United States
Environmental Protection
Adency
Robert S. Kerr Environmental
Research Laboratory
Ada OK 74820
Research and Development
EPA/600/S2-88/046 Sept. 1988
vvEPA Project Summary
Hydrologic - Hydrochemical
Characterization of Texas Gulf
Coast Saline Formations Used for
Deep - Well Injection of Chemical
Wastes
Charles W. Kreitler, M. Saleem Akhter, and Andrew C. A. Donnelly
This research was conducted to in-
vestigate fluid migration potential, direc-
tion and velocities in the regional
hydrologic environment of the Texas Gulf
Coast Tertiary formations in the context
of deep-well infection of hazardous
chemical wastes. The study has focuss-
ed on the Frio Formation because it is
the target of a large waste injection
volume and because a large database of
formation pressures and water chemistry
is available in the Frio.
Pressure data gathered from drillstem
tests and bottomhole pressure measure-
ments In onshore oil and gas wells were
used in evaluating pressure regimes.
Pressure-depth profiles and poten-
tiometric surfaces were constructed
from the pressure data and these reflect
existence of three hydrologic regimes: a
shallow fresh to moderately saline water
section in the upper 3-4 thousand feet,
an underlying 4-5 thousand feet thick
essentialy saline hydrostatic section,
and a deeper overpressured section with
moderate to high salinities. The com-
plexity of the hydrologic environment is
enhanced due to extensive depressuriza-
tion in the 4,000 to 8,000 ft depth inter-
val. This presumably results from the
estimated production of over 10 billion
barrels of oil equivalent and associated
brines from this interval alone in the past
50 years. Hydrologic analysis indicates
that transition to geopressured
sediments in some areas of the Gulf
Coast is encountered as shallow as 6,000
feet.
Due to variability in thickness and
pressure regimes, a composite poten-
tiometrlc surface of the entire Frio can-
not be constructed to determine natural
flow gradients or natural points of
discharge. Present conditions are
already quite altered from original ones.
Potentiometric surfaces representing
discrete depth intervals were mapped for
evaluating regional flow trends. Average
formation porosity and permeability
values were obtained from published
data. These values and the flow gra-
dients determined from potentiometric
surfaces were used to compute linear
fluid flow velocities ranging from 0.01
ft/year to 105 ft/year in the lateral
direction.
Potential for vertical fluid migration
was identified from equivalent en-
vironmental hydraulic heads. The
presence of widespread pockets of
depressured formations significantly af-
fects the direction and value of fluid gra-
dients, in as much as these depressured
oil and gas fields carry the risk of
becoming sinks for the injected chemical
wastes. Any subsequent pathway to
sources of fresh water will be determin-
ed by the capacity of faults and fractures
to act as conduits for flow, and/or the
presence of abandoned wells to facilitate
such flow.
Published water chemistry data was
supplemented by field sampling of
waters from thirty-two oil fields. Active
recharge of Frio by continental waters is
-------
not occurring. All waters sampled ap-
pear-in isotopic equilibrium with the rock
matrix. The chemical composition of
brines from the northern section in-
dicates that salt dome dissolution is the
primary reaction controlling water
chemistry in this region. Brines from the
deeper geopressured section may be
leaking into the hydrostatic section of
the central and southern Gulf Coast Frio.
This leakage does not appear to extend
all the way up to the fresh to moderatly
saline section.
An offshoot of the current research is
the evidence of microbial degradation of
organic material shallower than approx-
imately 7,000 ft. The lack of organic acids
and the alteration of Frio oils from these
zones suggests biodegradation. This
has useful implication for degradation of
injected chemical wastes and needs to
be further investigated.
A detailed analysis of the localized
hydrodynamics in Victoria County, Texas,
as a sample case study shows the ap-
plicability of the developed techniques
to injection facility siting and monitoring
process, where depressurization was
observed on a local, county-size scale.
This Project Summary was developed
by EPA's Robert S. Kerr Environmental
Research Laboratory, Ada, OK to an-
nounce key findings of the research pro-
ject that is fully documented in a
separate report of the same title (see Pro-
ject Report ordering information at
back).
