Title: Development of Wellhead Protection Areas for the Major Public Supply Wells in Hernando County, Florida
CITATION THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/WL00004529/00001
 Material Information
Title: Development of Wellhead Protection Areas for the Major Public Supply Wells in Hernando County, Florida
Physical Description: Book
Language: English
Publisher: HydroGeoLogic, Inc., Herndon, VA
 Subjects
Spatial Coverage: North America -- United States of America -- Florida
 Notes
Abstract: Jake Varn Collection - Development of Wellhead Protection Areas for the Major Public Supply Wells in Hernando County, Florida (JDV Box 91)
General Note: Box 23, Folder 1 ( Miscellaneous Water Papers, Studies, Reports, Newsletters, Booklets, Annual Reports, etc. - 1973 -1992 ), Item 32
Funding: Digitized by the Legal Technology Institute in the Levin College of Law at the University of Florida.
 Record Information
Bibliographic ID: WL00004529
Volume ID: VID00001
Source Institution: Levin College of Law, University of Florida
Holding Location: Levin College of Law, University of Florida
Rights Management: All rights reserved by the source institution and holding location.

Full Text










DEVELOPMENT OF
WELLHEAD PROTECTION AREAS
FOR THE MAJOR PUBLIC SUPPLY WELLS
IN HERNANDO COUNTY, FLORIDA



Final Technical Completion Report


-I






























































The Southwest Florida Water Management District (District) does not
discriminate upon the basis of any individual's disability status. This
non-discrimination policy involves every aspect of the District's
functions, including one's access to, participation, employment, or
treatment in its programs or activities. Anyone requiring reasonable
accommodation as provided for in the Americans With Disabilities Act
should contact Gwen Brown, Resource Projects Department, at 904-
796-7211 or 1-800-423-1476, extension 4226; TDD ONLY 1-800-
231-6103; FAX 904-754-6885/SUNCOM 663-6885.














DEVELOPMENT OF WELLHEAD PROTECTION AREAS
FOR THE MAJOR PUBLIC SUPPLY WELLS
IN HERNANDO COUNTY, FLORIDA



Final Technical Completion Report






Prepared for

Southwest Florida Water Management District
Hernando County




Prepared by

T. Neil Blandford
and
Tiraz R. Birdie

HydroGeoLogic, Inc.
1165 Hemdon Parkway, Suite 900
Hemdon, VA 22070


April 30, 1993

















This is to certify that I, Sandy Nettles, have reviewed the figures, tables, and text of the
following report, and have retained one copy for my files.
























-s --









Sandy Nttles, PG
FL Beg. No. 710


1165 Herndon Parkway. Suite 900, Herdon. Virginia 22070 USA
(703) 478-5186 FAX (703) 471-4180








TABLE OF CONTENTS



1 INTRODUCTION ..........................................1-1
1.1 Background and Objectives .............................1-1
1.2 Scope of Work .................. .................. 1-2
1.3 Organization of Report ...... ......................... 1-5

2 TECHNICAL APPROACH ....................................2-1
2.1 Background ........................................2-1
2.2 Technical Approach ...................................2-5
2.2.1 Level I Delineations Flow-Net Analysis ................ 2-5
2.2.2 Level II Delineations Analytical Modeling ............... 2-7
2.2.3 Level el Delineations Numerical Modeling .............. 2-8
2.3 Modeling Code Selection ............................... 2-9

3 HYDROGEOLOGICAL SETTING ...............................3-1
3.1 Introduction ........................................3-1
3.2 Geological Framework ................................3-1
3.3 Surface W ater .....................................3-4

4 DATA ANALYSIS ..........................................4-1
4.1 Potentiometric Surface Maps ............................4-1
4.2 Areal Recharge and Discharge ...........................4-5
4.3 Spring Discharge ................................... 4-6
4.4 Floridan Aquifer Parameters .............................4-9
4.4.1 Transmissivity ..................................4-9
4.4.2 Effective Porosity .............................. 4-14

5 ANALYTICAL CAPTURE ZONE DELINEATIONS .................... 5-1
5.1 Introduction and Background ............................5-1
5.2 Level II Delineations Analytical Modeling ................... 5-2
5.3 Level I Delineations Flow-Net Analysis .................... 5-10

6 REGIONAL GROUND-WATER FLOW MODELING ................... 6-1
6.1 Conceptual Model and Modeling Assumptions .................. 6-1
6.2 Grid Design ........................................6-4
6.3 Model Input Data ................................... 6-5
6.4 Regional Model Calibration .............................6-6
6.4.1 Calibration Procedure ............................6-6
6.4.2 Calibration Results ..............................6-6
6.4.3 Mass Balance ................................6-18
6.5 Regional Model Sensitivity Analysis ...................... 6-18










7 SUBREGIONAL GROUNDWATER FLOW MODELING AND
i CAPTURE ZONE DELINEATION .............................. 7-1
7.1 Subregional Ground-Water Flow Modeling ................... .7-1
7.2 Level III WHPA Delineations ............................7-4

8 FINAL WHPA DELINEATIONS ................................8-1
8.1 Technical Approach ..................................8-1
1 8.2 Rational for Selection of Final Delineation Method ............... 8-3
8.3 Final WHPA Delineation Results ........ ................. 8-5
8.4 Calculated Fixed Radius Calculations ...................... 8-10
8.5 Capture Zone Uncertainty Analysis ....................... 8-15

9 SUMMARY AND CONCLUSIONS ............................... 9-1

REFERENCES .. ... ...................................... R-1

APPENDIX A Basic Well Data




























iii
IV








LIST OF FIGURES


Figure Eage

1.1 Generalized study area base map ............................. 1-3

1.2 Location of existing and planned major public supply wells in Hernando
County .. ................... ..........................1-4

2.1 Schematic diagram of capture zone (WHPA) modeling terms for simplified
hydrogeological scenario (U.S. EPA, 1987) ........................ 2-2

2.2 Capture zone types that may be delineated for wellhead protection area
delineation purposes ............. ........................ .2-4

3.1 Generalized geologic cross sections for Hernando County (adapted from Fretwell,
1985) ................................................3-3

3.2 Contour map of top of the Upper Floridan aquifer. Adapted from Buono and
Rutledge (1978) .........................................3-6

3.3 Contour map of the bottom of the Upper Floridan aquifer. Adapted from Miller
(1986) ................ .............................. 3-7

4.1 May 1988 Upper Floridan potentiometric surface in Hernando County and
adjoining regions ............. .......................... 4-2

4.2 September 1988 Upper Floridan potentiometric surface in Hernando County and
adjoining regions .........................................4-3

4.3 General regions of ground-water recharge and discharge in Hernando County and
adjoining areas and locations of documented springs ............... 4-8

4.4 Observed 1988 hydrograph for the SWFWMD ROMP 105 deep well and monthly
recorded rainfall at Brooksville. ................................4-7

4.5 Observed Upper Floridan aquifer transmissivities in Hernando County and
adjoining regions from various sources ......................... 4-12

5.1 Ten-year Level II (WHPA code) capture zones for May 1988 conditions ...... 5-7

5.2 Ten-year Level II (WHPA code) capture zones for September 1988 conditions 5-8








Figure Pa=e

5.3 Ten-year Level II (WHPA code) capture zones for average 1988 conditions .... 5-9

i 5.4 Ten-year Level I (flow-net analysis) capture zones for May 1988 conditions ..5-11

5.5 Ten-year Level I (flow-net analysis) capture zones for September 1988
conditions ....... ... ............ ................... 5-12

6.1 Regional model grid and associated boundary conditions ................ .6-2

6.2 Simulated May 1988 Upper Floridan potentiometric surface .............. 6-7

6.3 Difference between the observed and simulated May 1988 Upper Floridan
potentiometric surfaces ............. ..... ................ .6-8

6.4 Simulated September 1988 Upper Floridan potentiometric surface. .......... 6-9

6.5 Difference between the observed and simulated September 1988 Upper Floridan
potentiometric surfaces ............. ..................... 6-10

| 6.6 Calibrated regional model transmissivities in ft2/d times 10,000 . . 6-14

6.7 Calibrated recharge to, and discharge from, the Upper Floridan for the May 1988
calibration period .............. ........... ........ ..... 6-15

6.8 Calibrated recharge to, and discharge from, the Upper Floridan for the September
I 1988 calibration period .................................... 6-16

6.9 Simulated potentiometric surface for first model sensitivity run in which Upper
i Floridan transmissivity was increased two-fold ..................... 6-21

F 6.10 Simulated potentiometric surface for second model sensitivity run in which Upper
t Floridan transmissivity was decreased 50 percent ................... 6-22

6.11 Simulated potentiometric surface for third model sensitivity run in which
I prescribed recharge to the Upper Floridan was increased two-fold ......... 6-23

6.12 Simulated potentiometric surface for fourth model sensitivity run in which
prescribed recharge to the Upper Floridan was decreased by 50 percent ..... 6-24

7.1 Location and identification number of subregional model domains .......... 7-2

7.2 Ten-year Level III (numerical) capture zones for May 1988 conditions ...... 7-5


v








Figure Page

7.3 Ten-year Level III (numerical) capture zones for September 1988 conditions .... 7-6

8.1 Final ten-year capture zones for Period 1 as outlined in Table 8.1 .......... .8-8

8.2 Final ten-year capture zones for Period 2 as outlined in Table 8.1 .......... 8-9

8.3 Schematic diagram of composite capture zone WHPA delineation approach. The
delineated WHPA would be the entire region encompassed by the outer lines of
the above diagram ...................................... 8-11

8.4 Final two, five and ten-year composite WHPAs for Hernando County ....... 8-12

8.5 One and two-year calculated fixed radius (CFR) WHPA delineations ....... 8-14

8.6 Final Period 1 ten-year capture zone and 85th, 90th and 95th ten-year capture
zone percentiles delineated for Spring Hill equivalent well .............. 8-16

8.7 End-member ten-year capture zones delineated using minimum transmissivity
value and maximum porosity value (a), and maximum transmissivity value and
minimum porosity value (b) ................................ 8-18








LIST OF TABLES


Table P.ag

5.1 Information for May 1988 Level I and Level I WHPA delineations ......... 5-4

6.1 Observed and simulated spring flows (ft3/s) for the regional model May and
September calibration periods ............................... 6-12

6.2 Regional model mass balance for May and September 1988 calibration periods 6-19

7.1 Summary of subdomain modeling grids ..........................7-3

8.1 Information for Final WHPA Delineations ......................... 8-6

A. 1 Basic information for major public supply wells in Hernando County ....... A-1








1 INTRODUCTION


1.1 Background and Objectives

The Amendments to the Safe Drinking Water Act (SDWA), which were passed in June 1986,
established the first nationwide program to protect ground-water resources used for public water
supplies from all (anthropogenic) potential threats. Unlike previous Federal programs, that have
tended to focus on individual contaminant sources, this new effort approaches the assessment and
management of ground-water quality from a more comprehensive perspective. The SDWA seeks
to accomplish this goal by the establishment of State Wellhead Protection (WHP) Programs that
"protect wellhead areas within their jurisdiction from contaminants which may have any adverse
effects of the health of persons." A WHP Program is part of a State's Ground Water Protection
Strategy.


One of the major elements of WHP is the determination of zones within which contaminant
source assessment and management will be addressed. These zones, called Wellhead Protection
Areas (WHPAs), are defined in the SDWA as "the surface and subsurface area surrounding a
water well or wellfield, supplying a public water system, through which contaminants are
reasonably likely to move toward and reach such water well or wellfield."


As of the writing of this document, the State of Florida has not yet developed a comprehensive
WHP program, although the Florida Department of Environmental Regulation (DER) has
conducted some preliminary planning and strategy sessions designed to investigate the potential
framework for such a plan. Some of the Water Management Districts, (most notably the
Southwest Florida and St. Johns River Districts) however, have been assisting local governments
within their jurisdiction with the delineation of WHPAs and the development of local WHP
measures; and at least seven local governments had enacted WHP ordinances by early 1992.
The local ordinances may be thought of as stop-gap measures utilized until the DER can
establish a state WHP program. It is expected that eventually the state will adopt local
protection areas as state protection areas (Blain and Evans, 1992).








The Southwest Florida Water Management District (SWFWMD), in cooperation with Hernando
County, Florida, jointly solicited proposals for the delineation of wellhead protection areas
(WHPAs) around the major public supply wells in Hernando County in the spring of 1991. The
major public supply wells were considered to be those pumping more than 100,000 gallons per
day (gpd). The delineated WHPAs were expected to be an integral component of Hemando
County's Ground-Water Protection Ordinance, which will impose certain restrictions on the use,
1 handling, storage, production and disposal of regulated substances in the vicinity of public
supply wells. HydroGeoLogic, Inc. submitted a proposal to the SWFWMD and was
; subsequently retained by the District and the County to perform the technical delineation of
WHPAs in Hernando County.



1.2 Scope of Work

r The scope of work for this project consists of the major tasks listed below:

The compilation and analysis of existing hydrogeological data for
Hernando County and immediately adjoining regions (Figure 1.1).

The delineation of WHPAs for the major public supply wells
(Figure 1.2) in the county using flow-net analysis (Level I) and
semi-analytical modeling (Level II) methods.

The development and calibration of a regional ground-water flow
model encompassing all of Hernando County and portions of
adjacent counties for May and September 1988 hydrologic
conditions.

The construction of four modeling subregions based upon the
regional model results, and the delineation of capture zones for the
major public supply wells in each subregion (Level I WHPA
delineations).

i' Conduct a comparison of the WHPA delineations obtained using
the various methods, and prepare a recommendation as to the final
WHPAs that should be implemented into the Ground-Water
Protection Ordinance.










































Figure 1.1 Generalized study area base map.





FT- I 1 777


Figure 1.2 Location of existing and planned major public supply wells in Hernando County.


~`T7 ~------I
i








Develop a comprehensive report detailing the data analysis and
technical work performed.



1.3 Organization of Report

This report is divided into nine chapters designed to lead the reader through the technical effort
in a sequential and logical manner. Chapter 1 provides background introductory materials, and
Chapter 2 outlines the general technical approach. Chapter 3 provides a synopsis of the
hydrogeological setting. Chapter 4 presents the data types and sources used, as well as any
technical analysis performed on the raw data. Chapter 5 presents the analytical (Level I and
Level II) capture zone delineations. Chapter 6 describes the construction and calibration of the
regional numerical ground-water flow model, and Chapter 7 presents the sub-regional numerical
modeling and capture zone delineations (Level III). Chapter 8 presents the final WHPA
delineations and the rationale for their selection, and Chapter 9 consists of a brief summary and
conclusions.








