Donna passed on September 10, 1960 (fig. 3). Only

PAGE 46

40 FLORIDA GEOLOGICAL SURVEY during September 10 to 14 was the discharge great enough to keep the canal flushed of salt water. At all other times during the period of record shown in figure 5, full openings of the gates permitted salt water to move upstream on a rising tide. This suggests that the control was operated at full-open position much longer than necessary. FACTORS AFFECTING INLAND MOVEMENT OF SALT WATER The manner in which the total gate opening is manipulated has a bearing on the landward extension of salt water in the canal. For example, a gate opening of 30 feet may be represented by a 7.5-foot opening for each of the 4 gates or a 15-foot opening for each of only two gates (the other 2 remaining closed).' The schematic flow diagrams of figure 21 illustrate the hydraulic advantage that attaches to uniform distribution of the gate opening. In figure 21a fresh water is forced to flow downward around the obstruction of the gate. This downward displacement of the fresh-water discharge section impedes the upstream intrusion of salt water. Uniform opening and submersion of the four sluice gates (in contrast to fully open gates) can be used as a device to distort the flow pattern deliberately in order to prevent saltwater intrusion. Questions arise as to whether the movement of salt water into a canal is ever justified on the basis that widest opening of the control gates will cause greatest discharge of fresh floodwater. The critical point of entrance of salt water into a canal is at the control dam, for here is the only place where intrusion can be controlled. The problem involves the total head at the dam, the amount of discharge at various gate openings under these various heads, and the influence of fresh-water head in preventing upstream movement of salt water beyond the control dam. Using the principle of static equilibrium for fluids of different density, salt water will not move upstream as long as the head differential between the fresh-water and salt-water sides of the control gate is greater than 0.3 foot (for a canal whose bottom is 12 feet below sea level). This value is obtained by balancing the 'The values used here only demonstrate a principle; the discharge area for 2 gates opened 15 feet is not equal to the discharge area for 4 gates opened 7.5 feet because the gates at wide-open position may be above the actual water surface by as much as 3 feet.

PAGE 47

REPORT OF INVESTIGATIONS NO. 24 41 pressure exerted by fresh water against the opposing pressure exerted by sea water at the bottom of the canal. Pf = p g lf P, p gls Pt g 1 = Ps g i letting if = h + 1, (See fig. 21a): pf (h + 1I) = ps, h = 1, (p --pf) Pf h = 12 (1.025 -1.000) = 0.3 ft. 1.000 Pf = pressure exerted by fresh water Ps = pressure exerted by sea water g = acceleration of gravity pf = 1.000, density of fresh water pS = 1.025, density of sea water 1f = length of fresh-water column, in feet 1i = 12 feet, length of sea-water column h = head differential between fresh-water and sea-water sides of control dam, in feet. Referring to the schematic flow diagram of figure 21b, when the gates are fully open, the head loss through the control structure itself is very small. The flood conditions would have to be very severe in order to maintain 0.3 foot of fresh-water head at the control dam near high tide, when the greatest possibility of inland movement of salt water exists. The reason for upstream penetration of salt water on September 15, 1960 (fig. 5), only 5 days after the passage of hurricane Donna, is evident. With fully open gates, the head differential and discharge decrease very rapidly until salt water is able to flow upstream during the rising tide. Submergence of the gates, however, tends to focus all the head loss at the control (fig. 21a). This is very desirable and if the four gates of S-29 are uniformly adjusted to maintain a head differential of 0.3 foot during the rising tide, no salt water will invade the canal. DISCHARGE RELATED TO GATE OPENING The quantity of water discharged under a sluice gate depends on the head differential across the gate and on the cross-sectional area through which the discharge occurs. Figure 22 shows the approximate discharge related to head differential for various openings of one gate of the control dam in the Snake Creek Canal. The plotted points represent the mean daily discharge at West Dixie Highway and the head differential,