Introduction
Liquid wastes generated by chemical
and manufacturing industries have been
disposed of by deep-well injection into Gulf
Coast Tertiary formations for nearly 35
years. The popularity of this method is
reflected in the ever increasing volumes
estimated at bout 6 billion gallons per year
in the mid 1980s in Texas. These practices
are regulated by federal and state agencies
concerned with preventing contamination
of ground water resources. Figure 1 shows
the active waste injection sites along the
Texas Gulf Coast.
The injection of large volumes of liquid
chemical wastes in the permeable Tertiary
sediments along the Texas Gulf Coast
raises concern about the hydrologic and
geochemical interaction with formation
fluids. A description of this interaction re-
quires first the characterization of formation
fluid hydrology and hydrochemistry, and
then the evaluation of superposition of in-
jected fluids onto the existing system. This
study has dealt with the first part, namely
describing the hydrologic regimes existing
in the Tertiary formations, the potential for
fluid movement, and the chemistry of for-
mation waters. The chemical processes
controlling interaction of injected chemical
wastes with the insitu fluids and formation
rocks are being investigated by the Bureau
of Economic Geology under a separate
cooperative agreement with the EPA.
Potential for fluid migration in the subsur-
face is controlled by formation hydrologic
properties (permeability and porosity) of the
sedimentary formations, existence of a flow
gradient and pathways for flow. The
methodology described in this study
facilitates determination of flow gradients
and velocities. This can be integrated with
regional geologic information about flow
paths such as depositional facies, faults
and fractures for compiling a geohyrologic
flow model. They hydrochemical informa-
tion from thiSxH-eport can be used for
hydrogeologic interpretation and for analyz-
ing chemical interaction and degradation
processes.
Procedure
The main focus of this research is the
description of Texas Gulf Coast Frio
hydrologic and hydrochemical environment
using formation fluid pressures and water
chemistry. The formation pressures used
to construct pressure-depth profiles and
potentiometric surfaces are taken from
drillstem tests (DSTs) and bottomhole
pressure measurements in oil and gas
wells. Nearly 17,400 pressure values in Frio
were gleaned from a large commercial
database after careful screening. Figure 2
plots these pressures versus depths and
was the starting point in identifying the dif-
ferent hydrologic regimes. Reliability of DST
data was varified by plotting histograms of
initial and final shut-in test pressures and
by evaluating their convergence ration.
These pressures were further separated by
test depths in 2,000 ft thick intervals, with
the objective of segretating the shallow
hydropressured section from the deeper
saline hydrostatic and overpressured sec-
tions. Additionally, brine chemistry data
were used to confirm the delineation of dif-
ferent regimes. Fluid pressures were con-
verted to equivalent fresh water and brine
heads for constructing poteniometric sur-
faces. Selection of which fluid gradient to
use for conversion to hydraulic heads was
based on analysis of water salinity data. A
surface contouring package CPS-1 was
utilited for making the potentiometric con-
tours. Data in each horizontal depth slice
were carefully screened to cull abnormally
high and low values and were selected from
similar time intervals (usually 10 years) tc
minimize their dated nature.
Potential for vertical fluid migration was
analyzed by constructing residual brine-
equivalent potential surfaces. This involv-
ed subtracting the potentiometric surface
of one (shallower)depth interval from the
other (deeper) depth interval.
The hydrochemistry effort consisted o
evaluating nearly 850 Frio analyses frorr
previously published reports. These most-
ly contained major cation and anion data.
Thirty-two additional oil field waters were
sampled by the Bureau of Economic
Geology. These were analyzed for major
and minor ions, isotopes, organic acids,
and organic composition of oils. Samples
were collected from depths between 3,000
and 10,000 feet for a good representation
of normal as well as potentially biodegrad-
ed oils. Various plots were generated to
determine correlation between chemistry,
origin and migration patterns of these
waters. Figure 3 provides an overview of the
range of salinity values encountered along
the Gulf Coast Frio. For ease of handling
the large pressure and chemistry data, the
study area was divided into three regions:
A, B, and C; corresponding to north, cen-
tral and south Gulf Coast.