2 TECHNICAL APPROACH


2.1 Background

The primary objective of this study was to delineate reasonable, technically valid WHPAs for
each of the major public supply wells in Hernando County. To this end, a variety of delineation
methods were applied. The methods range from fairly simple (flow-net analysis) to highly
complex (fully three-dimensional numerical modeling). The methods were applied in ascending
order of technical sophistication, with each phase of the WHPA delineation study adding
increased insight and technical basis for the following phase. Based upon the results obtained
using each of the three methods, a methodology for completing the final WHPA delineations to
be incorporated into the County's Ground-Water Protection Ordinance was identified and
implemented.


An overview of the technical approach followed for the delineation of WHPAs in Hernando
County is presented later in this section. Prior to initiating the technical discussion, however,
some basic terms and principals that are important in WHPA delineation modeling are presented.
Figure 2.1 illustrates a schematic diagram of the ground-water flow field in the vicinity of a
pumping well for a simplified hydrogeological scenario. Although in most instances the
hydrogeological setting in Hernando County is more complex than that illustrated in Figure 2.1,
the illustration is useful for the presentation of general concepts.


In Figure 2.1 the zone of contribution (ZOC), zone of influence (ZOI), and zone of transport
(ZOT) are all clearly illustrated. The ZOC is defined as the entire volume of aquifer that
contributes ground-water recharge to the pumping well. In a hydrogeological setting that has
a sloping water table or potentiometric surface under ambient (non-pumping) conditions, the
extent of the ZOC is defined downgradient of the well by the location of a stagnation point, and
it is defined upgradient of the well by some hydrogeological condition such as ground-water flow
divide. The extent of the ZOC is determined by site conditions such as pumping rate, areal
recharge or leakage to the aquifer, and aquifer hydraulic properties. The ZOI of a well is the








---ZOT (10 YR)--I
S---zoc I -
SZOI *I GROUND-WATER
I I Z I -I I DIVIDE
J PUMPING DV
',I WELL
83 r T~we~A'


PUMPING WELL


AK


LEGEND: (B) PLAN VIEW
2 Water Table
S" 10 Year Zone of Transport
Direction of Ground-water Flow
ZOC Zone of Contribution
ZOI Zone of Influence
ZOT Zone of Transport


Schematic diagram of capture zone (WHPA) modeling terms for simplified
hydrogeological scenario (U.S. EPA, 1987).


Figure 2.1


_ ,, I 1 111 1








volume of aquifer within which there is a change in the water table or potentiometric surface due
to the withdrawal of water at the well. In cross-section, the ZOI is commonly referred to as the
cone of depression caused by the well. Note that ground water outside the ZOI of a well may
eventually reach the well, and conversely some ground water within the ZOI will never enter
the well. The zone of transport (ZOT) is simply a ZOC that is modified using a time of travel
(TOT) criterion such as 5 or 10 years. ZOTs in this report are referred to as ZOCs modified
l using a time of travel criterion.


The term capture zone is often used when delineating WHPAs. In general, the term capture
zone corresponds to the ZOC, or time-modified ZOC, of a pumping well. Three types of
i capture zones, as classified by Blandford and Huyakorn (1991), are presented in Figure 2.2.


. A steady-state capture zone is the surface or subsurface region surrounding a pumping well that
will supply ground-water recharge to the well over an infinite period of time. The open-ended
I; shape of the capture zone in Figure 2.2 is due to the fact that, given enough time, any particle
of water upgradient of the well within the capture zone will eventually travel to the well. In
practice, the upgradient end of the capture zone would be "capped" in some manner due to
physical and/or managerial restrictions. For example, a steady-state capture zone may terminate
l at a ground-water flow divide.


SA time-related capture zone is the surface or subsurface region surrounding a pumping well that
will supply ground-water recharge to the well within a specified period of time. This type of
capture zone is most commonly used in practice. A time-related capture zone is always
. represented by some closed shape. In general, time-related capture zones are less conservative
(enclose smaller areas) than steady-state or hybrid capture zones. As the specified time
increases, however, differences between the three capture zone types in the proximity of the
pumping well become negligible.


As the name implies, a hybrid capture zone is a combination between a steady-state and a time-
r related capture zone. The hybrid capture zone is identical to the steady-state capture zone in all













WATER DIVDE STREAMLINE


STEADY-STATE
CAPTURE ZONE


TIME-RELATED
CAPTURE ZONE


HYBRID
CAPTURE ZONE


Figure 2.2


t=T


AMBIENT
4" FLOW


CAP
BOUNDARY








WATER DIVIDE STREAMLINE


Capture zone types that may be delineated for wellhead protection area
delineation purposes.


I








respects expect that it is "capped" on the upstream end (Figure 2.2). A particle of water
released from the mid-point of the capping segment will reach the pumping well within some
specified time. Therefore, the cap on the hybrid capture zone approximates a segment of a time-
related capture zone. The hybrid capture zone can be viewed as an implementable alternative
to the steady-state capture zone.



2.2 Technical Approach

This section summarizes the technical approach for the delineation of WHPAs in Hernando
County. The technical basis and the assumptions and limitations of each delineation procedure
is outlined. The various approaches are compared and contrasted, and the relative strengths and
weaknesses of each approach are highlighted.



2.2.1 Level I Delineations -.Flow-Net Analysis

The initial WHPA delineations for Hernando County were obtained using methods that are
"intermediate" in terms of their level of sophistication. The first method applied was flow-net
analysis. This method involves the delineation of the ZOC, or capture zone, for a given well
or well field based upon the configuration of an existing potentiometric surface map. If the
aquifer of interest is assumed to be isotropic, and ground-water flow within the aquifer is
predominantly two-dimensional, then the limiting pathlines that form the capture zone may be
delineated by drawing lines that intersect the potentiometric surface contours at right angles.
The capture zone is delineated upgradient from the well or well field until it is "capped"
according to a time of travel criterion (e.g. 5 yrs).


The major advantages of this method are two-fold. First, the major data input requirement,
potentiometric surface maps of the Floridan aquifer, are readily available. These maps are
constructed bi-annually by the USGS for the months of May and September. These two periods
are believed to be indicative of the extreme potentiometric surface fluctuations within the
Floridan aquifer system. The May map represents the potentiometric surface following the








relatively dry period in spring, which is usually a period of relatively large aquifer withdrawals.
The September map represents the effects of recharge to the Floridan aquifer following the wet
summer period which is usually a period of relatively small aquifer withdrawals.


The second major advantage to the flow-net analysis method is that, if the assumptions of
isotropic hydraulic conductivity and predominantly two-dimensional ground-water flow in the
Floridan aquifer are reasonable, then the method intrinsically accounts for all of the physical
complexities (e.g. spatially variable recharge and discharge), known and unknown, that affect
the capture zone. This is so because the capture zones are delineated directly from
potentiometric surface maps, which are themselves a product of all of the factors that affect the
ground-water flow system. Therefore, although the flow-net method of WHPA delineation is
not the most sophisticated technically, it has some appealing aspects.


There are, however, several significant weaknesses and limitations associated with using flow-net
analysis for the delineation of capture zones in Hemando County. The most serious limitation
is that the Floridan aquifer potentiometric surface maps are constructed on a regional scale.
Because the cones of depression caused by pumping wells in Hernando County are relatively
small, they are not evident in the potentiometric surface maps. This condition makes it
impossible to apply the flow-net analysis method in the vicinity of the wellhead. To circumvent
this problem, the Level I delineations were performed using the Level II delineations (described
in the next section) as a starting point (see Chapter 5).


Another significant limitation of this method is that travel times are approximated in a rather
crude fashion by dividing pathlines into multiple segments over which the hydraulic gradient is
relatively uniform. Using Darcy's Law, an estimate of the advective travel time for ground
water along each segment can be made. This procedure is fairly labor intensive.


Finally, this method of WHPA delineation has no predictive or sensitivity analysis capability.
The ZOCs are constructed for a given set of existing conditions defined by the configuration of
a potentiometric surface map; changes in the capture zones due to changes in the potentiometric








surface map may not be evaluated until the new potentiometric surface map is constructed. This
could pose a significant limitation when topics such as the effects of future pumping increases
need to be addressed.



2.2.2 Level II Delineations Analytical Modeling

The second method applied to delineate WHPAs in Hernando County is analytical ground-water
flow modeling. Analytical models are exact mathematical solutions to the partial differential
equation of ground-water flow., However, in order to obtain an exact mathematical solution, a
number of simplifying assumptions must be made concerning the physical system. These
assumptions generally include, but may not be limited to, two-dimensional areal ground-water
flow; fully penetrating pumping/recharge wells; homogeneous aquifer properties (e.g.
transmissivity and porosity); and fully penetrating boundary conditions such as streams or
barriers. Different analytical solutions may be superposed to simulate flow fields of varying
complexity. For example, the analytical solution for a pumping well in a homogeneous, leaky-
confined aquifer may be added to the analytical solution for uniform ambient ground-water flow
within the aquifer. When the limiting assumptions are not violated too badly, analytical models
are an efficient, powerful tool for the delineation of WHPAs. The widely used EPA WHPA
code (Blandford and Huyakorn, 1991) was selected for use to perform the analytical delineations.


The analytical modeling delineation approach requires that a variety of site-specific data be
acquired for each well or well field. The typical input parameters to the WHPA code are the
well pumping rate, the aquifer transmissivity and thickness, the angle of regional (ambient)
ground-water flow and the ambient hydraulic gradient. Once these key data were collected, the
WHPA code was used to delineate the two, five and ten-year capture zones for May and
September, 1988 (the reasons for selecting 1988 are presented later in the report).


For sites where the use of analytical modeling is appropriate, this method has significant
advantages over the flow-net delineation method. Using the WHPA code, capture zones may
be delineated extremely efficiently, and it is a simple effort to perform sensitivity analysis on








model input parameters (such as porosity) or predictive simulations during which certain
parameters, such as the well pumping rate, may change. The analytical WHPA delineations
were also used to determine the extent of numerical modeling subregions used to obtain the
Level III delineations.


The primary drawback of the analytical modeling delineation method is that a series of
simplifying assumptions concerning the ground-water flow system must be made. The major
assumptions are that the aquifer is homogeneous and has a constant thickness; pumping wells
discharge at a constant rate and penetrate the entire aquifer thickness; and that ground-water flow
is two-dimensional in the areal plane. However, it is often the case that these assumptions are
not violated to such a degree that renders the analytical models inapplicable. If such is the case,
it is highly desirable to use analytical models due to their efficiency, versatility, and ease of use.



2.2.3 Level HI Delineations Numerical Modeling

Following the Level II WHPA delineations, but prior to the Level III delineations, a regional
ground-water flow model encompassing Hernando County and immediately adjoining areas was
developed. The purpose of the regional model was to provide a computer simulation capability
for predicting the Upper Floridan potentiometric surface on a regional scale. Using the
numerical model, physical complexities such as heterogeneous aquifer conditions and spatial
variations in areal recharge and discharge rates could be addressed. The numerical modeling
approach is also ideally suited to conducting sensitivity and predictive analysis. The
discretization of the regional model domain is too coarse (each model cell is 1 mile square) to
obtain accurate ground-water flow pathlines in the vicinity of individual wells. To implement
a finer discretization more suitable to the delineation of capture zones, four subregional models
were constructed based upon the regional modeling results. The subdomain model input
parameters and boundary conditions (prescribed head) were obtained from the regional model.
The USGS ground-water flow code MODFLOW (McDonald and Harbaugh, 1988) was used for
the regional and subregional ground-water flow modeling.








The subregional modeling results (hydraulic heads and aquifer properties) were used as input to
a three-dimensional pathline (capture zone) delineation code. Using this approach, the two, five
and ten-year capture zones were estimated. The major problem encountered during this stage
of the project was that, since the cones of depression caused by the Hernando County wells are
small, it was impractical to develop subregional model grids that were fine enough to provide
a detailed configuration of each cone of depression. For this reason, the numerical results
proved unsuitable for providing a detailed representation of each capture zone, but the general
configuration of the capture zone for each well could be estimated.



2.3 Modeling Code Selection

To perform the analysis outlined in the previous section, three computer modeling codes were
required: one for analytical capture zone delineation; one for three-dimensional numerical
ground-water flow modeling; and one for three-dimensional ground-water flow pathline
delineation. The rational behind the selection of the three codes applied is presented in this
section.


The WHPA code version 2.0 (Blandford and Huyakorn, 1991) was selected for the analytical
delineation of WHPAs in Hernando County. The WHPA code was developed by
HydroGeoLogic, Inc. for the U.S. EPA Office of Ground-Water Protection (OGWP). It is an
easy-to-use, widely applicable tool for WHPA delineation. The WHPA model contains four
major computational modules: RESSQC, MWCAP, GPTRAC, and MONTEC. The latter three
of these modules were developed specifically for the EPA using state-of-the-art technology and
some recently published studies available in the literature.


Due to its user-friendly nature, the WHPA code is a highly efficient tool for delineating WHPAs
using the analytical modeling approach. Input data sets may be efficiently created, manipulated
and stored using the menu-driven preprocessor. Output options include the ability to scale
capture zone plots to any of the standard USGS map scales or any alternative scale, and capture
zone plots may be saved in a variety of formats that make the transfer of model output into the


2-9








ARC/INFO Geographic Information System (GIS) or a computer aided design (CAD) software
package trivial.


The USGS three-dimensional ground-water flow code MODFLOW (McDonald and Harbaugh,
1988) was selected for the regional and local ground-water flow modeling in Hernando County
because it is a widely accepted, public domain code developed by the USGS; it has been used
in numerous previous studies to model regional ground-water flow in various parts of Florida,
including Hernando County; it has the capability to incorporate the appropriate system features;
and it is computationally efficient and relatively easy to use. There is also a great deal of
accessory software that enhances use of the model by providing efficient pre- and postprocessing
capabilities. In addition, SWFWMD staff were already familiar with the MODFLOW code and
have used it in their District-wide modeling effort as well as numerous other modeling studies.
MODFLOW is designed to simulate steady-state or transient ground-water flow through
heterogeneous, anisotropic porous media in three dimensions, subject to a variety of complex
boundary conditions. The code, therefore, is quite versatile in that it can be used to simulate
a wide variety of hydrogeological conditions that may exist in the field.