PAGE 48

42 FLORIDA GEOLOGICAL SURVEY (0) 0U WATER SURFACE ft st SRESH WATER SALT WATER (b) WATER SURFACE FRESH WATER INTERFACE C SALT WA TER Figure 21. Schematic flow diagrams: a, with gate partially submerged; b, with gate fully open. as obtained from the difference in mean daily gage height at West Dixie Highway and Biscayne Bay at NE 84th Street. Comparison of past records indicated that the mean daily gage heights of Biscayne Bay at NE 84th Street and the now discontinued gaging station just below the control dam in Snake Creek Canal were comparable when the control dam was closed or the discharge was within the range shown in figure 22. The curves were calculated by using the formula Q = ACV2 gh where Q is the discharge, A is the cross-sectional area, C is. the coefficient of friction, g is the acceleration of gravity, and h is the

PAGE 49

REPORT OF INVESTIGATIONS No. 24 43 DISCHARGE, IN CFS 0 200 40 600 0 800 1000 1200 1400 30 o. H 0 / Z 0 z 0 o0 0 -20 0 \ --/ ILJ // ~-H EAD DIFFERENTIAL THAT WILL PREVENT .I UPSTREAM MOVEMENT OF SALT WATER SAT BOTTOM OF CONTROL DAM Figure 22. Graph showing relations of head differential and discharge for various openings of one gate of the control dam in the Snake Creek Canal. head differential through the gate. As Q, A, and H were known for each of the points in figure 22, the coefficient of friction (C) was evaluated for the different gate openings. Then, using the average value of C and holding the area (A) constant for the individual gate openings, the curves were constructed through the mean value of the plotted points by solving the formula for Q while letting the head differential (h) vary. The curves may be used to estimate low discharge for the control structure provided the upstream tidal range is less than 0.5 foot. The accuracy decreases when the tidal range of the upstream stage recorder exceeds 0.5 foot during a given day. The following table shows: the recession iof discharge following the passage of hurricane Donna on September 10, 1960.

PAGE 50

44 FLORIDA GEOLOGICAL SURVEY Date Mean daily head Mean daily (September) differential (feet) discharge (cfs) 11 0.48 2,430 12 .38 2,180 13 .31 1,800 14 .28 1,660 15 .23 1,510 16 .17 1,290 17 .19 1,240 18 .22 943 Salt water invaded the canal on a rising tide on September 15. A comparison of the measured discharges in the table (all gates fully open) with the plotted points in figure 22 (only one gate partially open) indicates that the four gates can be partly closed within a week after a flood peak has occurred to maintain a head differential of at least 0.3 foot. This would keep the salt water out of the canal. The small increase in head required to maintain the same discharge would be insignificant from the standpoint of flood height above the dam. There is a distinct difference between holding salt water back, as shown in the schematic flow diagram of figure 21a, and allowing salt water to invade the canal, as in figure 21b. Inland penetration of salt water does not contribute to seaward discharge because, as previously defined, true removal of floodwaters must only mean removal of fresh floodwaters. Salt water which penetrates landward in the canal merely replaces fresh water; once it has gained entry, energy must be expended to remove it. In the velocity profiles of figures 8A and 8B, the velocities near the bottom of the canal are landward during the rising tide, and are small relative to the seaward velocities in the upper part of the canal during the falling tide. These velocities reflect two things: (1) Sea water provides back pressure which is greatest at the bottom of the canal. (2) Energy is expended in driving the intruded salt water back to the ocean on the falling tide. As the noncontributing landward flows and the reduced seaward flows of salt water must be integrated with the positive flow of fresh water to yield the mean daily discharge, the mere presence of salt water in the canal, in effect, reduces the mean daily discharge of true (fresh) floodwaters. In some respects, salt water acts as a movable dam in the bottom of the canal; it effectively reduces the depth of the canal