Results
Pressure-depth profiles generated
separately and integrated for different
regions and different well types in the Gulf
Coast reflect the complexity of regional
pressure regimes. Two major hydrologic
regimes are evident on Figure 2: a brine
hydrostatic regime (with a slope of 0.465
psi/ft) which extends to depths of
10,000-11,000 ft, and, a geopressured
regime (with a slope approaching 0.9 psi/ft)
which extends as shallow as - 7,000 ft and
is shallower than previously recognized.
The hydrostatic regime represents a hydro-
logic zone of potentially active ground-water
circulation. The geopressured regime re-
presents a zone of restricted circulation.
Chemical wastes are injected into the
hydrostatic section. Also evident are the
large areas of depressurization correctable
to hydrocarbon producing fields. Poten-
tiometric surfaces for Frio slices are quite
flat in the shallow fresh to moderately saline
sections above 4,000 ft depth. But the
deeper saline section in the 4,000-8,000 ft
interval contain widespread sub sea level
potential contours indicating depressured
hydrologic conditions. These result in
horizontal flow gradients significantly
steeper than those in the shallow sections
Still deeper sections are dominated by
highly positive potentiometric contours due
-------
100°
Explanation
• 74 Deep-well injection
1224) Inactive well
167C Commercial Well
-------
8-
O-U:
<^§S;f •
. .."•s- • •
Figure 2.
2000 4000 6000 8000 10000 12000 14000 16000 18000
Depth from Surface (feet)
Pressure-depth diagram for Frio Formation; regions A, B, and C data. 17,411
pressure measurements used in this figure.
to the transition into geopressured
sediments. Residual surfaces constructed
to assess vertical flow potential also recon-
firm the tendency of flow to be directed
toward depressured sections, whith the flow
gradient being a function of the degree of
depressurization in a slice. No regional flow
into shallow fresh aquifiers is observed.
Analysis of Frio water chemistry shows
a gradual shift from Na-CI water in the north
and north central retion to a Na-Ca-CI
character in the south central and a Na-Ca-
CI-SO4 water in the south. Chlorides in-
crease with depth in the north Frio, in the
salt domes region. This correlation is
variable in the south region and is reversed
in San Patricio and Nueces Counties, which
may be tied to the transition to geopres-
sured conditions. Sulfate concentrations
are generally low. Sodium concentrations
mostly increase linearly with chloride ex-
cept for the south. Calcium versus chloride
is variable and may be linked to cation ex-
change in clays. Similarly, relatively high
magnesium values may be either derived
from clay reactions or dolomitization. A plot
of bromide versus chloride breaks up in two
trends; one in increasing Br with Cl, and
the other is constant Br with increaing Cl
(Figure 4). Plotting CI/Br ratio versus Cl,
and Na/CI ratio versus Cl similarly shows
separation of two populations. The CI/Br
and Na/CI ratios for the low Br waters
indicate halite dissolution in the Houston
Embayment salt dome region. Source of the
high Br in the central and southern region
suggests upward leakage of deeper waters.
Total field titrated alkalinity provides a
qualitative estimate of the organic acids in
deep formation waters. Two trends are
observed in the total alkalinity versus
organic alkalinity correlation; increasing
organic acid concentration with total
alkalinity, and almost zero organic alkalini-
ty for a limited total alkalinity value in some
samples. In the first trend, nearly 50% of
the total alkalinity is attributable to organic
acids. This suggests existence of a decar-
boxylation reaction. For the second trend,
absence of organic acids for total alkalini-
ty values less then 800 mg/l suggests
biodegradation. These waters were col-
lected from depths shallower than 7,000 ft.
Gas chromatographic analyses of eleven
oil samples were performed to test for
evidence of biodegradation. Normal paraf-
fins (NC) and isoprenoids (IP) appear to
dominate the composition of these oils.
Paraffins between C5 and C13 are most
susceptible to biodegradation. The ratio of
NC17 to IP19 as reflected in the loss of
NC17 peak in comparison to the IP19 peak
is another indicator of biodegradation. The
phenomenon is oberved in the sample in
Figure 5a, which shows a loss of nearly all
organic compounds. The oil sample of
Figure 5b in comparison is not biodegrad-
ed. Five out of the eleven oil samples in-
dicate varying degree of biodegradation. All
degraded oils were collected from depths
shallower than 7,000 ft.