The GPT3 (general particle tracking in three-dimensions) code was selected for the Level HI
(numerical model) capture zone delineations. GPT3 is a public domain code developed for the
U.S. Department of Energy. GPT3 has the capability to delineate ground-water flow pathlines
and time-related capture zones in three dimensions for saturated and variably saturated porous
media. The code is a direct outgrowth of the GPTRAC module of the WHPA code. GPT3 uses
the semi-analytical pathline delineation technique presented by Pollock (1989). This technique
is superior to traditional numerical-integration pathline delineation methods used in older codes
because the refraction of pathlines at the interface of different aquifer materials is simulated
precisely. Pathline delineation codes that do not use this recently developed method will smooth
the angle of refraction as a pathline crosses from one region of hydraulic conductivity into
another. Required inputs for GPT3 are the three-dimensional grid configuration, the hydraulic
conductivity distribution, and the hydraulic head field obtained from numerical modeling (finite
difference or finite element) analysis.


2-10


T








3 HYDROGEOLOGICAL SETTING


3.1 Introduction

The geological and hydrogeological setting of Hernando County has been described by numerous
authors. Fretwell (1985), Ryder (1985) and SWFWMD (1987) provide three of the most recent
and comprehensive discussions. Useful descriptions can also be found in Trommer (1987),
Yobbi (1989), Fretwell (1983) and Cherry et al. (1970). The following sections are not intended
to reproduce, but rather to summarize, the previous body of relevant literature as it pertains to
the study at hand.



3.2 Geological Framework

A generalized geologic column, adapted from Fretwell (1985), is presented in Table 3.1. Two
representative cross sections are presented in Figure 3.1. In general the subsurface within the
study area is composed of the Tertiary age Suwannee Limestone, the Ocala Limestone, and the
Avon Park Formation. The Avon Park Formation contains both the Lake City Limestone and
the Avon Park Limestone often cited in earlier studies. The Tampa Limestone may overlay the
Suwannee Limestone at some places along the Brooksville Ridge, but it is generally absent
(Fretwell, 1985). The Oldsmar and Cedar Keys Formations lie below the Avon Park Formation,
but these units are not of interest for the current study (see Section 3.4). The Tertiary
limestones are overlain by surficial Quaternary deposits consisting of undifferentiated sands, silts
and clays which may have thicknesses up to 100 ft in the Brooksville Ridge area.


No major structural discontinuities within the Tertiary carbonate rocks due to faulting have been
reported. The major structural feature within the study area, the Ocala "uplift", is described by
Miller (1986). The uplift is about 20 mi wide and trends approximately 25* west of north; its
axis passes through Hernando County along the east side of Brooksville. Miller states that
because the Ocala "uplift" only affects sediments younger than the middle Eocene (i.e., the









Table 3.1 Generalized geologic column for Hernando County and adjoining regions (adapted
from Fretwell (1985)).


System Series Stratigraphic Thickness Lithology Water-producing
unit (ft) characteristics
Quaternary Holocene Undifferentiated 0-100 Soil, sand, and clay of marine and Generally not a
and deposits estuarine terraces, alluvial, lake, source of water.
Pleistocene and windblown deposits
Tertiary Pliocene Alachua and 0-100 Predominantly clay; some grayish- Confining layer in
and Hawthorn green, waxy; some interbedded sand some places;
Miocene Formations and and limestone, phosphatic clay, generally not a
Tampa marl, calcareous sandstone, source of water.
Limestone limestone residuum
Oligocene Suwannee 0-150 Limestone, cream to tan colored, Many domestic and
Limestone fine-grained, fossiliferous, thin- irrigation wells
bedded to massive, porous produce water from
the lower part.
Eocene Ocala Limestone 100-500 Limestone, white to tan, Yields large
(upper) fossiliferous, massive, soft to hard, quantities of water
porous to wells completed
above evaporites.
Eocene Avon Park 200-800 Limestone and dolomite. Limestone
(middle) Formation is light to dark brown, highly
fossiliferous, and porosity is
variable in lower part. Dolomite is
gray to dark brown, very fine to
microcrystalline and contains porous
fossil molds, thin beds of
carbonaceous material, and peat
fragments. Formation generally
contains evaporites in lower part.





















A'
(EAST)


A
(WEST)


FEET


2-. 100
100
SEA
LEVEL
100

200

300
400
* 4003 --


0*
o o


FEET
200

100
SEA
LEVEL
100

200

300

400


8240' 8230' 82020' 82010'
I I I I


28S40'-


[w-8318 I


O 10 MILES
0 8 16 KILOMETERS


EXPLANATION

t707 WELL AND FLORIDA BUREAU
OF GEOLOGY REFERENCE
NUMBER



SAND AND CLAY UNITS
INCLUDING THE ALACHUA
AND HAWTHORN FORMATIONB8
UNDIFFERENTIATED



B-- LINE OF SECTION


VERTICAL SCALE GREATLY EXAGGERATED


Generalized geologic cross sections for Hernando County (adapted from Fretwell,
1985).


(NORfT)

I :
81 J


Figure 3.1








Suwannee Limestone, the Ocala Limestone, and the upper portion of the Avon Park Formation),
it is not a true uplift, and was produced by sedimentational (rather than tectonic) processes.


Sinkholes occur due to the solution of carbonate rocks over time. As a sufficient volume of rock
is dissolved and carried away by ground water, the remaining infrastructure will eventually
collapse under the weight of the overburden. The collapse may be sudden or occur very
gradually over time. Numerous sinkholes exist in Hernando County. For a detailed discussion
of their distribution and genesis refer to SWFWMD (1987), Trommer (1987), Fretwell (1985)
and the references contained therein.



3.3 Surface Water

Surface water features in Hernando County consist of rivers, lakes and swamps. The only
perennial streams inland from the coast are the Withlacoochee and Little Withlacoochee Rivers
in the eastern part of the county. A number of other perennial streams exist along the coast and
are spring fed; the largest of these streams are the Weeki Wachee River fed by Weeki Wachee
Springs, and the Chassahowitzka River fed by the Chassahowitzka spring complex (the
Chassahowitzka River is just north of Hernando County). None of these streams drains a
substantial portion of the county land mass; Fretwell (1985) states "Most drainage in Hernando
County is internal, as is typical in karst terrain .... After heavy rainfall, small intermittent
streams flow to sinkholes where the water either percolates rapidly or ponds to form prairie
lakes."


The major swamps within the county exist in the coastal lowlands west of Highway 19 or in the
extreme eastern end of the county south of the Little Withlacoochee River. The swamps in the
eastern portion of the county are generally regions of small net recharge to the ground-water
system, while the coastal swamps are regions of ground-water discharge.


Numerous small lakes occur throughout Hernando County. Fretwell (1985) states "Some lakes
appear to be surface expressions of water tables perched on impermeable materials; others are








directly connected to the Floridan aquifer system through sinkholes and reflect the potentiometric
surface of the aquifer." Most lakes in Hernando County, with the exception of those that lie in
the extreme western portion of the county along the border of the coastal swamps, probably act
as a source of recharge to the ground-water flow system.



3.4 Ground Water

The Floridan aquifer system is the major source of ground water within the study area. Tibbals
(1990) states "The top of the Floridan (aquifer system) is defined as the first occurrence of
vertically persistent, permeable, consolidated carbonate rocks." Throughout most of Florida,
the Floridan aquifer system has two distinct producing zones separated by a middle
semiconfining unit. The upper production zone is referred to as the Upper Floridan aquifer, or
simply the Upper Floridan (likewise for the Lower Floridan). What is termed the Upper
Floridan in most reports is the only aquifer considered in this study. Ryder (1985) explains

"The base of the Upper Floridan aquifer in west-central Florida (including
Hernando County) is generally at the first occurrence of vertically persistent,
intergranular evaporites. This base is equivalent to the 'middle semiconfining
unit' used regionally... Hydraulic tests of rocks with intergranular evaporites
indicate that they have extremely low permeabilities. The rocks below the
'middle semiconfining unit' have known or estimated low transmissivity;
therefore, for all practical purposes, freshwater flow is limited to rocks above the
section with intergranular evaporites in west-central Florida."

In this report, as in Ryder (1985), the terms Upper Floridan and Floridan aquifer are assumed
to be synonymous.


Figures 3.2 and 3.3 are contour maps of the top and bottom of the Upper Floridan, respectively.
Figure 3.3 was adapted from Miller (1986), but Figure 3.2 was adapted from Buono and
Rutledge (1978) rather than Miller, since they used a greater number of control points to develop
their map and it is believed to be more accurate for Hernando County. The thickness of the
Upper Floridan ranges from about 700 ft along the coast up to 900 ft in the vicinity of
Brooksville. In the vicinity of the coast, the top of the Floridan aquifer is generally at or


---mm
















































Contour map of top of the Upper Floridan aquifer. Adapted from Buono and
Rutledge (1978).


Figure 3.2


-- -- ------- ---------r -I------ ---._...,,.~ -mICIC-r.
















































Contour map of the bottom of the Upper Floridan aquifer. Adapted from Miller
(1986).


Figure 3.3


7-7








slightly below land surface. Throughout the remainder of the county, the top of the Upper
Floridan may be from 0-100 ft below land surface (see Figure 3.1).


The Floridan aquifer in Hernando County generally exists under unconfined conditions (Miller,
1986; Fretwell, 1983 and 1985; Camp and Barcelo, 1988; Ryder, 1985; and Anderson and
Laughlin, 1982). Locally, the aquifer may be semiconfined or perched water tables may exist.
Fretwell (1985) states

"In some places, as on hills in the Brooksville Ridge area, a perched water table
in the surficial aquifer of limited areal extent occurs above the Floridan aquifer
system due to separation of the sand from the underlying limestone by clay of
very low permeability. In most parts of the county, there are sufficient breaches
in the clay layer to allow percolation of water from the sand into the underlying
limestone. In areas where saturated sand lies directly above limestone, water in
the sand is hydraulically connected to the Floridan aquifer system."

Ground-water flow in the Floridan aquifer in Hernando County is generally toward the coastline
(Section 4.1) and, since the majority of ground water is discharged at a series of coastal springs,
the freshwater flow system probably terminates in the vicinity of the coast. This conclusion is
supported by Fretwell (1983), who illustrates a relatively sharp saltwater-freshwater transition
zone that extends only 2-3 miles inland of the coast at a depth of about 400 ft below msl.









4 DATA ANALYSIS


4.1 Potentiometric Surface Maps

Potentiometric surface maps for the Upper Floridan aquifer are constructed bi-annually by the
USGS for the months of May and September. These two periods are generally indicative of the
extreme potentiometric surface fluctuations within the Floridan aquifer system. The May map
represents the potentiometric surface following the relatively dry period in Spring, which is
usually a period of relatively large aquifer withdrawals. The September map represents the
effects of recharge to the Upper Floridan following the wet summer period, which is usually a
period of relatively small aquifer withdrawals. In Hernando County, the potentiometric surface
may be considered to be in an essentially undeveloped (or redevelopment) state. Fretwell
(1985) states "Even though water levels have fluctuated seasonally over the years, the (well)
hydrographs (in Hernando County) do not indicate any long-term trend toward higher or lower
levels". This conclusion is supported by Ryder (1985), who states "... the ground-water
resource from about the Pasco-Hernando County line northward is relatively undeveloped ...
The flow system in the northern area, typified by high rates of natural recharge and spring
discharge, has probably changed little from redevelopment conditions." Since the ground-water
flow system exists in an essentially undeveloped state, the amount of rainfall, rather than ground-
water pumping, is the primary factor affecting fluctuations of the potentiometric surface in
Hernando County.


The observed May 1988 and September 1988 Upper Floridan potentiometric surfaces within
Hernando County and adjoining regions are illustrated in Figures 4.1 and 4.2, respectively.
Throughout most of Hernando County, the May and September potentiometric surfaces are very
similar, and the May surface is generally no more than about 2-3 ft lower than the September
surface.


Ground-water flows from regions of high potentiometric surface values to potentiometric surface
lows; ground-water flow pathlines may be drawn at right angles to the potentiometric surface


















































May 1988 Upper Floridan potentiometric surface in Hemando County and
adjoining regions.


Figure 4.1


____I_ __I~ _I__ ___~_ _~rP ___i_ mi___ __II*_ _:__ 1__1_ ____I ______~___1_ _C__ 111__























*F C CHE/ /






"70 DE
Ioo G o



Figure 4.2 September 1988 Upper Floridan potentiometric surface in Hernando County and
adjoining regions.








contours if two-dimensional areal flow is assumed and if the aquifer is isotropic in the areal
plane. Figures 4.1 and 4.2 indicate that ground water enters Hernando County from the south
and east. Ground-water inflow from the east is the result of ground-water recharge in the
vicinity of the Green Swamp potentiometric high roughly centered about Polk City in Polk
County to the southeast of the study area. The Pasco potentiometric high is not as large as the
Green Swamp potentiometric high, but it is a prominent feature of the Upper Floridan
potentiometric surface in south-central Hernando County and central Pasco County. In
Hernando County, ground-water flows from the Pasco high to the northwest, north and
northeast. There is a prominent trough in the potentiometric surface (often referred to as a
reentrant) to the east of the Pasco high in the vicinity of the Withlacoochee River, which is
indicative of concentrated ground-water discharge (Fretwell, 1985). Areal recharge to the Upper
Floridan occurs throughout most of Hernando County, except in the coastal lowlands where areal
discharge and concentrated ground-water discharge at numerous springs occurs.


An important feature of the potentiometric surface in Hernando County with respect to wellhead
protection is the fact that the cones of depression due to individual or multiple pumping wells
are not evident. This observation is consistent with the fact that aquifer transmissivities and
areal recharge rates are generally high throughout the County, and these physical conditions
inhibit the formation of substantial cones of depression. This observation is supported by
Fretwell (1985), who presents the results of two aquifer tests conducted about 3 mi northeast of
Weeki Wachee. In the first test, the stressed well was pumped at 1,080 gallons per minute
(gpm) for 48 hours. Although 55 ft of drawdown was measured at the well bore (90 percent
of which occurred during the first minute of pumping), no drawdown was measured in two
observation wells 309 ft and 1,063 ft away. For the second test, one of the observation wells
of the first test was pumped at 1,850 gpm for 48 hrs. At the end of the test, only 5.4 ft of
drawdown was observed in the pumping well and zero drawdown was observed in the nearest
observation well 754 ft away. The water levels in both production wells approached a steady-
state condition by the end of the testing period. Fretwell (1985) concludes "... the Upper
Floridan aquifer in this area can yield large quantities of water without the development of
extensive cones of depression".