PAGE 51

REPORT OF INVESTIGATIONS No. 24 45 through which fresh water can flow seaward. For example, if salt water occupies the lower 6 feet of the canal when the control is fully open (fig. 8), the effective flow section for fresh water is reduced to a thickness of about 6 feet (total depth of canal is 12 feet). If the gates are lowered to an opening of, say, 7 feet, the fresh-water discharge will be impeded and this will cause the freshwater head to rise until salt water is driven out of the lower part of the canal. Conceivably, the head might build up sufficiently so that the fresh water would flow through the full 7-foot section below the sluice gate. However, there is no assurance that this will occur-fresh water may discharge through only the upper part of the 7-foot thick section while salt water continues to flow landward through the lower part. This leads to an important point. The effect upon head and discharge caused by constriction from lowering the gates, as opposed to constriction and back-water effect from salt water in the bottom of the canal, is indeterminate without controlled tests. The optimum gate opening probably occurs when the salt-water wedge is held just downstream from the control and only fresh water is flowing beneath the sluice gates. This gate opening will discharge the maximum amount of water without permitting salt water to invade the canal and aquifer above the control. Future studies should be directed toward evaluation of the relations of fresh-water to salt-water discharge with various gate openings, in the immediate vicinity of the control dam. SUMMARY The Snake Creek Canal cuts through highly permeable limestone of the Biscayne aquifer of southeastern Florida. The canal extends inland about 18 miles from the salinity-control dam located near the shoreline. The canal is somewhat representative of a number of canals in the area and information learned during-the study is generally applicable to other canals. Salt water was traced in the Snake Creek Canal under various circumstances of control-dam operation during a special test in March 1961. In the first phase of this test all four gates of the control dam were opened wide. Salt water moved inland about 2 miles during the rising tide and seaward during the falling tide. Salt water penetration into the aquifer was negligible during the 3 days the gates were fully open. In the second phase of the special test the control was completely closed at high tide, trapping salt water above the dam. A density

PAGE 52

46 FLORIDA GEOLOGICAL SURVEY current developed in the trapped salt water, which caused it to flow inland along the bottom of the canal at a rate of about 900 feet per hour until it reached a deep rock pit which acted as a hydraulic sink. A flushing operation removed salt water from the canal, but not that in the rock pit. Measurements of electrical conductivity during the subsequent months in the rock pit showed that the salt dispersed upward by a combination of wind, tidal mixing, and molecular diffusion. After reduction of salinity by this upward dispersion, the salt water in the rock pit was flushed seaward through the canal. When salt water is trapped in a canal by control-dam operation or by overtopping the dam during abnormally high tides, it is not necessary to wait for nor to have high fresh-water head behind the dam in order to perform a flushing operation. The falling tide (level) of the ocean creates a seaward hydraulic gradient and a condition favorable for a high rate of discharge from the canal. Calculations and observations show that flushing of 2 to 3 miles of the downstream reach of a canal is feasible during one falling tide. The canal should tap a reasonably large reservoir of fresh water and the control dam should be the type that permits rapid operation so that the gates can be opened just after high tide and closed at low tide. Although some fresh water is lost during flushing, good water management is better served by wasting a small portion of a vast reservoir of fresh water than by allowing a small volume of salt water to move inland to contaminate a much larger volume of fresh water. The movement of salt water in the aquifer adjacent to a canal is closely related to happenings in the canal itself. When the gates are closed, the downstream controlled reach of the canal is immediately converted from a condition of drainage to a condition of recharge. Salt water trapped behind the dam during the test entered the aquifer and quickly appeared in wells adjacent to the canal. Sections perpendicular to the canal showed that the saltwater outseepage from the canal moved much more rapidly horizontally than it did vertically. The residual effect of downward movement of salt water in the aquifer was recognizable for 6 months at one of the observation stations. The fluctuations of chloride content in shallow wells indicate that some of the lowsalinity water in the upper part of the aquifer was eventually recovered and flushed seaward through the canal during subsequent gate openings.

PAGE 53

REPORT OF INVESTIGATIONS NO. 24 47 The movement of high salinity water deep in the aquifer responds to head changes related to long-term operations of the control dam. The chloride content in deep wells increased during the fall of 1960 when the control remained open because of flood conditions. The year of 1961 was abnormally dry; infrequent opening of the control caused heads to remain sufficiently high so that general expulsion of salt water from the aquifer occurred throughout 1961. It is a contradictory situation that the greatest inland movement of salt water took place in a wet year and the greatest expulsion of salt water took place during a dry year. Nevertheless, this fact indicates that the means are available for very effective control of the movement of salt water in the Biscayne aquifer. CONCLUSIONS AS TO OPERATING CRITERIA The data in this report indicate that salt water need not and should never be trapped in a canal. As the position of salt water in a canal has only a tenuous relation to head, the following conclusions are made with regard to the operation of the gates: (1) Control gates, which operate automatically by sensing the head differential between the upstream and downstream sides of the dam, cannot be operated effectively on an automatic basis until the relations of salt-water movement and head are properly evaluated. It is entirely possible for the automatic operations to trap salt water inadvertently. (2) In.addition to the upstream and downstream stage recorders which are being installed at the new automatically controlled dams, a conductivity recorder should be installed in each canal slightly upstream from the control dam. Data from these three sources will permit future evaluation of the relation of salt-water movement to total head in the canal and to head differential through the control dam. (3) Persons responsible for gate operation should be furnished a portable conductivity instrument to determine the salinity at the bottom of the canal at different points. Knowledge of the presence and movement of salt water in the canal will aid greatly in proper operation of the control. The following suggestions are made for operation after heavy rainfall: (1) After the first inland penetration of salt water during a rising tide (which may occur within a week after a flood