An insitu pH of 5-6 was estimated for the
oil-field water sampled. A linear trend of
higher pH for increasing alkalinities was
observed. Degassing of CO2 does not ap-
pear to cause significant loss of inorganic
alkalinity. Isotope composition of hydrogen
(52H) versus oxygen (<518O) for Frio waters
shows a general trend of isotopic enrich-
ment of 518O away from the meteoric water
line (Figure 6). Increasing 618O but cons-
tant 62H values with depth and isotopic
equilibration of Frio waters with formation
clays is observed in field sampled brines
and other available data. Recently recharg-
ed meteoric waters (light 518O values) were
not found.
Discussion
The Frio pressure-depth profiles indicate
hydrostatic and subhydrostatic conditions
in sediments above 10,000 ft. This coexists
with overpressures observed as shallow as
6,000 ft. Large scale depressurization is
linked to hydrocarbon production. This
variability of pressure regimes is also
reflected on the potentiometric surfaces
which tend to be flat in the shallow sections
and show steeper gradients in the deeper
sections. Because of this depressurization,
injected wastes may be constratined from
migrating upward into fresh-water aquifiers
but may migrate toward the depressured oil
-------
1000.
_, 3000 i
| 5000
*• 7000
^ 5000-
£ 77000-
13000
15000
7000 •
3000 :
5000 •
7000 •
5000 -
7 7000 •
73000 •
75000 .
4
* • ^
" i • "«
/• *; . " •
"•"•^VT" «
" \ '
o 50000 roooo 150000
Chloride (mg/L)
0 50000 10000 150000
Chloride Img/L)
Frio outcrop
/ 1000-
-,. 3000 •;
g 5000 :
^ 700O:
f. 5000:
,S 1 7 7000-
73000 •
1 *nnn -
• •» • •
» '"i^1 *
9" •:'^"} *
•**!??>•'•'•
:?:>''
50000 70000 750000
Chloride (mg/L)
Mexico
7000 •
9000
11000 -
13000 •
15000
50000 10000 150000
Chloride (mg/L)
Figure 3.
Chloride distribution for the Frio Formation for various depths from Northgulf.
Northcentral. Southcentral and Southgulf regions. Chemical compositions in
mg/L.
and gas fields. Overpressured and hydro-
static conditions exist at same depth range
and suggest hydrologic discontinuity in the
form of compartmentalization. This may be
advantagous in locating future waste injec-
tion facilities.
County scale potentiometric surface map
(Figure 7) show the general coincidence of
depressurization and oil fields. The coun-
ty scale is considered an appropriate scale
to map the depressurization; the county
area is large enough for sufficient data
coverage, but small enough to map pertur-
bations of the potentiometric surface. Such
maps should be integrated with structure
maps locating faults and salt domes, as
well as with maps of oil and gas fields, and
with deep-well/abandoned-well maps. These
integrated maps would be valuable in
evaluating permeability pathways and di-
rection for potential fluid flow for injection
factilty siting decisions.
The geochemical environment in the
4,000-7,000 ft Frio depth range used for
deep-well injection has implication for the
long-term confinement, migration and
degradation of these chemical wastes. This
environment is typically slightly acidic (pH
300
200
700-
HighBr- Trend
Low Br-Trend
0 20000 40000 60000 80000
Cl (mg/L)
Figure 4. Plot of bromide versus chloride
from brines collected for this
study for Frio Formation. Note
two different populations of
data. High Br trend from central
and south Texas regions sug-
gests leakage of brine from
deeper geopressured sedi-
ments. The low Br trend from
northern Houston embayment
region results from halite
dissolution.
5-6), saline (Cl range from 20,000 to above
60,000 ppm), reducing (presence of NH4),
warm (less than 80°C), and biologically ac-
tive (evidence of microbial degradation of
oils). The varying salinities can affect the
degree of mixing of injected wastes with for-
mation brines. Degradation of wastes
through hydrolysis proceeds favorably un-
der high or low pHs. The slightly acidic pH
of formation waters does not enhance reac-
tion rates. Biodegradation wastes will oc-
cur faster than abiotic reactions in the re-
latively shallow moderately warm (70-80°C)
hydrostatic section. But, significantly higher
temperatures are required to accelerate
abiotic processes such as hydrolysis.