4.2 Areal Recharge and Discharge


Areal recharge to the Upper Floridan occurs throughout most of Hernando County. Previous
studies have generally delineated the central portion of Hernando County, from about the town
of Weeki Wachee in the west to the Withlacoochee River in the east, as a region of high (greater
than 10 in/yr) recharge potential (Aucott, 1988; Anderson and Laughlin, 1982). SWFWMD
(1987) states that from the sandhills east of U.S. Highway 19 to the Brooksville Ridge there is
a high recharge potential due to the high permeability of the soils and the general lack of a
confining layer between the surficial soils and the Upper Floridan aquifer. There are also a
substantial number of sinkholes that are believed to be directly connected to the Upper Floridan
aquifer in this area (Trommer, 1987). Within the Brooksville Ridge area the soils are generally
poorly drained (low permeability), but the recharge potential is still high due to the presence of
sinkholes and breaches in the confining unit. There are several large sinkholes within Hernando
County that act as concentrated points of recharge to the ground-water system during wet
periods. The largest of these are Peck Sink, about 2 mi southwest of Brooksville, and Blue
Sink, about 5 mi northeast of Brooksville. Immediately following periods of substantial rainfall,
significant quantities of water may recharge the Upper Floridan through these sinkholes, and the
potentiometric surface measured locally may differ substantially from that observed on a regional
scale (Trommer, 1987). A generalized map of recharge and discharge areas in Hernando
County is provided in Figure 4.3. This map was developed using several different sources, but
it is based primarily upon a similar map presented in SWFWMD (1987).


Fretwell (1985) suggests that recharge in Hernando County can be conservatively estimated to
be at least 15 in/yr, and could be as high as 20 in/yr (for average conditions) if average
evapotranspiration is considered to be 36 in/yr. Maximum recharge rates in Hernando County
derived from modeling studies have been set at 22 in/yr (Adams, 1985), 20 in/yr (Ryder, 1985),
and 30 in/yr (Yobbi, 1989). Each of these studies was conducted for average, redevelopment
conditions. During a modeling study conducted for Pasco County to the south, Fretwell (1988)
states that recharge to the surficial aquifer in internally drained areas could reach a maximum
of 28 in/yr.








At Brooksville, the average rainfall for the period of record (1901-1990) is 56 in/yr. During
1988, however, the annual rainfall at Brooksville was significantly above average at 68 in/yr.
If evapotranspiration for 1988 can be considered to be about 36 in/yr or so, the maximum annual
recharge rate for Hernando County in 1988 may have exceeded 30 inches.


The amount of recharge to the water table that occurs in response to a rainfall event is highly
variable and depends, among other things, upon the depth of the unsaturated zone, the soil type,
the antecedent soil conditions, the intensity and duration of the rainfall event, and climatic
conditions such as temperature. There is also a time delay between a rainfall event and the
water-table response. In Hernando County, the time delay may range from about 2 weeks to
2 months, depending upon site-specific conditions. Figure 4.4 illustrates the observed 1988
hydrograph of the ROMP 105 deep observation well at Brooksville and the total monthly rainfall
at the Brooksville station. The superimposed curves indicate that at the ROMP 105 deep well,
there is about a 1-2 month delay between rainfall events and the corresponding water-table
response. For example, a substantial amount of rainfall occurred during the months of July,
August and September, but the observed hydrograph did not peak until mid-November.


Most discharge from the ground-water flow system in Hernando County occurs through
numerous coastal springs which are discussed in the next section. In addition to spring
discharge, however, areal discharge occurs in the coastal swamps and lowlands west of the 10
ft topographic contour which generally lies about 1 mi or so west of U.S. Highway 19. Areal
discharge rates reported for this region range from about 3 to 20 in/yr (Adams, 1985; Ryder,
1985; and Yobbi, 1989). These areal discharge rates incorporate discharge due to numerous
undocumented springs and seeps within the coastal lowland region.



4.3 Spring Discharge

The vast majority of ground water in Hernando County is discharged at springs along, and
slightly inland, of the coast (Figure 4.3). The largest and best known spring in the study area
is Weeki Wachee, which had an average discharge in 1988 of about 130 MGD (200 fte/s).





7n3 ""~'-7,
t :?1~


High rleherge pCOeniei somewh
aII c.he bul rll dl eonbllynrl *n4
breeched numer.o laneo by rnlief
| lnkhnele


Figure 4.3 General regions of ground-water recharge and discharge in Hernando County and

adjoining areas and locations of documented springs.


r-f V -


1 -'71 7_717











42.00




41.00




40.00


39.00




38.00




37.00


36.00




35.00




34.00


20.0


0
-o




"3


0
-0"


:3
10.0
3


CD
U)


S0.0
989


Observed 1988 hydrograph for the SWFWMD ROMP 105 deep well and monthly
recorded rainfall at Brooksville.


Figure 4.4








Substantial discharge also occurs at the group of springs in the vicinity of Chassahowitzka just
north of Hernando County in the extreme southwestern corner of Citrus County. The springs
in this region (Chassahowitzka, Crab Creek, Ruth, Potter, and several smaller springs) also had
a combined average discharge of about 130 MGD in 1988. Ryder (1985) states that the ground-
water flow system in the vicinity of large springs, such as Weeki Wachee and Chassahowitzka,
is probably relatively shallow and "consists of interconnected channels or conduits that collect
and feed water to (the) springs at high rates of flow."


A list of the known springs within the study area that could be identified and located, along with
their estimated discharge for May and September 1988 is provided in Table 4.1. Most of the
discharge values listed in Table 4.1 were obtained from Yobbi (1991), who conducted a series
of measurements at various times during 1988. To derive the May and September discharges
listed in Table 4.1, the available measurements listed in Yobbi (1991) were averaged for the
time period closest to May and September. For example, many of the September discharges
listed in Table 4.1 are actually averages of a series of spring discharges measured during
October of 1988. There is obviously some degree of uncertainty inherent in this approach, but
the approach is deemed reasonable since no program exists to measure spring discharge on a
regular basis for the springs within the study area.



4.4 Floridan Aquifer Parameters

4.4.1 Transmissivity

The observed transmissivity values of the Upper Floridan within Hernando County and adjoining
regions are generally quite high (in the hundreds of thousands of ft/d). Figure 4.5 illustrates
the transmissivity values obtained at various locations within the study area from aquifer tests
or flow net analysis. The highest reported transmissivity value of 2,100,000 ft2/d is at Weeki
Wachee Springs. A very high transmissivity of 2,000,000 ft2/d is also reported for the coastal
region of Citrus County. It is also evident from observation of the potentiometric surface maps








Table 4.1 Observed or estimated spring discharge for known springs within the study area.
Several small springs that could not be located are not included in the table.


Discharge (ft'/s) Regional Model Parameters
Spring Name
Sprig N e May Sept. Row Col Pool Elev. (ft)

Bobhill Springs 1.49 3.35 26 2 9.4
Magnolia Springs' 1.08 1.03 26 1 1.5
Gator Spring* 0.21 0.31 26 1 1.5
Boat Spring' 0.50 1.71 26 1 1.5
Weeki Wachee Springs 177.41 223.69 20 6 10.0c
Jenkins Creek Spring' 18.30 25.50 20 2 3.7
Hospital Spring 0.25^ 0.38 19 3 4.7
Salt Spring 30.54 30.30 18 4 4.7
Mud Spring2 28.25A 43.46 18 3 4.6
Ryle Creek Head Spring3 1.76A 2.70 9 4 2.4c
Ryle Creek Lower Spring4 8.10A 12.48 9 4 2.4c
Blue Run Head Spring5 7.81 7.70 9 5 2.5c
Beteejay Head Spring6 6.85 6.71 8 5 2.3c
Beteejay Lower Spring6 5.65 1.68 8 5 2.3c
Rita Maria Spring6 4.04 5.11 8 5 2.3c
Baird Creek Head Spring 1.37 6.20 7 6 1.7c
Chassahowitzka Springs 60.19A 92.60 7 6 1.7c
Crab Creek Spring 51.40 45.20 7 6 1.7
Lettuce Creek Spring 3.52 7.03 7 6 1.7
Salt Creek Head Spring 0.50 0.38 6 6 2.0
Ruth Spring 12.40 16.93 6 5 1.9
Potter Spring 8.80 25.50 6 5 1.9
Unnamed Spring No. 1* 5.00c 11.00c 26 1 1.0


4-10
















Total


1 Also
2 Also
3 Also
4 Also
5 Also
6 Also


518.65


699.35


called Unnamed Spring No. 5
called Mud River Spring
called Unnamed Spring No. 11
called Unnamed Spring No. 10
called Unnamed Spring No. 12
called Unnamed Spring No. 9 (3 springs together)


* Spring located with prescribed head node during numerical simulations
A May discharge value obtained by decreasing the September discharge
value by 35 percent after Yobbi (1991)
RDischarge value from Rosenau et al. (1977)
c Pool elevation from Yobbi (1989)


4-11


Discharge (ft'Is) Regional Model Parameters
Spring Name
nMay Sept. Row Col Pool Elev. (ft)

Blind Spring 50.70' 68.40 11 3 1.7
Unnamed Spring No. 7 32.50A 50.00' 11 3 1.7


















































Observed Upper Floridan aquifer transmissivities in Hernando County and
adjoining regions from various sources.


Figure 4.5


_ ___~_








for Hernando County (Figures 4.1 and 4.2) that the aquifer transmissivities are quite high, since
there are no cones of depression discernable.


In general, the highest transmissivities occur at major points of concentrated discharge, such as
springs, and transmissivities probably decrease with radial distance from the springs. This
conceptualization stems from the observation that although ground-water flow converges toward
the major springs, there is in general no significant increase in hydraulic gradients in the vicinity
of the springs. It can therefore be surmised that the ability of the aquifer to transmit water
increases with proximity to a spring.


In addition to the observed transmissivity values presented in Figure 4.5, there are a number of
numerical modeling studies during which transmissivities were estimated as a calibration
parameter. Most of these studies consisted of a regional-scale modeling effort; the best-known
study is probably that of Ryder (1985). Ryder used transmissivities ranging from 100,000 -
2,000,000 ft2/d in Hernando County. Throughout most of the county the model transmissivity
was 250,000 ft2/d or greater. The largest value of 2,000,000 ft/d was assigned to the model
cell that contained Weeki Wachee Springs.


Throughout much of the SWFWMD in the region south of the study area, it has been well
documented that there are two distinct production zones in the Upper Floridan (Wetterhall, 1964
and Ryder, 1985). The first production zone occurs near the bottom of the Suwannee
Limestone,. and the second occurs near the top of the Avon Park Formation. These two
production zones are separated by the relatively low-permeability Ocala Limestone. It is
generally believed, however, that there are not two clearly discernable production zones in the
Upper Floridan in Hernando County. This is because in Hernando County the Suwanee
Limestone is generally thin or absent, and the Ocala Limestone lies near the surface and, as a
unit, is more permeable then it is farther to the south. Ryder (1985) states

"North of Pasco County, where the Ocala Limestone is at or near land surface,
the permeability of this formation increases substantially, as indicated by its
cavernous nature, high recharge rates, and numerous large springs."


4-13








4.4.2 Effective Porosity


The effective porosity (hereafter referred to as porosity) of the aquifer is the ratio of the volume
of the voids in the aquifer through which water flows to the total volume of the aquifer. The
effective porosity is always (sometimes substantially) less than the total, or core, porosity. In
aquifers when ground-water flow occurs primarily through fractures and/or solution conduits,
which is the case with the Floridan aquifer, effective porosities obtained from field tests may
be quite low (less than 10 percent). For example, Burklew (1989) determined an Upper Floridan
porosity of 0.052 (5.2 percent) based upon the results of a tracer test conducted in Brevard
County. Most field tests, however, have a short duration and do not adequately measure the
long-term response of the aquifer. If the tests were conducted for a longer term and a greater
volume of aquifer were incorporated, the calculated porosities would generally approach that of
a typical porous media (about 0.1-0.25).


The Upper Floridan aquifer in Hernando County is generally unconfined, and the effective
porosity may therefore be approximated by the specific yield of the aquifer. However, there are
no known aquifer tests conducted in Hernando County for a duration sufficient to determine the
specific yield. There is some precedent, however, for aquifer porosities used during previous
studies. Ryder (1985) used a specific yield of 0.2 in Hernando County and adjoining
(unconfined) regions to conduct transient ground-water flow simulations. Mahon (1989) used
an effective porosity of 0.15 to calibrate a cross-sectional ground-water flow and solute transport
model in the southwestern portion of Hernando County. Fretwell (1985) estimated the specific
yield for Hernando County to be 0.15 based on findings in other counties by Stewart (1966) and
Hickey (1979). Although data concerning the effective porosity of the Upper Floridan aquifer
in Hernando County is quite limited, it would seem that a value of 0.15 is reasonable.


4-14








5 ANALYTICAL CAPTURE ZONE DELINEATIONS


5.1 Introduction and Background

In this chapter, the first and second level capture zone delineation methodologies and results are
discussed. The delineation methods applied at this stage of the study are generally less complex,
and easier to apply, than those described in Chapter 7 (Level m delineations). For reasons
presented below the Level II delineations are presented before the Level I delineations. Prior
to discussing the actual delineations, however, some important conceptual aspects behind the
approach are presented in the following paragraphs.


For reasons explained later in this section, the capture zones for most of the Hernando County
public supply wells are quite long and narrow; many have widths of only 150-500 ft. The
location, or orientation, of each capture zone depends solely upon the observed direction of
ambient ground-water flow. As explained in Section 4.1, however, the direction of ground-
water flow in the Upper Floridan is continually changing due to seasonal influences. The well
capture zones, therefore, are not stationary; they change as the regional orientation of the
potentiometric surface changes due to seasonal fluctuations. To adequately delineate and protect
the public supply well capture zones in Hernando County, it is essential that the seasonal
changes in the position of the capture zones be taken into consideration.