PAGE 54

48 FLORIDA GEOLOGICAL SURVEY peak), the gates should be partly closed as conditions permit. (2) The gates should always be opened uniformly so that maximum fresh-water pressure is applied against the intruding salt water by increased velocities of the fresh water as it flows under the submerged sluice gates. (3) The instantaneous dynamic head that will prevent inland movement of salt water cannot be determined without controlled tests. However, the calculated hydrostatic head that will prevent intrusion at the bottom of S-29 is 0.3 foot. Therefore, salt water will not enter the bottom of the canal above the control structure if an instantaneous head of 0.3 foot is maintained between upstream and downstream sides of the dam at the time of minimum head differential, about 2 hours before high tide. Some experimentation will be necessary to establish specific operating criteria. Closing the control gates so that 0.3 foot of head will be maintained during the rising tide will tend to increase the stage in the upstream part of the canal by an amount not greater that 0.3 foot. This is a tolerable increase in stage even during floods.

PAGE 55

REPORT OF INVESTIGATIONS NO. 24 49 REFERENCES Cooke, C. W. 1939 Scenery of Florida interpreted by a geologist: Florida Geol. Survey Bull. 17. Ferguson, G. E. (see Parker, G. G.) Kohout, F. A. 1960 Cyclic flow of salt water in the Biscayne aquifer of southeastern Florida: Geophys. Research Jour., v. 65, no. 7, p. 2133-2141. 1961 Case history of salt water encroachment caused by a storm sewer in Miami: Am. Water Works Assoc. Jour. v. 53, no. 11, p. 1406-1416. Leach, S. D. 1963 (and Sherwood, C. B.) Hydrologic studies in the Snake Creek Canal area, Dade County, Florida: Florida Geol. Survey Rept. Inv. 24, pt. 3. Love, S. K. (see Parker, G. G.) Parker, G. G. 1951 Geologic and hydrologic factors in the perennial yield of the Biscayne aquifer: Am. Water Works Assoc. Jour., v. 43, no. 10, p. 817-835. 1955 (Ferguson, G. E., Love, S. K., and others) Water resources of southeastern Florida, with special reference to the geology and ground water of the Miami area: U. S. Geol. Survey WaterSupply Paper 1255. Sherwood, C. B. (see Leach, S. D.)

PAGE 56

-FLORIDA-GEOLOGICAL-SURVEY COPYRIGHT NOTICE © [year of publication as printed] Florida Geological Survey [source text] The Florida Geological Survey holds all rights to the source text of this electronic resource on behalf of the State of Florida. The Florida Geological Survey shall be considered the copyright holder for the text of this publication. Under the Statutes of the State of Florida (FS 257.05; 257.105, and 377.075), the Florida Geologic Survey (Tallahassee, FL), publisher of the Florida Geologic Survey, as a division of state government, makes its documents public (i.e., published) and extends to the state's official agencies and libraries, including the University of Florida's Smathers Libraries, rights of reproduction. The Florida Geological Survey has made its publications available to the University of Florida, on behalf of the State University System of Florida, for the purpose of digitization and Internet distribution. The Florida Geological Survey reserves all rights to its publications. All uses, excluding those made under "fair use" provisions of U.S. copyright legislation (U.S. Code, Title 17, Section 107), are restricted. Contact the Florida Geological Survey for additional information and permissions.