Presence of microorganisms at tempera-
tures above 80°C (deeper horizons) is not
expected. Presence and relative activities
of aerobic and anaerobic bacteria associ-
ated with chemical degradation needs fur-
ther study.
Active recharge of continental waters is
not occuring in the Frio. All sampled waters
appear in isotopic equilibrium with the rock
matrix. Total dissolved solids are also much
higher than in the underlying Wilcox For-
mation, where deeper penetration of
meteoric waters is observed. Although
evidence from organic acids and biode-
graded oils suggests deep circulation of
-------
05
<0
Figure 5a. Gas chromatograph trace 'of whole oil sample TB€G-34. Sample appears to be degraded Sample depth is 6,244 ft
meteoric waters for transport of bacteria,
such an active circulation is not supported
by chemical data. The presence of Br-rich
brines in the hydrostatic sections of central
and south Gulf Coast suggests upward
leakage of fluids from the undercompacted
geopressured section. Constant Br values
in Houston Embayment region indicate the
importance of halite dissolution and reflect
lack of upward fluid leakage there. This
contradicts the general concept that growth
faults and salt domes m this region are
pathways for upward migration of deeper
brines. The process of salt dissolution and
fluid leakage suggest an active rather than
a stagnat hydrologic environment in the
Frio, in the context of geologic time.
However, natural flow rates are probably
slow enough to have no impact on the con-
finement of injected wastes.
Conclusions
The use of bottomhole pressures and
water chemistry data in conjunction with
regional structure and geology is an ap-
propriate technique for hydrologic-
hydrochemical characterization of Gulf
Coast saline formations being used for
deep-well disposal of chemical wastes.
Flow gradients in the Frio Formation
calculated from potentiometric surfaces
and available permeability values were
used to determine horizontal linear veloci-
ties ranging from 0.01 to 105 ft/year. The
depressurization in and around oil and gas
fields seems to overwhelm the natural con-
ditions. Natural hydrologic conditions may
be better delineated through hydrochemical
data. Upward migration of water in the
hydrostatic section is presently constrain-
ed by the depressurization and density dif-
ferential between shallow fresh to
moderately saline aquifiers and the deeper
saline aquifiers. Decisions on injection fa-
cility siting sould be evaluated in the con-
text of local hydrologic conditions which
can be better described with county-scale
maps.
Hydrochemical data suggests that the
Frio is not being actively recharged in
geologic time scale by continental meteoric
waters. Brines from the deeper geopres-
sured section may be leaking up into the
hydrostatic section of the Frio in the cen-
tral and southern regions. Vertical leakage
does not appear to be occuring in the nor-
thern region, where salt dome dissolution
is the dominant geochemical process. The
presence of degraded hydrocarbons sug-
-------
-------
-100
-10
• EPA Data
+ Other Data
6118O/0/00)
Figure 6. Hydrogen versus oxygen isotopic composition of Frio waters collected for this
study. Global meteoric water line from Craig (1961).
-------
a Oil/gas field in 4000- to 5000-ft
sub-sea depth interval
' Contour level (ft)
0 5 mi
0 5 10km
Contour interval 200 ft
Datum sea level
380,000
400.000
420,000
CM 9793
figure 7. Potentiometric surface, 4,000-4,900-ft slice, Victoria County. Frio and Catahoula, all classes, 1945-1984 data. Includes formation
pressure at Du Pont injection facility. Equivalent brine heads.
-------
Charles W. Kreitler, M. Saleem Akhter, and Andrew C. A. Donnelly are with
the University of Texas at Austin. Austin, TX 78713.
Jerry Thornhill is the EPA Project Officer (see below).
The complete report, entitled "Hydrologic-Hydrochemical Characterization of
Texas Gulf Coast Saline Formations Used for Deep- Well Injection of Chemical
Wastes," (Order No. PB 88-242 573/AS; Cost: $25.95, subject to change)
will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Robert S. Kerr Environmental Research Laboratory
U.S. Environmental Protection Agency
P.O. Box1198
Ada. OK 74820
10
-------
O
o
O
atz ru
oo < o
rnutM
»» 30
3WF* IS
O
5
3
0)
C
*
cr
K:
a
a
o
<^
O
•*»
35
n
3
J|
4 g'
S (
------- |