To achieve this goal, a pseudo-transient capture zone delineation approach was followed in
which the capture zone for each well was delineated for May and September 1988 conditions.
For example, the five-year capture zone for May was delineated assuming that the May 1988
potentiometric surface is invariant for five years, and likewise for September. In this way, the
fact that the well capture zones continually shift or "swing" between the May and September
conditions is duly accounted for. It should also be pointed out that one of the reasons that the
year 1988 was selected to perform the capture zone delineations is that, at most major public
supply wells in the county, the spread (difference in angles) between the May and September
directions of ground-water flow was largest of the years (1987-1989) examined. This approach








is reasonably conservative from a protection standpoint and incorporates a safety margin in the
WHPA delineations.



5.2 Level II Delineations Analytical Modeling

The second level capture zone delineations were performed using the MWCAP module of the
WHPA code (Blandford and Huyakorn, 1991). In this approach, ground-water flow pathlines
are delineated using ground-water flow velocities that are determined analytically. The major
assumptions and limitations involved in the application of this approach are as follows:

Ground-water flow is assumed to be two-dimensional in the areal
plane (horizontal flow).

The ground-water flow velocities are assumed to be at steady-state;
that is, they do not change in magnitude or direction with time.

The aquifer is assumed to be homogeneous and isotropic
(anisotropy in the horizontal direction could be handled, however,
if certain critical parameters could be identified)

The aquifer is assumed to be confined.

Local areal recharge within the cone of depression caused by a
given well is neglected.

Although the Floridan aquifer in Hernando County is generally unconfined, the ratio of
drawdowns due to pumping to the initial redevelopmentn) saturated thickness of the aquifer is
small (less than 10 percent), and therefore the confined aquifer assumption is valid. As
explained in the previous section, capture zone delineations were performed for May and
September of 1988, and therefore the delineation approach may be considered pseudo-transient.


To delineate the capture zone for each well, the following physical input parameters were
required: well location, well pumping rate, transmissivity, thickness and porosity of the aquifer,
ambient hydraulic gradient and the angle of ambient flow. Since the MWCAP module does not
account for well interference (which is a reasonable approach for the Hernando County wells








due to their limited cones of depression), the required input parameters had to be identified and
entered for each well. The angle of ambient flow and the ambient hydraulic gradient were
measured using the May and September 1988 potentiometric surfaces. The assumption that the
observed potentiometric surface for May and September of 1988 is representative of the ambient,
or redevelopment, surface is reasonable in Hernando County because the ground-water flow
system has not been substantially affected by pumping. The location of each well was
determined by the District using a global positioning satellite (GPS) system, and was supplied
in Universal Transverse Mercator (UTM) coordinates. The withdrawal rates for each well were
metered during 1988 and were supplied by the District. For the May delineations, the average
pumping rate for April and May was used, and for the September delineations, the average
pumping rates for August and September were used. An effective porosity value of 0.15 (fifteen
percent) was used for each delineation.


The aquifer transmissivity and thickness values at each well were determined as follows. For
wells that had a documented transmissivity value from an aquifer test available, the reported
transmissivity was used. At wells that did not have a documented transmissivity value, a
transmissivity value was derived based upon those reported in Ryder (1985). Since the
transmissivity values of Ryder are representative of the entire thickness of the Upper Floridan,
they were divided by the aquifer thickness (which is generally 700-800 ft in Hernando County,
see Chapter 3) to obtain an estimate of hydraulic conductivity. The hydraulic conductivity was
then multiplied by the aquifer thickness selected for each well (which is less than the thickness
of the Upper Floridan since the wells are partially penetrating) to yield a transmissivity value
that was input into the WHPA code. The "effective" aquifer thickness was determined as the
casing depth subtracted from the well depth. Since the wells in Hernando County do not
penetrate the entire thickness of the Upper Floridan, the aquifer thickness derived in this manner
will yield more meaningful simulation results than using the entire aquifer thickness would. The
well-specific input parameters for each of the public supply wells in Hernando County for which
Level II capture zone delineations were performed are listed in Table 5.1.










* 4 @0 @0 q @ @ 0 *- @ o o @ -H -
II. I S I !

000 0000 0 0 00



R. i en
a r4 00 fn An so >
a S 8 x x x x x x x x x x x x x x x x x x x~



S,| 0 0 C> 9 !$ 2 0 8 3 C, s is)
S G R I SS .
aI "" TTTTTTTTTTT TT T




V4 N V %n %n 4 %n %n IV V W4 No V V @ @ r4
- R "'- -t "'" ".S------- -- S 8 "1
*% 4 @0 ar @3

@0 t' S @0 -
No e 0 s 0 e 'n s a



e S~- 0@ 0 00 % eS @ 0SS S 1 0 1 "
%n -, I-;A A 1 1 ,

-~ 0i 0 0 0 0 0 0 0n 0 p
x x x x x x x x x x x x x x x x




01
Sr-: 0r 1'- @0 eq @0 eq eq n lr Os eq r- 4 t- e




C44

% o e 4 om e ov o- oe v W W oa % e ; m





^ ~~ ~%~ "i i^Tl i m 'Z 'T TT 's
m Go 00 en 0 0


G! m e en
0 0 00 0 0 0 0 0' 0 @0 @0 00 @ 0 0-0

as q T 1 0
% n 0e %n e v @ @ v 0 0 @ 0
'0 @0 4 -~ '~ 0% 0 oeq olcc @0 ~ @0 H e




G o en BO 0 o m e
":St m n ~ #



la eq '0^n n e n 4 4 @ @ Q






MIL ~ l l~ r% -In %ol r- 1
%3on m ol
oe Go Go 00 m0 C04 N0 11, Go C4 @0 @ @ 0 0 @




(1 N G G o 0 %0 a,
Noee~@ N0 -n 0 11-
"I"Ien A "IMM"ImP" ~ t







cC s Y! L OD Y Y, r m vr JN r- 0
4Ae ~ ; : 0 00 0 0








-, *4 @0 4 0 -; 0 1; '0 Ic V0 ~ 0 0 0 V2 19 %0

r 'C
eq 0 4 @0 @ 4 @0 4











0 0 0 8 8 '0 0 8 0 8 0 0 0 0 0 0 0 8 0 0

g SS S: 8,L S So So 2 St SI S S22
u 1 u 1 u 1 1 1 u u 1 1 1 1
Bs0 '0 '0 @0 4' 4 4 @ @








m m m 9
4 4'ugu~y~ @0-(4e 03. 04@


0 u uI u @jI ui o r E0v
V a 0

$16 Ii s I *i Iee8 I, ; 16 1 1 1
00 00 00 o66 666 U 1 11
C. C.)l8 C)~s8I I I I C.)8 8 81 1 C.C) C. C) C) C. C) C) .


___ I l











I n n n n R q qTT O
R RC R



x x x x x x x x x x x x x x x
__ > n 'i fi ti ^ m O f r o vc
S o 0 0 0 0 0 0 0 0 0 0 0 0

SfR fn 4n oi ll P' O i Nt S i It
*en m II I V I
01^ S- ^ 0% SC m 0% '4 (
c oo C4 %n C O Be Wo Vn '4-

-40 000 0 000 00 000en
C4 e









x x x x x x x x x x x x x x x
12 C 000 0 0 00
l o b 44 b 44 fl '- '4 '0


C4 % -c




0 '0 SC4 S A S ?
Q t Io -, f

v ,v z8 88 8 W







- x x x x x x x x x x x x x x x
SN 44 N V 0 ^ 4 0 '
0'' '0 b SC'' b



""^~ < '~^' 0 '4 0 "^ 3"" '" ~ '0 0%" V~" ~
a Soo '0 0% 0% 6




Oi 'i V r'. -o i w ^ -, W% V "V 'f ? Vi V
co 10 ci







i e 0 W 0




ni SA Go % I C4 enD ; 0 0 S C -
C44' V ~4 4 C14 ('4 %0 eft
'0 Is C4 0 r- C~4 0 '
SC d I Is 0% (o '4 r' ei o e




00 ^"* 0 Is a, C4 w0

vi w, C0 4 r- I 42, q4'' V v~
s =4






~~ii x 'l4* 4.%x Ig'"T-5T
*'n en me e4no o r ^ t' ^



e>n m m0 r l C SC
en '4 e (n en en en m' Af e4n-4"


'0 -g -B
s s^ Sz S"^ 9 s
Xg lo E; 1" tl l S ^ S ?S S ~ ''< o o' o 'b 'o

la o" r r^ ^ f~ r s r^ ^ f
fP c o o wi v r Wi v i -fr in
5 ~ ~~ ~ cc- w n w f^ d t ^ o ol^ t
E0 N 7~ ""
3~p _-,P o^-SN"^ e


I-


d r
0 0
, I


0


.0








Note that in Table 5.1 several groups of wells are "lumped" together to perform the capture
zone analysis. The wells grouped together to form an "equivalent" well lie very close to one
another; in these situations it is reasonable to consider a composite capture zone delineated for
all of the wells. For the equivalent well delineations, the individual well locations were
averaged to obtain an equivalent well location, and the pumping rates of the individual wells
were summed to obtain the equivalent well pumping rate. The aquifer thickness for each
equivalent well is based upon the maximum open-hole interval determined using the reported
casing and well depths for each well (i.e., the minimum casing depth was subtracted from the
maximum well depth). The transmissivity values were subsequently determined by multiplying
the aquifer thickness by the hydraulic conductivity.


The ten-year capture zones for May, September and average 1988 conditions delineated using
the analytical modeling approach (Level II) are presented in Figures 5.1-5.3, respectively. The
orientation of each capture zone is variable between the May, September and average 1988
delineations since the angle of ambient flow is generally different for each case. The capture
zone widths may be variable between the time periods for a given well due to variable pumping
rates and/or different values of the ambient hydraulic gradient; the capture zone widths will
increase with higher pumping and decreasing values of the ambient hydraulic gradient.


The most notable feature of the majority of capture zones is that they are quite long and narrow.
Physically, this capture zone shape is due to two (in some areas three) reasons:

1) The transmissivity of the Floridan aquifer in Hernando County is
generally high (on the order of hundreds of thousands ft/d)

2) The withdrawal rates of the major public supply wells in Hernando
County are generally not large with respect to the transmissivity of
the system

3) In some regions of the county, particularly in the vicinity of the
Pasco high, ambient hydraulic gradients are relatively steep

The delineated capture zones that are significantly wider than the majority or tend more towards
a circular shape are wells 2 and 10 in the Hernando County West System, Hemando County -


____________________-------* -9-----------






r7--, 777
I i~l~.i". 7,


Figure 5.1. Ten-year Level II (WHPA code) capture zones for May 1988 conditions.


-
I i -;I










































Figure 5.2. Ten-year Level II (WHPA code) capture zones for September 1988 conditions.





















CITRUS CO
*. HERNANDO CO :




0
x


L-J
o

1b 0 PiPBROOKSVILLE
'0 0


Figure 5.3. Ten-year Level II (WHPA code) capture zones for average 1988 conditions.








Dogwood Estates well 1, and Brooksville well 1. Wider capture zones are predicted at these
wells due to observed transmissivity values that are significantly lower (in part due to limited
production zone thickness) than those of the surrounding wells.



5.3 Level I Delineations Flow-Net Analysis

The delineation of capture zones using flow-net analysis involves the delineation of ground-water
flow pathlines based upon the configuration of an existing potentiometric surface map. If the
Floridan aquifer is assumed to be isotropic (in the horizontal plane), and ground-water flow
within the aquifer is predominantly two-dimensional, then the limiting pathlines that form the
zone of contribution to a well may be delineated by drawing lines that intersect the
potentiometric surface contours at right angles. The upgradient end of the zone of contribution
may be terminated based upon some time of travel criterion (e.g. 2 yrs, 5 yrs, etc.). Note that
seepage velocities of ground-water flow along individual pathlines may be computed by a simple
application of Darcy's law.


The application of this method of analysis to the Hernando County wells is problematic because
the observed potentiometric surfaces do not indicate any cones of depression in the vicinity of
the pumping wells. Flow paths in the vicinity of the pumping wells, therefore, cannot be
delineated based solely upon the observed potentiometric surface. To circumvent this difficulty,
the Level II delineations were used to obtain the capture zones in the vicinity of the wells, and
flow path delineation based upon the observed potentiometric surfaces were conducted only for
those regions of Level II capture zones where the direction of regional ground-water flow was
significantly different from that measured at the well.


The Level I delineation results are presented in Figures 5.4 and 5.5. In general, the Level I
capture zones did not have to be adjusted until the five-year or greater time of travel point was
reached. The most significant differences between the Level I and Level II delineations are for
some of the Spring Hill Utilities wells in southwestern Hernando County and for the Hernando


5-10























CITRUS CO
P HERNANDO CO






o
O
w


0

DROOKSVILLE
r '3
I-


Figure 5.4. Ten-year Level I (flow-net analysis) capture zones for May 1988 conditions.


i--7;F




















































Figure 5.5.


Ten-year Level I (flow-net analysis) capture zones for September 1988 conditions.


------ Il~i~-~ I-TTTa








County Hill N'Dale wells near the intersection of Interstate 75 and State Highway 50 (these
wells are in close proximity to the northern tip of the Pasco potentiometric high).


5-13








6 REGIONAL GROUND-WATER FLOW MODELING


6.1 Conceptual Model and Modeling Assumptions

For the purposes of regional ground-water flow modeling, the Floridan aquifer within the model
domain (Figure 6.1) was conceptualized as a single, unconfined aquifer layer. This approach
is consistent with that followed in numerous other modeling studies such as Ryder (1985).
Although the Upper Floridan was conceptualized as an unconfined aquifer, ground-water flow
within the aquifer was simulated using the MODFLOW code confined aquifer option. This
approach is valid since fluctuations in the Upper Floridan potentiometric surface due to pumping,
and/or seasonal variations in recharge and discharge, are small relative to the entire saturated
thickness of the aquifer.


The regional model was calibrated to two potentiometric surface maps; May 1988 (Barr, 1988)
and September 1988 (Lewelling, 1988). Some model input parameters, such as transmissivity,
remained unchanged between the calibration periods. Other parameters, such as pumping rates
and prescribed recharge and discharge, differ between the calibration periods. It was necessary
to calibrate the regional model to May and September conditions so the subregional models could
be developed, and capture zones could subsequently be delineated for, each of these periods as
were the Level I and II delineations. All model simulations were steady state. Areal recharge
to, and discharge from, the Upper Floridan aquifer was prescribed throughout the model
domain. The prescribed recharge and discharge values are conceptualized as net fluxes that
implicitly incorporate complex processes such as losses due to evapotranspiration.


Ground-water efflux from the Upper Floridan due to coastal springs was simulated using a head-
dependent flux boundary condition at the model cells that contained springs (Figure 6.1). Using
this modeling approach, the discharge from a spring (Q) is calculated by the model using





S7 --- --- --

















0 .-








I *i
X




"20
Q\ 0@








h)0


Figure 6.1 Regional model grid and associated boundary conditions.


i, i 'A-.I


--- --" -----~ ;-----,
'"









Q, = C,(h,-h,


where h, is the spring pool elevation, h, is the head in the Upper Floridan simulated by the
model, and C, is an empirical conductance term unique to each spring and each modeling
configuration. The DRAIN package of the MODFLOW code was used to implement this
modeling conceptualization. For input into this package, the spring pool elevation is entered as
the head in the drain, and the spring conductance is entered as the conductance of the interface
between the cell and the drain. If the head in the Upper Floridan drops below or is equal to that
of the spring pool elevation (h1, 5 h, the resulting efflux (Q) is set to zero. Since the model
calculates the discharge at springs based upon the above equation, the observed spring discharges
are critical calibration targets for the model.


The exchange of water between the Upper Floridan and the Withlacoochee and Little
Withlacoochee Rivers was also simulated using a head-dependent flux boundary condition at the
roughly appropriate model cells (Figure 6.1). Aquifer discharge to, or recharge from, either
of these streams was calculated using the relation:

-K' (h, h) (6.2)
b'

where q, is the vertical Darcy flux entering or exiting the Upper Floridan, h, is the water-surface
elevation in the river, h, is the hydraulic head in the Upper Floridan. Since the Withlacoochee
River flows in a limestone channel over much of its course (Camp and Barcelo, 1988), the
conductance term K'/b' can be approximated as the vertical hydraulic conductivity of the Upper
Floridan beneath the river divided by one-half of the Upper Floridan thickness. The initial
approximation of b' is based upon the single-layer modeling approach, in which a model node
is conceptualized as being at the center of a grid block. Note that if h, is less than hk, the q,
term is negative and water discharges, rather than recharges, the system. The term hu is
calculated by the flow model, while the remaining terms on the right-hand-side of equation 6.2
(h,, K', b') are input parameters. Each of the four terms used to calculate q, may exhibit


(6.1)








substantial spatial variability. The MODFLOW code RIVER package was used to implement
the variable flux recharge condition described above.


All ground-water pumping values used in the regional model were obtained from the District.
V Average pumping rates for April and May were used for the May simulations, and average
pumping rates for August and September were used for the September simulations. Some water
use categories, such as public supply, major industrial users and agriculture, are metered and
the pumping estimates should be relatively accurate. Agricultural water use estimates were
: based on metered pumping data obtained from the District's Agricultural Irrigation Monitoring
(AIM) Program. Non-metered, non-agricultural water use estimates were based on permitted
pumping rates. There may be a substantial degree of uncertainty in the non-metered pumpage
estimates. The utilization of 2-month average pumping estimates is consistent with the Level
II (analytical modeling) delineations. It is believed that the average pumping for two months
may provide a better estimate of general levels of dry season and wet season pumping then
would selecting the pumping observations and estimates for a given month. This approach is
also consistent with the physical conceptualization that the potentiometric surface at any given
time is a function of various hydrologic conditions over some period (several weeks to several
months) prior to that time.



S6.2 Grid Design

The regional model grid is a single layer block-centered finite difference grid representing the
Upper Floridan aquifer; it encompasses all of Hernando County and adjoining portions of Pasco,
I Citrus and Sumter Counties (Figure 6.1). The grid consists of 33 rows and 43 columns. Thus,
the total number of nodes in the grid is 1,419, of which only 973 were active due to the
V configuration of the boundary conditions. Note that in Figure 6.1 the first two rows and the last
two columns of the model domain are not illustrated; these rows and columns were used during
V! the initial modeling runs but were subsequently deleted from the active model domain when the
boundary conditions were finalized. The grid is uniform; each cell is 1 mi square. During the
initial stages of the regional modeling a sensitivity run was conducted in which the grid density

6-4
L:








was doubled (each cell 0.5 mi square); there were no substantial differences between the
simulated head fields obtained using the two grids.


The boundaries of the model domain were selected based upon two criteria:

1) The regional model had to incorporate all of Hernando County, but
only some reasonable portion of adjoining counties as technical
considerations warranted.

2) The location of the model boundary should coincide with natural
boundary conditions, such as regional ground-water flow pathlines,
as much as possible.

The northern and southern model boundaries generally coincide with regional ground-water flow
pathlines that encompass Hernando County. The western portion of the northern model
boundary that is horizontal was selected to lie above the Chassahowitzka spring group, but below
the Homosassa spring group in Citrus County farther to the north. It was unnecessary for the
purposes of this study to simulate the ground-water flow system in the vicinity of Homosassa
springs. The eastern portion of the model domain was terminated a couple of miles outside
Hernando County; the natural extension of this boundary would continue southeast to the Green
Swamp potentiometric high in Polk County. For the purposes of this study, it was unwarranted
and unnecessary to extend the model domain any farther to the southeast. The western model
boundary was selected to lie at or near the coastline. Most of this boundary is no-flow, since
the fresh ground-water flow system terminates (pinches out) in the vicinity of the coastline. In
the southwestern corer of the model domain, the western boundary is prescribed head since it
is located within the fresh ground-water flow system.



6.3 Model Input Data

A list of the required regional model input parameters is provided below:

Boundary condition type (prescribed head or no-flow) and location
(Figure 6.1)

Upper Floridan transmissivity


I









* Prescribed recharge to, or discharge from, the Upper Floridan


Spring pool elevations and conductance terms for input to the
MODFLOW DRAIN package

Water level elevations and conductance terms for the
Withlacoochee and Little Withlacoochee Rivers for input into
MODFLOW RIVER package

Well pumping rates

Some input parameters, such as the prescribed boundary head values, prescribed recharge, and
water level elevations in the rivers were calibration-period specific (the values of these
parameters changed between the May and September calibration periods).



6.4 Regional Model Calibration

6.4.1 Calibration Procedure

Model calibration, also referred to as history matching, is the general procedure of adjusting
model input parameters within reasonable ranges until the model output (in this case hydraulic
head in the Upper Floridan and simulated spring flows) resembles conditions observed in the
field within some prescribed error tolerance. In this study, the model calibration parameters
were transmissivity of the Upper Floridan, areal recharge to the Upper Floridan (May and
September), spring conductances and pool elevations, and prescribed river heads and river bed
conductance terms. The major calibration targets were observed hydraulic head in the Upper
Floridan and observed spring flows. To a lesser extent ground-water discharge to the
Withlacoochee River was also used as a calibration parameter.



6.4.2 Calibration Results

Figures 6.2-6.5 present the regional model calibration results with respect to hydraulic head.
Figures 6.2 and 6.3 illustrate the simulated May 1988 potentiometric surface and the difference


I



















































Figure 6.2 Simulated May 1988 Upper Floridan potentiometric surface.


__~___





















































Difference between the observed and simulated May 1988 Upper Flofidan
potentiometric surfaces.


Figure 6.3


i


~--~h
(


r--








































Figure 6.4 Simulated September 1988 Upper Floridan potentiometric surface.
















31 R* i U C" 0
ERNANDO CO *








S(


Figure 6.5


Difference between the observed and simulated September 1988 Upper Floridan
potentiometric surfaces.


-----7 *---


'~Tm -p- --?
C `








between the observed and simulated May 1988 potentiometric surfaces, respectively. In terms
of hydraulic heads, the May calibration is quite good. Differences between the observed and
simulated head values are generally 2 ft or less. There are two relatively small regions, one
northwest of Brooksville and the other centered on the Little Withlacoochee River in eastern
Hernando County, in which the head differences are 4 ft or greater. There are, however, no
major public supply wells within these regions.


Figures 6.4 and 6.5 illustrate the simulated September 1988 potentiometric surface and the
difference between the observed and simulated September 1988 potentiometric surfaces,
respectively. The September calibration is also quite reasonable in terms of hydraulic heads.
Differences between the observed and the simulated head values are less than 2 ft throughout
most of the model domain. There is one region in south-central Hernando County in which head
differences locally exceed 6 ft. A portion of this region is in the vicinity of the Pasco
potentiometric high, which was a difficult feature to simulate during this study. In this region,
the observed hydraulic gradients are very steep, and a slight misplacement of a contour line on
the observed potentiometric surface map, or a small shift in the simulated hydraulic head values,
may easily cause discrepancies of 4 ft or greater.


There is another, larger region in which head differences are 4 ft or greater; this zone is in the
vicinity of the Withlacoochee River in Hernando County south of the confluence of the
Withlacoochee and Little Withlacoochee Rivers. The simulated heads in this region are too
high; they could be reduced (and were reduced during some calibration runs) if simulated
ground-water discharge to the Withlacoochee River could be increased. Discharge to the river
in this region could be increased by increasing the river bed conductance, or by decreasing the
prescribed water level elevation in the river, but neither of these alternatives were deemed
reasonable given the present conceptualization of the system.


The model calibration results in terms of simulated versus observed spring discharge is presented
in Table 6.1. The percent discrepancy between the observed and simulated discharge at the
major springs is generally 10-15 percent or less, which is an acceptable match. On a model-


6-11








Table 6.1 Observed and simulated spring flows
September calibration periods.


(ft/s) for the regional model May and


6-12


May 1988 Calibration Period September 1988 Calibration Period
Row Col
Observed Simulated Percent Observed Simulated Percent
Discharge Discharge Discrepancy Discharge Discharge Discrepancy
26 2 1.49 2.08 40.0 3.35 3.06 -8.7
20 2 18.30 20.10 9.8 25.50 23.06 -9.6
20 6 177.41 157.54 -11.2 223.69 255.92 14.4
19 3 0.25 0.29 16.0 0.38 0.33 -13.2
18 4 30.54 29.46 -3.5 30.30 33.92 11.9
18 3 28.25 32.04 13.4 43.46 36.70 -15.6
11 3 83.20 91.84 10.4 118.40 109.32 -7.7
9 4 9.86 10.85 10.0 15.18 15.22 0.3
9 5 7.81 5.86 -27.3 7.70 8.20 6.5
8 5 16.54 12.26 -25.9 13.50 18.41 36.4
7 6 116.48 112.84 -3.1 151.03 148.50 -1.7
6 6 0.50 0.35 -30.0 0.38 0.48 26.3
6 5 21.20 29.55 39.4 42.43 33.13 -21.9
Totals 511.83 504.88 -1.4 675.30 686.25 1.6









wide basis, the percent discrepancy is less than 2 percent for each calibration period. Note that
if a given model cell contained multiple springs, the springs were simulated as a lumped, or
equivalent, spring for which a total discharge equal to the sum of the individual spring discharge
was set as a calibration target.


The calibrated Upper Floridan transmissivity values are presented in Figure 6.6. Throughout
the model domain the calibrated transmissivity values match reasonably well with those observed
at individual wells or those obtained through flow-net analysis. The model transmissivity values
increase with proximity to the major springs (such as Weeki Wachee) or spring groups.
Transmissivity values also generally increase moving north across the model domain. Both of
these transmissivity trends have been documented by Ryder (1985) and other authors.


Transmissivities in the southern portion of the model domain in Pasco County and in the vicinity
of the Pasco potentiometric surface high are substantially lower than elsewhere in the system.
Transmissivity values throughout much of Pasco County less than 100,000 ff/d have been well
documented (Hutchinson, 1984). The lowest model transmissivity is 20,000 ft2/d and occurs in
the vicinity of Pasco potentiometric high. This low transmissivity is not substantially out-of-line
with observed transmissivities in adjoining regions of Pasco County, which are in the 28,000-
48,000 ft/d range. Low transmissivities were required throughout the northern portion of the
Pasco potentiometric high to adequately match the pronounced mound shape in that region.


Figures 6.7 and 6.8 are the prescribed recharge rates applied to the Upper Floridan for the May
and September calibration periods, respectively. The recharge rates illustrated on the figures
are in in/yr, but the rates actually input to the model were in ft/d. Although the recharge rates
are presented in in/yr, they are not annual rates of recharge; in reality the recharge that occurs
throughout the year may vary substantially. The highest rates of recharge are along the
Brooksville Ridge and the sandy, rolling hills that lie to the west of the ridge but to the east of
State Highway 19. A very high recharge rate of 60 in/yr was prescribed during the September
calibration period for the model cells that contain a large quarry northwest of Brooksville. It
has been reported that during a rainfall event, most of the surface runoff at the bottom of the


6-13












7 -o ,o I- -1












S0- 030 0
LL -
2.


S0 40-80
0200
*40-80

S 0-200


Figure 6.6 Calibrated regional model transmissivities in ft2/d times 10,000.


7-7 I8-~- --- I







































Calibrated recharge to, and discharge from, the Upper Floridan for the May 1988
calibration period.


Figure 6.7










































Calibrated recharge to, and discharge from, the Upper Floridan for the September
1988 calibration period.


Figure 6.8








quarry seeps into fractures in the limestone surface almost immediately (John Parker of
SWFWMD, personal communication, 1992). The lowest recharge rates are in the vicinity of
the Withlacoochee and Little Withlacoochee Rivers in the eastern portion of the County.
Discharge occurs within the coastal swamps region. In general, the May recharge rates are
about 30 percent less than the September recharge rate.


For the purposes of this study it was necessary to perform May and September (or dry and wet
period) calibrations. However, most previous modeling conducted for central Florida has
focused on average redevelopment conditions or average hydrologic conditions for a given year.
When this modeling approach is used, the averaged potentiometric surface is assumed to be the
average of the May and September surfaces, and the model parameters (such as recharge) are
assumed to be average annual values.


Following this conceptualization, the May and September recharge values obtained during this
study, if averaged together, may be compared to those obtained from previous modeling studies.
If this approach is followed, the highest recharge rate in the study area is about 33 in/yr in the
region west of the Brooksville Ridge but east of State Highway 19. This recharge rate is
reasonable, since previous studies have indicated recharge as high as 30 in/yr (Yobbi, 1989) for
average redevelopment conditions. During 1988, rainfall was significantly above average
(Section 4.2), and consequently the recharge that occurred during 1988 should be greater than
average. Following the above analysis, it is evident that the prescribed recharge rates within
other regions of the model domain are in general agreement with those reported in previous
studies.


The modeling approach of conducting steady-state model calibrations to the May and September
periods is a rough approximation for simulating the physical system. In reality, the ground-
water flow system exists in a transient state driven by seasonal climatic conditions, and storage
effects (neglected in steady-state analysis) may have a significant impact upon the observed
potentiometric surface. From a practical perspective, it would be difficult to model the ground-


6-17








water system in a transient mode due to the high degree of variation in the rainfall and
evapotranspiration rates that occur throughout the year.


During this study, the prescribed recharge and discharge rates derived for the May and
September calibration periods were conceptualized as average rates representative of a.period
of several months (2-3) prior to the potentiometric surface measurements conducted for each
period. It should be noted, however, that the effective recharge rates developed for the May
calibration period incorporate, to an unknown degree, storage effects not explicitly simulated.
In other words, the May potentiometric surface is a function of recharge to the aquifer due to
rainfall that occurred previously (probably 2 weeks-2 months), and the release of water from
storage due to the falling water table. The calibrated recharge rate obtained during the May
calibration is reflective of both of these processes. The corollary is true for the September
calibrated recharge rates, only during this calibration period increased storage of water is
generally occurring.



6.4.3 Mass Balance

The regional model mass balance for the May and September calibration periods is presented
in Table 6.2. The mass balance summary confirms the conceptual hydrogeologic framework
outlined in Chapters 3 and 4. Most inflow to the ground-water system is due to areal ground-
water recharge within the study area due to rainfall and leakage from surface water bodies
(lakes). Ground-water discharge occurs primarily as spring flow, discharge to the
Withlacoochee River, and as ground-water outflow along the western and northern model
boundaries.



6.5 Regional Model Sensitivity Analysis

A series of four sensitivity runs were conducted to observe the changes in the simulated
potentiometric surface due to changes in the model input parameters. Each of the sensitivity
runs was conducted for the September 1988 simulation period; similar results would be obtained


6-18








Table 6.2 Regional model mass balance for May and September 1988 calibration periods.


May Calibration Period September Calibration Period

Source Percent of Total Percent of Total
Flux (fP/d) Inflow/Outflow Flux (ft/d) Inflow/Outflow
Inflows
Prescribed Head 2.3734 x 107 26.0 2.1465 x 10 18.6
Boundary Conditions
Areal Recharge 6.6325 x 10 72.7 9.385 x 107 81.3
River Leakage 1.2235 x 106 1.3 1.4662 x 10 0.1
Outflows
Prescribed Head 2.4869 x 107 27.2 3.1317 x107 27.1
Boundary Conditions
Pumping Wells 1.1411 x 107 12.5 8.0816x 106 7.0
Springs 4.3621 x107 47.8 5.9291 x10 51.4
Areal Discharge 2.0566x 106 2.3 5.0083 x 106 4.3
River Leakage 9.3248 x106 10.2 1.1764 x107 10.2


6-19








for the May 1988 simulation period. The sensitivity simulation results, therefore, should be
compared with Figure 6.4 of Section 6.4.2. Each sensitivity run involved changes in the two
key model input parameters that more than any other parameter affect the configuration of the
simulated potentiometric surfaces; Upper Floridan transmissivity and prescribed areal recharge.


Figures 6.9 and 6.10 illustrate the effects of a two-fold increase, and a 50 percent reduction
respectively in the Upper Floridan transmissivity. Increasing the transmissivity makes it easier
for water to flow through the aquifer, and therefore the regional hydraulic heads and hydraulic
gradients decline substantially throughout the model domain. Simulated heads decreased by as
much as 15 ft at the crest of the Pasco potentiometric high.


Decreasing the Upper Floridan transmissivity by half decreases the aquifer's ability to transmit
water, and therefore simulated hydraulic heads and hydraulic gradients increase significantly
throughout the model domain. Simulated heads at the Pasco high increased by as much as 45
ft relative to the September 1988 simulation results.


Figures 6.11 and 6.12 illustrate the effects of a two-fold increase, and a 50 percent reduction
respectively in prescribed recharge to the Upper Floridan. Note that Figure 6.11 (increase in
recharge) is very similar to Figure 6.10 (decrease in transmissivity), and Figure 6.12 (decrease
in recharge) is very similar to Figure 6.9 (increase in transmissivity). This result is not
surprising since it is well documented that for steady-state conditions the effects of transmissivity
and areal recharge on the ground-water flow system are nearly equal and opposite. This is
because if recharge is increased, more water must be passed through the aquifer system and
hydraulic gradients must increase if transmissivity is not changed. However, more water can
be passed through the system without changing the hydraulic gradient if transmissivity is
increased. Stated another way, a given hydraulic gradient may be maintained by decreasing
recharge and transmissivity in conjunction, or by increasing recharge and transmissivity in
conjunction.


6-20










































Simulated potentiometric surface for first model sensitivity run in which Upper
Floridan transmissivity was increased two-fold.


Figure 6.9


I I-III I I IIII _~---p-----^-~,~,-~.---~-~ rl^-r~-.~--_li~-i~*_-C-~-~^~--__





* I 7 *.


Figure 6.10 Simulated potentiometric surface for second model sensitivity run in which Upper
Floridan transmissivity was decreased 50 percent.









0 5 mi


-50- Simulated potentlometric
surface in feet above NGVD
for sensitivity run 3


Figure 6.11 Simulated potentiometric surface for third model sensitivity run in which
prescribed recharge to the Upper Floridan was increased two-fold.




777 41-7-7 r
A.,7


Figure 6.12


Simulated potentiometric surface for fourth model sensitivity run in which
prescribed recharge to the Upper Floridan was decreased by 50 percent.








The above observation raises the question of how can a reasonable ratio between areal recharge
and Upper Floridan transmissivity be identified during the modeling process? For the Hernando
County model, the calibrated values of areal recharge and Upper Floridan transmissivity were
identified by matching to the observed (or estimated) May and September 1988 spring discharges
and by keeping model transmissivities and recharge rates within reasonable bounds as determined
by hydrologic analysis and literature review.


6-25








7 SUBREGIONAL GROUNDWATER FLOW MODELING AND
CAPTURE ZONE DELINEATION



7.1 Subregional Ground-Water Flow Modeling

It was known a priori that the regional model discretization was too coarse to obtain accurate
capture zone delineations. The scope of this project, therefore, included the construction of
subregional models, based upon the regional model results, to obtain accurate Level II capture
zone delineations. Four subregional models were constructed; the identification number and
location of each subdomain is illustrated in Figure 7.1.


Within each of the four subdomains, a subregional model with a much finer discretization than
the regional model was constructed. A summary of each of the subregional model grids is
presented in Table 7.1. In addition to constructing a finer discretization of the areal plane within
each subregion, the thickness of the Upper Floridan was divided in 5 layers of equal thickness.
This vertical discretization was conducted to examine the effects of partially penetrating wells.


The aquifer parameters and boundary conditions for each subregion are based upon the regional
model simulations. This approach maintains the mass-balance of ground water within each
subregion. All subregional model boundary conditions are prescribed head for the relevant
calibration period (May or September, 1988). Areal recharge is applied to the top model layer,
as is the Withlacoochee River boundary condition for Subregion 3. Pumping from each well
was divided equally between the top two model layers, as an analysis of Hernando County well
depths revealed that most wells are completed within the upper half of the Floridan aquifer.


Weeki Wachee spring is incorporated in the northwest corner of Subregion 1. When this
subregion was first constructed, the simulated potentiometric surface did not match well with the
regional simulated surfaces in the vicinity of the spring. The subregional model did not permit
a sufficient efflux of water from the spring due to the finer discretization of the Upper Floridan
in both the areal and vertical dimensions. To circumvent this problem and maintain a good mass




------_ .---------------^-~ .~~7~i '-7i--




5. O 5 mi

2 Subreglonal model
70- 2 -domain Identification
S BUSHNELL number

|'.| E :EEE 4 E

-- --- -- I 111 -- 42 -
@

A e*
'C, .: .I BROOKSVILLE --

SWEEKI 3
WACHEE o
t^ t I 1 r'


Figure 7.1 Location and identification number of subregional model domains.








Table 7.1 Summary of subdomain modeling grids.


Grid Parameter Subregional Modeling Domain Number
1 2 3 4
Number of Columns 111 71 47 32
Number of Rows 117 74 27 25
Minimum Column Spacing (ft) 200 386 383 440
Maximum Column Spacing (ft) 1,268 1,441 1,522 1,320
Minimum Row Spacing (ft) 306 356 450 704
Maximum Row Spacing (ft) 920 1,320 1,405 2,075








balance, in Subregion 1 Weeki Wachee springs was incorporated into the model as a fully
penetrating pumping well rather than as a head-dependent flux boundary condition. The
pumping rate of the well was set equal to the simulated spring flux obtained from the regional
model. This approach yielded a much better agreement between the Subregion 1 and regional
model simulated head fields, although a small cone of depression about the spring exists in the
Subregion 1 simulation results which was not obtained in the regional model. An alternative
approach would have been to recalibrate the spring conductance term and the leakances between
model layers in the vicinity of the spring. This approach could prove timely and cost-
consuming, however, and it was not deemed necessary in terms of the major objectives of this
study.


For each of the subregions, the simulated May and September 1988 potentiometric surfaces were
compared with that obtained from the regional model. For each of the subregions the match is
very good. The subregional model potentiometric surfaces are not presented in this report
because, for all practical purposes, they are the same as those presented in Chapter 6.



7.2 Level I WHPA Delineations

The ten-year Level III capture zones were delineated based upon the subregional modeling
results using the GPT3 code (Park and Huyakorn, 1990); the Level III capture zones for May
and September 1988 are illustrated in Figures 7.2 and 7.3. The ground-water flow pathlines
illustrated in Figures 7.2 and 7.3 are the surface expression of two particle tracking paths; one
particle was released at the center of each of the upper two layers of the model (the layers in
which pumping was assigned). It is evident from the Level III capture zones that, even though
the subregional model domains were assigned a substantially finer discretization than was the
regional model, the discretization was still too coarse to obtain well-defined cones of depression
during the simulations. To obtain well-defined cones of depression in the vicinity of each
pumping well, the results of which could be used to obtain accurate, well-rounded capture zones
in the vicinity of each well, extremely small grid spacings, perhaps on the order of tens of feet
in many regions, would have to be used. Although a finer discretization of the subregional





































































Ten-year Level III (numerical) capture zones for May 1988 conditions.


Figure 7.2


---`1 -- (-1-- r~~'ll
:


ii CI_77 -7











































DADE
CITY


0 5 ml
6R I


Level III (numerical) Sept.
1988 ten-year capture
zones


Ten-year Level III (numerical) capture zones for September 1988 conditions.


Figure 7.3








model domains could have been implemented, it was deemed unnecessary for comparison
purposes since the ground-water flow pathlines delineated provide a clear indication of the length
and orientation of each capture zone.


Some of the Level I delineations are quite similar to the Level II (analytical modeling)
delineations. Compare, for example, the delineations at Spring Hill wells 6, 7, 10, 17 and 21;
the Hernando County West System wells 6-9, 12 and 13; and the Brooksville number 3 and 4
wells. At other wells, there are various degrees of differences between the Level II and Level
m capture zones. Significant differences between the Level II and Level m capture zones may
be attributed to:

1) The subdomain model transmissivities may be different, in some
cases substantially, than that used for the analytical delineations.

2) The Level lm delineations incorporate Floridan aquifer
heterogeneity (transmissivity is variable), which in turn may
influence the velocity and direction of ground-water flow (pathlines
refract as they pass from one transmissivity zone into another).

3) Simulated hydraulic gradients and angles of regional ground-water
flow differ, at least slightly, in most places from the observed
values.

4) There is a vertical component of ground-water flow incorporated
in the Level III delineations that is neglected in the Level II
delineations.

5) The Hernando County West System wells 3, 4, 5, (Weeki Wachee)
and 13 have Level III capture zones that are substantially
influenced by the simulated cone of depression about Weeki
Wachee Spring. The cone of depression (which is not evident on
the observed potentiometric surface maps) causes the capture zones
for these wells to be "deflected" to the north of Weeki Wachee.

The above analysis generally holds true for a comparison of the Level I and Level HI capture
zone delineations as well, since the Level I delineations were based largely upon the Level II
delineations. The one region where there is a significantly better match between the Level I and
Level HI delineations than there is between the Level II and Level n delineations is in








southwestern portion of the county, at Spring Hill wells 2, 5, 8, 9 and 19. In this region, the
analytical assumption of a uniform regional ground-water flow direction is violated due to
configuration of the observed potentiometric surfaces; regional ground-water flow upgradient of
the wells has a more northerly component than does the flow in the immediate vicinity of these
wells.








8 FINAL WHPA DELINEATIONS



8.1 Technical Approach

Based upon knowledge obtained and conceptualizations developed during the Level I-II capture
zone delineations, a technical approach for delineating the final WHPAs to be implemented into
,the County WHP Ordinance was developed. The final delineation strategy is based primarily

upon two fundamental observations of the Level I-llI capture zones: 1) the capture zones are
generally long and narrow, and 2) the orientation of the capture zones is dependent upon the
direction of regional ground-water flow. Therefore, to adequately protect those portions of the
l Floridan aquifer that supply ground water to the major public supply wells in Hemando County,
it is critical that seasonal and annual fluctuations in the direction of ground-water flow be
L incorporated into the final delineation methodology. The procedure upon which the final WHPA
- delineations are based is outlined below:

1) The May and September potentiometric surfaces for 1987, 1988
: and 1989 were examined to determine the regional direction of
ground-water flow in the vicinity of each public supply well. For
each well, the boundary angles (largest and smallest) and regional
^ hydraulic gradients immediately upgradient of the well were
identified.

2) The estimated year 2000 discharge for each public supply well was
identified. This information was determined by the District in
conjunction with the various utilities.

3) The aquifer transmissivity and thickness at each well were
determined as they were during the Level II delineations, with the
exception that for wells that did not have a documented
transmissivity value obtained from an aquifer test, the
transmissivity was determined at the well using the numerical
1 ground-water flow model developed during this study. As was
done during the Level II delineations, the model-derived
transmissivities were adjusted (decreased) based upon the various
well depths.



8-1








4) Using the information obtained as outlined above, the two, five
and ten-year capture zones for each well were delineated using
analytical modeling. Two delineations were performed for each
well, one for each of the two bounding angles of regional ground-
water flow identified.

5) The analytical capture zones delineated during step 4 were
critically examined to determine if the direction of regional
ground-water flow changed significantly over the length of the
capture zone (the analytical modeling approach assumes that the
direction of regional ground-water flow is constant). For capture
zones within which a significant change in flow direction did
occur, the capture zones were modified so that their boundaries are
perpendicular to the observed potentiometric surface.

6) The final WHPAs were delineated as the regions encompassing the
composite capture zone for each well (the composite capture zone
is the region consisting of the two bounding capture zones for a
given well and the enclosed region between the bounding capture
zones).


As explained in Chapter 5, for groups of multiple wells that lie close to one another, it is
reasonable for capture zone delineation purposes to treat these wells as one equivalent well. For
the equivalent well delineations, the individual well locations were averaged to obtain an
equivalent well location, and the estimated year 2000 pumping rates of the individual wells were
summed to obtain the equivalent well pumping rate. The transmissivity and aquifer thickness
values for the equivalent well were determined based upon the largest composite open hole
interval determined by examination of each well's casing and completion depth.


The final analytical capture zone delineations were performed for most of the public supply wells
using the MWCAP module of the WHPA code. This particular module neglects the effects that
well interference (overlapping cones of depression) may have upon delineated capture zones.
This approach is reasonable for most of the public supply wells in Hernando County. However,
there are several groups of wells in which one well lies immediately upgradient of another well;
in these situations the effects of well interference could be important and therefore the GPTRAC
module of the WHPA code, which accounts for well interference effects, was utilized. The final








capture zones were delineated using GPTRAC for the following wells: Spring Hill wells 2 and
8; Hernando County Dogwood Estates wells 1 and 2; Hernando County Hill N' Dale wells 1
and 2; and Seville wells 11, 14 and 15 (wells 11 and 14 were combined to form an equivalent
well). For the above well sets, the transmissivities and aquifer thicknesses determined for each
well were averaged to obtain one value for input to GPTRAC. For the Seville wells, the
observed angles of regional ground-water flow and hydraulic gradients were slightly different
for each of the wells and were thus averaged to obtain a single value for input to the code.


In general, the assumption of uniform regional ground-water flow (constant hydraulic gradient
and direction of flow) is reasonable within the five-year capture zones. For most capture zones
that were adjusted to account for varying direction of ground-water flow, the adjustment was
only necessary within the five-year to ten-year time-of-travel portion of the capture zones.
Although the orientation of the capture zone boundary was adjusted to lie perpendicular to the
observed potentiometric surface, the length of the delineated capture zone was not modified.
This approach is consistent with the assumption that the capture zone width does not change;
changes in the regional hydraulic gradient are assumed to be compensated for by changes in
aquifer transmissivity such that the ground-water flux within the capture zone remains constant.
This approach is reasonable with respect to the data availability for the ground-water flow
system.



8.2 Rational for Selection of Final Delineation Method

The analytical modeling method (implemented using the WHPA code), modified at some
locations using flow-net analysis, was selected to conduct the final delineations because this
approach has several significant advantages over the numerical modeling approach for the
prevailing hydrogeological conditions in Hemando County. First of all, using the WHPA code,
it is a straight-forward matter to conduct multiple capture zone delineations using various angles
and hydraulic gradients of regional ground-water flow. As presented in the next section, the
bounding angles of regional ground-water flow (upon which the composite capture zones are
based) for many of the wells occurred in May or September of 1987 and 1989, which are four







periods not covered during the numerical modeling. It would have been costly and time-
consuming to attempt to model the entire 1987-1989 period using numerical modeling.
Furthermore, it is very difficult, and at many places impossible in a practical sense, to create
a simulated potentiometric surface that matches the observed potentiometric surface in sufficient
detail such that the angles of regional ground-water flow are precisely the same. With respect
to the long, narrow capture zones that occur in Hernando County, a numerical model simulation
error of only several degrees in the direction of ground-water flow may cause the delineated
capture zones to be significantly in error.


Although the numerical simulation results were not utilized to delineate the final capture zones
for this study, the numerical model developed for Hernando County is quite useful in several
ways. First of all, the transmissivities obtained from the numerical model were used to derive
the transmissivities at individual wells. Previous numerical modeling studies, such as Ryder
(1985) which was utilized during the Level II delineations, did not focus on Hernando County
in particular but rather encompassed large regions of which Hernando County is only a small
portion. Local hydrogeologic conditions in Hernando County, therefore, are better represented
in the model developed during this study.


Secondly, the numerical model was used to assess the potential impacts of the increase of public
supply well discharge rates to the estimated year 2000 values. The model indicates that
drawdowns due to the increased withdrawal rates will be very minor, and therefore it is
reasonable to base the delineated WHPAs upon a ground-water flow system conceptualization
based upon current hydrogeologic conditions.


Finally, the development and calibration of the regional ground-water flow model assisted with
the conceptualization of the ground-water flow system. For example, the Level III capture zone
delineations indicate that there are vertical components of ground-water flow within the Upper
Floridan due to areal recharge. Therefore, ground-water flow to a well has both horizontal
(lateral) and vertically downward components. Since the final capture zone delineations were
obtained through analytical modeling, which does not account for vertical ground-water flow,








the final delineations are conservative in this respect (they enclose a larger region -than they
would if vertical flow effects were accounted for). This analysis also indicates that the portion
of the aquifer below a given well depth probably contributes very little water to that well.


8.3 Final WHPA Delineation Results
The required input parameters for the final capture zone delineations are presented in Table 8.1.
For each well or group of wells, the potentiometric surface maps from which the bounding
angles of regional ground-water flow were identified and measured are indicated. For example,
for Hernando County Ridge Manor well number 1, the observed regional ground-water flow
directions for May 1989 and September 1987 bound, or enclose, all other ground-water flow
directions in the vicinity of that well for the period examined (1987-1989). The Period 1
information pertains to one of the bounding angles of regional ground-water flow at each well,
and Period 2 pertains to the second bounding angle. Most of the Period 1 ground-water flow
directions and hydraulic gradients are for the May potentiometric surface map of a given year,
and likewise Period 2 ground-water flow parameters were generally obtained from a September
potentiometric surface for a given year. However, the bounding hydrologic conditions at each
well were not always observed during the May and September periods; see, for example,
Hernando County West System well number 3. Analytical capture zone delineations that were
subsequently modified using flow-net analysis are marked by an asterisk (*) in Table 8.1.


The final ten-year capture zone delineations for Period 1 and Period 2 are presented in Figures
8.1 and 8.2. In general these capture zones are similar to the Level I and II capture zones
presented in Chapter 5. Significant differences do exist, however, between the final delineations
and the Level I and II delineations for the following reasons:

The estimated year 2000 well discharge rates are generally higher
than those reported for 1988




-


Table 8.1 Information for Final WHPA Delineations


Period 1 Parameters Period 2 Parameters Calculated
UTILITY NAME SWFWMD UTM-X UTM-Y 2000 Trans. Poros. b Fixed Radius'
WELL ID (m) (m) Q 1at well (ft)
(tWELL ID) (at weld) Date Gradient Alpha Date Gradient Alpha 1 yr 2 yr
___ /d) (_ _(nf) (t)

HERNANDO CO.- Ridge Manor I 384683 3152861 15144 49,200 0.15 250 May 1989 7.50x104 136.4 Sept. 1987 6.00x104 168.1 217 306
HERNANDO CO.- Ridge Manor 2 383089 3153862 34614 37,176 0.15 79 May 1989 7.20x104 128.0 Sept. 1987 6.20x104 158.4 583 824
HERNANDO CO.- Ridge Manor 4 379656 3155805 59985 142,012 0.15 400 May 1989 2.80x 104 95.7 Sept. 1988 2.10x 104 131.1 341 482
HERNANDO CO.- Ridge Manor 5 379275 3155840 2753 24,142 0.15 68 May 1989 2.80x 104 95.7 Sept. 1988 2.00x 104 129.6 177 250
HERNANDO CO,- Dogwood Estates I 368544 3161311 8403 15,243 0.15 56 May 1989 2.80x10-4 178.5 Sept. 1987 3.20x104 101.4 341 482
HERNANDO CO.- Dogwood Estates 2 368625 3161187 8403 35,112 0.15 129 May 1989 2.80x104 178.5 Sept. 1987 3.20x104 101.4 225 318
HERNANDO CO.- West System 2 350951 3158289 53366 18,800 0.15 256 May 1988 4.80x104 174.5 Sept. 1988 4.50x104 162.5 402 568
HERNANDO CO.- West System 3,4,5 343951 3157211 32019 347,609 0.15 205 Sept. 1989 4.50x104 151.4 Sept. 1988* 6.00x104 144.0 348 492
HERNANDO CO.- West System 6 349071 3157913 21346 550,498 0.15 215 May 1987 2.60x104 181.5 Sept. 1988* 4.40x104 162.0 277 392
HERNANDO CO.- West System 7 349128 3158128 64039 472,340 0.15 185 May 1987 2.60x104 182.5 Sept. 1988* 4.40x104 163.0 518 732
HERNANDO CO.- West System 8 349242 3158557 53366 486,582 0.15 265 May 1987 2.60x104i 183.6 Sept. 1988* 4.60x104 161.2 395 559
HERNANDO CO.- West System 9 349301 3158895 170770 469,748 0.15 258 May 1987 2.60x104 183.7 Sept. 1988' 4.60x104 160.5 716 1013
HERNANDO CO.- West System 10 354025 3158466 85385 231,920 0.15 223 May 1989 4.30x104 174.3 Sept. 1987 3.80x104 185.5 545 770
HERNANDO CO.- West System 11 354058 3158897 10673 215,280 0.15 207 May 1988 5.05x104 174.4 Sept. 1987 3.80x104 186.6 200 283
HERNANDO CO.- West System 12 349361 3159295 170770 546,985 0.15 300 May 1987 2.60x 104 182.8 Sept. 1988' 4.80x 104 161.0 664 939
HERNANDO CO.- West System 13 346788 3157974 53366 238,961 0.15 92 May 1989' 4.25 x104 155.2 Sept. 1989 4.56 x 104 167.1 670 948
HERNANDO CO.- West System 14 352286 3154178 65746 358,146 0.15 255 May 1988 5.05x104 160.5 Sept. 1987* 5.10x104 139.0 447 632
HERNANDO CO.- West System 15 352538 3154791 65746 491,573 0.15 350 May 1988 5.10x 104 164.4 Sept. 1987* 5.10x 0o4 140.1 381 539
HERNANDO CC.- West System 16 352454 3154545 65746 463,483 0.15 330 May 1988 5.20x 104 162.5 Sept. 1987 5.10x 104 139.0 393 556
HERNANDO CO.- Hill N'Dale 1 373375 3155717 3453 38,100 0.15 110 May 1988 3.16x104 96.9 Sept. 1989 1.76x10' 58.5 156 221
HERNANDO CO.- Hill N'Dale 2 373350 3155871 25324 71,294 0.15 303 May 1988 3.16x104 96.9 Sept. 1989 1.76x10s 58.5 254 360












Table 8.1 Continued.


Period I Parameters Period 2 Parameter Calculated
UTILITY NAME SWFWMD UTM-X UTM-Y 2000 Trans. Poros. b Fixed Radius'
WELL ID (m) (m) Q at well (ft)
(ft'/d) (fte'd) Date Gradient Alpha Date Gradient Alpha I yr 2 yr
(ft) (ft)
HERNANDO CO.- Springwood Est. 2 357142 3151040 13373 70,900 0.15 400 May 1988 3.87x104 137.2 Sept. 1987 6.20x104 160.1 161 228
HERNANDO CO.- Springwood Est. 3 355663 3152412 13373 112,234 0.15 138 May 1988 4.36 x 104 144.5 Sept. 1987* 6.60x 104 155.0 274 387
BROOKSVILLE 1,2 364114 3158884 138411 205,796 0.15 450 May 1989 2.86x104 179.5 Sept. 1987 3.20x104 144.0 488 690
BROOKSVILLE 3 363757 3160873 69205 57,542 0.15 285 May 1989 2.96x104 177.0 Sept. 1987 3.83x104 138.1 434 613
BROOKSVILLE 4 363211 3156047 69205 358,997 0.15 404 May 1988 2.67x104 178.1 Sept. 1988 3.56x104 154.5 364 515
BROOKSVILLE 5 363349 3156199 20355 422,088 0.15 475 May 1988 2.64x104 181.5 Sept. 1988 3.56x104 156.5 182 258
SEVILLE 11,14 351629 3174181 376097 38,100 0.15 125 May 1987 3.75x104 140.8 Sept. 1989 3.80x104 127.0 609 862
SEVILLE 15 351773 3173733 10639 53,699 0.15 50 Sept. 1988 2.82x104 137.3 Sept. 1989 5.65xl04 125.5 406 574
SPRING HILL 2 342396 3148734 38801 133,700 0.15 164 May 1989 3.91x 104 150.5 Sept. 1988 5.30x 104 157.0 428 605
SPRING HILL 5 342301 3147689 81483 267,400 0.15 194 May 1989' 4.10 x 10 149.5 Sept. 1988* 5.78 x 104 158.0 570 807
SPRING HILL 6,17,21,26 352075 3149302 388012 333,699 0.15 377 May 1988 6.00x104 141.0 Sept. 1987 4.80 x 04 136.4 893 1263
SPRING HILL 7,10,25 347806 3152604 213407 763,081 0.15 375 May 1988 3.80x 10 146.0 Sept. 1988 5.90x 104 150.6 664 939
SPRING HILL 8 342234 3148798 77602 78,000 0.15 130 May1989 3.91x104 150.5 Sept. 1988 5.30x104 157.0 680 962
SPRING HILL 9,19,23 346814 3148143 236688 420,231 0.15 375 May 1988 5.14x 104 134.0 Sept. 1988* 7.30x 104 143.0 699 989
SPRING HILL 13,16,20 351422 3154630 329811 652,344 0.15 462 May 1988 4.80x04 160.5 Sept. 1988 4.70x 04 144.0 744 1052
SPRING HILL 15,18,24 348221 3150331 310410 391,035 0.15 393 May 1987 3.75x104 131.2 May 1988 6.00x104 146.7 782 1106


* Analytical capture zone adjusted for changing direction of regional ground-water flow


SSee Section 8.4


I




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs