as V'pqa^ NV 891 1
c/EPA
-tjiOM ami Development
Groundwater
Quality Monitoring
Recommendations for
In Situ Oil Shale
Development
-------
EPA-600/4-83-045
September 1983
GROUNDWATER QUALITY MONITORING RECOMMENDATIONS FOR
IN SITU OIL SHALE DEVELOPMENT
By
L.G. Everett
K.E. Kelly
E.W. Hoylman
Kaman Tempo
Santa Barbara, California 93102
Contract No. 68-03-2449
Project Officer
Leslie G. McMillion
Environmental Monitoring Systems Laboratory
Las Vegas, Nevada 89114
ENVIRONMENTAL MONITORING SYSTEMS LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
LAS VEGAS, NEVADA 89114
-------
DISCLAIMER
This report has been reviewed in accordance with the U.S. Environmental
Protection Agency's peer and administrative review policies and approved for
publication. Mention of trade names or commercial products does not consti-
tute endorsement or recommendation for use.
-------
PREFACE
On behalf of the Environmental Protection Agency (EPA), Kaman Tempo is
conducting a three-year program dealing with the design of groundwater quality
monitoring programs for western oil shale. The type of oil shale development
being evaluated in this study is modified in situ (MIS) retorting as presently
proposed for Federal Prototype Lease Tracts C-a and C-b in western Colorado.
This study focuses specifically on the monitoring for in situ retorting on the
Federal Lease Tracts and has followed a stepwise monitoring methodology.
This report represents the final phase of this research program. The
goals of the first phase were to:
Review MIS monitoring programs and development with regard to po-
tential impacts on groundwater quality
Use this review to identify the key issues, uncertainties, and
unknowns with regard to design and implementation of groundwater
quality monitoring programs for MIS retorts.
The goal of this phase was to:
Protect groundwater quality by developing specific monitoring
recommendations for the use of geophysical logs and aquifer test-
ing methods for MIS oil shale development.
Monitoring recommendations are provided for the use of the following
logs: temperature, caliper, gamma, spinner, radioactive tracer, velocity,
sonic, density, electric, and seisviewer. In addition, monitoring recommenda-
tions are presented on four hydraulic testing methods and a complete program
for sample collection, preservation, and handling at MIS operations.
iii
-------
CONTENTS
PREFACE iii
FIGURES v11
TABLES ix
ABBREVIATIONS AND SYMBOLS xi
ENGLISH/METRIC CONVERSIONS xili
ACKNOWLEDGMENTS xiv
Section Page
1 INTRODUCTION 1
Background 1
Federal Prototype Lease Development 1
Previous Work 3
Present Study 3
2 SUMMARY 6
Hydrogeologic Characterization 6
Geophysical Methods 6
Hydraulic Methods 7
Sampling Methods 7
Well Design 8
Monitor Well Placement 11
Sample Collection Methods 11
Sampling Frequency 14
Sample Preservation and Handling 15
Selection and Preservation of Constituents for Monitoring 16
Sample Analysis 18
Interpretation of Water Quality Data 18
3 HYDROGEOLOGIC CHARACTERIZATION METHODS 20
General Basin Hydrogeology 20
Lower Aquifer 22
Upper Aquifer 22
Alluvial Aquifers 23
Geophysical Methods 24
Temperature Log 27
Caliper Log 28
Gamma-Ray Log 32
Spinner Log 32
Radioactive Tracer Log 36
Three-Dimensional Velocity Log 38
Acoustic Log 39
-------
Section Page
Density Log 43
Electric Logs 52
Seisviewer Log 58
Hydraulic Test Methods 61
Drill Stem Tests 62
Single Packer Tests 66
Dual Packer Tests 72
Long-Term Pump Tests 76
Evaluation of Mine Development Data 83
4 SAMPLING METHODS 85
Well Construction Factors 85
Well Construction 85
Well Size 89
Annular Seal 89
Casing Material 90
Well Security and Protection 92
Well Design and Sampling Costs 92
Well Design Costs 92
Sampling Costs 94
Monitor Well Placement 95
Sample Collection Methods 95
Bailing 97
Pumping 105
Swabbing 114
Sampling Frequency 115
Sample Handling and Preservation 116
Field Data Collection 116
Field Notes and Records, Sample Labels 120
Field Handling and Preservation Techniques 121
Sample Shipment 123
Chain of Custody 128
Selection of Constituents for Monitoring 129
Enrichment Factors 129
Indicator Constituents 142
Stable Isotopes 144
Sample Analysis and Costs 145
Trace Elements 145
Organic Methods 149
Other Inorganic Species 150
Interpretation of Water Quality Data 152
Data Analysis 152
Data Presentation 154
Data Interpretation and Reporting 156
References 157
VI
-------
FIGURES
Number page
1 Location of Tracts C-a and C-b study area in Piceance
Creek Basin 4
2 Geologic section through the Piceance Creek Basin along
north-south line between Tracts C-a and C-b 21
3 Computer plot of a typical temperature log from Tract C-a 29
4 Spinner log calibration plot 34
5 Acoustic porosity analog and aquifer production zones
for Tract C-b Well 32x-12 41
6 Density porosity analog and aquifer production zones
for Tract C-b Well 32x-12 46
7 Birdwell elastic properties log for Well 32x-12, Tract C-b 48
8 Density porosity analog and spinner survey for Tract C-a
Well CE-705A 49
9 Variable matrix density-porosity analog and spinner survey
for Tract C-a Well CE-705A 51
10 Correlation electric log for Tract C-a Well CE-705A 53
11 Detail electric log for Tract C-a Well CE-705A 54
12 A portion of a seisviewer log for Tract C-b Well 32x-12 59
13 Jacob's straight-line solution for T 65
14 Single packer injection test setup 68
15 Plots of simulated, multiple pressure, permeability tests 70
16 Dual packer steady flow injection test 75
17 Rose diagram of surface joint strikes in vicinity of MDP
area, Tract C-a 78
vii
-------
Number Page
18 Rose diagram of photo linear strikes within MOP area, Tract C-a 79
19 Illustration of parameters used by Hantush 81
20 Illustration of parameters used by Hantush 81
21 An example of multiple completion well, Tract C-b Well SG-21 87
22 Typical recompleted Upper Aquifer monitoring well for Tract C-a 93
23 USGS Upper and Lower Aquifer monitoring well design 94
24 Features of the modified Kemmerer bailer 97
25 Variation in specific conductance and temperature with depth,
Upper Aquifer Well GS-13, Tract C-a 99
26 Variation in specific conductance and temperature with depth,
Lower Aquifer Well D-17, Tract C-a 100
27 Variation in specific conductance and temperature with depth,
Lower Aquifer Well D-18, Tract C-a 101
28 Well diagram of Upper Aquifer Well GS-13, Tract C-a 103
29 Typical pump apparatus configuration 106
30 Variation in specific conductance with continued pumping,
USGS Well 75-1A, 1980 108
31 Variation in specific conductance with continued pumping,
USGS Colorado Core Hole #3, 1980 109
32 Variation in specific conductance with continued pumping,
USSGS Well TH75-1B, 1980 109
33 Variation of temperature of pumped discharge, USGS Well
TH75-1B, 1980 110
34 Variation in pH with continued pumping, USGS Well 75-1A, 1980 110
35 Comparison of pump locations and the volume of water necessary
for extraction before representative aquifer water is obtained 113
36 Water quality, data display using vectors 155
37 Trilinear diagram for displaying water quality data 155
viii
-------
TABLES
Number Page
1 Well Construction and Sampling Costs for Deep Aquifer Wells 10
2 Geophysical Data Collection, Tract C-a 25
3 Geophysical Data Collection, Tract C-b 26
4 Cost Schedule for Temperature Logs 30
5 Cost Schedule for Caliper Logs 31
6 Cost Schedule for Gamma-Ray Logs 33
7 Cost Schedule for Spinner Surveys 35
8 Cost Schedule for Radioactive Tracer Logs 37
9 Cost Schedule for 3-D Velocity Logs 39
10 Cost Schedule for Acoustic/Sonic Logs 44
11 Cost Schedule for Density Logs 50
12 Cost Schedule for Various Resistivity Logs 57
13 Sampling Costs 96
14 Variation in Water Quality with Depth in Selected
Deep Aquifer Wells, Tract C-a 102
15 Water Chemistry of Samples Collected after Discharge of
Varying Well Volumes, USGS Wells, Piceance Basin, 1980 112
16 Flow Rates of the Upper Aquifer, Piceance Creek Basin,
Estimated by Three Studies 116
17 Recommendation for Sampling and Preservation of Samples
According to Measurement 117
18 Chemical Analysis of Samples Taken from Alluvial Well A-6
for Three Different Times of Analysis 124
IX
-------
Page
Chemical Analysis of Samples Taken from Alluvial Well A-9
for Three Different Times of Analysis 125
20 Chemical Analysis of Samples Taken from Alluvial Well A-12
for Three Different Times of Analysis 126
21 Representative Concentrations in Groundwaters Adapted for
This Study 130
22 Species Enriched in the Lower Aquifer 132
23 Enrichment Factors Estimated for Spent MIS Oil Shale Leachate 133
24 Enrichment Factors for Retort Waters 136
25 Relative Likelihood of Detection of Mobility from Various
Sources to Upper and Lower Aquifers and Springs Based on
Estimated Enrichment Factors 138
26 Comparison of Analytical Techniques for Trace Element
Determinations 146
-------
LIST OF ABBREVIATIONS AND SYMBOLS
ABBREVIATIONS AND SYMBOLS
bpd barrels per day
°C degrees Centigrade
cubic feet per second
cfs
EPA
U.S. Environmental Protection
Agency
EMF
°F
ft/mi n
ft
g
gm/cc
gpm
electromotive force
degrees F
feet per minute
foot, feet
square foot, square feet
gram(s)
grams per cubic centimeter
gallons per minute
gal /ton gallons per ton
CHEMICALS, IONS, CONSTITUENTS
C02 carbon dioxide
copper sulfate
dissolved organic carbon
su If uric acid
phosphoric acid
HN03 nitric acid
CuS04
DOC
H2S04
MDP mine development phase
meq mi Hi equivalent
mg/1 milligrams per liter
MIS modified in situ
ml milliliter(s)
PVC polyvinyl chloride
RBOSC Rio Blanco Oil Shale Company
SP spontaneous potential,
self-potential
SPI secondary porosity index
USGS U.S. Geological Survey
ymho/cm micromhos per centimeter
usec microsecond(s)
3-D three dimensional
I iodine
MBAS methylene blue active
substances
NaHC03 nahcolife
NaOH sodium hydroxide
NTA nitrilotriacetic acid
TDS total dissolved solids
xi
-------
FORMULAE ABBREVIATIONS
A length of test section
c hydraulic resistance
Cu conductivity coefficient,
unsaturated
Cs conductivity coefficient,
saturated
hi static water column head
h2 applied pressure
H effective head
k hydraulic conductivity
K permeability coefficient
kD aquifer transmissivity
Ko Bessel function
L leakage factor
m slope
Q constant recovery (drawdown)
discharge
q-j ith flow interval
qn last flow interval
r distance from pumping well
s drawdown
S storage coefficient
Sp inflection point
T transmissivity
Te effective transmissivity
ti flow time for each change in rate
tn total flow time
Tn transmissivity in the direction
(9+a) with the x-axis
tp time corresponding to Sp
Tx transmissivity on major flow axis
Ty transmissivity on minor flow axis
AS change in slope
At interval transit time
Atf fluid interval transit time
At^ matrix interval transit time
Ap change in pressure
u porosity
x percentage of unsaturated strata
<(> porosity
xii
-------
ENGLISH/METRIC CONVERSIONS
°Fahrenheit
1 gallon
1 barrel
1 cubic yard
1 cubic foot
1 acre-foot
1 pound
1 acre
1 quart
1 foot
1 square mile
1 ton (short)
(°Centigrade x 9/5) - 32
3,846.2 cubic centimeters; 3.86 liters
0.16 cubic meter
0.77 cubic meter
0.028 cubic meter
1,250 cubic meters
0.0005 tonne (metric ton); 487.8 grams
0.0004 hectare
0.9463 liter
0.305 meter
2.49 square kilometers
0.909 tonne (metric ton)
xm
-------
ACKNOWLEDGMENTS
Dr. Guenton C. Slawson, Jr. was a principal initial contributor to the
report. Dr. Slawson's involvement with the report ceased when he joined the
Rio Blanco Oil Shale Company as Manager of Environmental Affairs. His insight
into monitoring requirements is highly appreciated.
Technical consultation and review for this study were provided by Mr.
Glen A. Miller, U.S. Geological Survey, Conservation Division, Area Oil Shale
Supervisor's Office.
In addition, Kaman Tempo wishes to acknowledge the support and coopera-
tive interaction of representatives of Tract C-a and C-b developers:
Rio Blanco Oil Shale Company
Ms. Rosalie Gash
Ms. Maria Moody
C-b Oil Shale Venture
Mr. R.E. Thomason
Mr. C.B. Bray
xiv
-------
SECTION 1
INTRODUCTION
BACKGROUND
Synthetic petroleum products recovered from western oil shales are ex-
pected to play an important part in supplying energy needs of the United
States during the later part of the 1900's. Various estimates of the magni-
tude of western oil shale reserves have been made. The U.S. Geological Survey
estimates that an equivalent of about 4,000 billion barrels* of oil are con-
tained in the oil shales of the Green River Formation of Utah, Colorado, and
Wyoming. These oil shale resources account for 80 percent of the known world
resources but, of course, are not completely recoverable. Recoverable re-
sources are a function of mining and retorting technology and economics, but
may amount to about 1,800 billion barrels of oil (Hendricks and Ward, 1976).
As the estimated remaining world ultimate oil resources are about 2,000 bil-
lion barrels (Tiratsou, 1976), of which less than 150 billion barrels are in
the United States, western oil shale is clearly a significant energy resource.
Federal Pnjtotyjie Lease Development
The current Federal Prototype Oil Shale Leasing Program, administered by
the U.S. Department of Interior, was initiated in 1969. Program planning and
environmental evaluation efforts by various government interagency and indus-
try groups culminated in preparation of a draft environmental impact statement
in 1971. Informational core hole drilling by firms interested in obtaining
oil shale leases was conducted in the 1971 through 1973 period. This led to
nomination of 20 potential lease tracts in Colorado, Utah, and Wyoming. The
Department of Interior selected six tracts for the prototype leasing program.
The environmental impact statement was finalized in 1973. Later in 1973, the
first lease sale was initiated. In January 1974, successful bidders for the
two Colorado lease tracts (C-a and C-b) and for two Utah tracts (U-a and U-b)
were announced. No bids were received on the proposed Wyoming lease tracts.
Environmental baseline and operation design studies were conducted over
the two years following the lease initiation. In 1976, Detailed Development
* See p. xiii for conversion to metric units. English units are used in this
report because of their current usage and familiarity in industry and the
hydrology-related sciences.
-------
Plans (DDP) were submitted for both Tracts C-a and C-b. The initial DDP for
Tract C-a called for open pit mining, surface retorting, and off-tract loca-
tions for processing facilities, overburden, and processed shale disposal. A
number of serious problems, in particular approval for off-tract disposal of
wastes, could not be resolved and a lease suspension was requested. This sus-
pension was granted in September 1976. During this suspension, a revised DDP
for Tract C-a was prepared calling for modified in situ (MIS) development plus
surface retorting of the oil shale mined from development of the MIS retorts.
This revised DDP was submitted in May 1977 and was subsequently approved by
the Area Oil Shale Supervisor (AOSS).
Initial development plans on Tract C-b (by Ashland Oil, Inc. and Shell
Oil Company) were submitted in February 1976. This plan called for a deep
mining and surface retorting (and disposal) operation. Development was sus-
pended later in 1976. In November of that year, Shell withdrew from the C-b
Oil Shale Project and Ashland formed a new venture with Occidental Oil Shale,
Inc. A revised DDP proposing MIS operations was submitted in February 1977.
Site development was initiated in the fall of 1977.
Shale deposits in the Piceance Basin that can potentially be exploited by
in situ technologies underlie an area of considerable topographic variation
that is largely undeveloped. A wide range in both hydrologic and geologic
conditions occurs throughout the area containing the deposits. Several in
situ technologies are available, each of which could have characteristic im-
pacts. There has not been sufficient experience with the various retorting
methods to determine which is the most suitable in terms of minimizing envi-
ronmental harm in the Piceance Basin. It may appear at first glance that in
situ retorting has less potential for impact to the environment than surface
retorting; however, the long-term impact to the subsurface environment may
prove this assumption to be wrong.
Monitoring of groundwater quality impacts associated with in situ oil
shale development will be difficult. Retort waters produced by small-scale in
situ operations have resulted in the identification of a wide spectrum of po-
tential pollutants. Research to date indicates that many of these pollutants
have only recently been classified, while others are still under investiga-
tion. It is not clear if the quality of the retort waters from small-scale in
situ retorting will be similar to those waters produced by large-scale commer-
cial in situ retorts.
The Federal Water Pollution Control Act Ammendments of 1972 (P.L. 92-500)
and the Safe Drinking Water Act of 1974 (P.L. 92-523) provide for protection
of groundwater quality. These mandates call for programs to prevent, reduce,
and eliminate pollution of both navigable waters and groundwater and for par-
ticular protection of drinking water resources. Similar goals are embodied in
the Toxic Substances Control Act of 1976 and the Resource Conservation and Re-
covery Act of 1976. The national responsibility for these various activities
is given to the U.S. Environmental Protection Agency (EPA). Various State
agencies also have similar responsibilities via State enabling legislation.
-------
PREVIOUS WORK
A companion report to this study, developed at Tempo and prepared by
Slawson (1980b), presents the results of a groundwater monitoring design study
of in situ oil shale development. The approach used in that study is the gen-
eral monitoring methodology developed by Tempo as follows:
Monitoring Step Description
1 Select Area for Monitoring
2 Identify Pollution Sources, Causes, and Methods of Disposal
3 Identify Potential Pollutants
4 Define Groundwater Usage
5 Define Hydrogeologic Situation
6 Describe Existing Groundwater Quality
7 Evaluate Infiltration Potential of Wastes at the Land Surface
8 Evaluate Mobility of Pollutants from the Land Surface
9 Evaluate Attenuation of Pollutants in the Saturated Zone
10 Prioritize Sources and Causes
11 Evaluate Existing Monitoring Programs
12 Identify Alternative Monitoring Approaches
13 Select and Implement the Monitoring Program
14 Review and Interpret Monitoring Results
15 Summarize and Transmit Monitoring Information
In particular, the companion report focused on modified in situ development as
proposed for Federal Prototype Lease Tracts C-a and C-b in Colorado by devel-
oping data required for an initial pass through methodology steps 1 through
13, although step 13 is not fully implemented. The methodology, in general,
and its application to monitoring design problems are described in several
other reports (Everett, 1979, 1980; Todd et al., 1976; Slawson, 1979) and will
not be presented here in detail.
A preliminary monitoring design/implementation framework has been devel-
oped for MIS retorts in the companion report. This work lead to the identifi-
cation of areas of uncertainty with regard to implementation of groundwater
quality monitoring programs for in situ facilities. These uncertainties were
found to be primarily within (1) hydrogeologic characterization and (2) sam-
pling methods utilized at the MIS retorts.
PRESENT STUDY
This study addresses the two primary groups of uncertainties regarding
the implementation of a groundwater quality monitoring program for MIS oil
shale development such as proposed for Federal Prototype Lease Tracts C-a and
C-b (see Figure 1). Hydrogeologic characterization, an essential element in
-------
0 5 10 19
*m
SCALE IN MILES
LOWER COLORADO
BASIN
NEW MEXICO
COLORADO BASIN
BOUNDARY
DRAINAGE BASIN
SOUNOAOr
Figure I. Location of Tracts C-a and C-b study area in Piceance Creek Basin.
-------
siting monitor wells and for the design of the wells to obtain consistent and
representative samples, is discussed in terms of geophysical and hydraulic
methods that are employed on the Federal Tracts. These methods are also
appropriate for other areas with oil shale stratigraphy. Geophysical and hy-
draulic methods are evaluated and ranked relative to cost, potential effec-
tiveness, and availability of testing equipment in the oil shale region.
Sampling methods are discussed, covering a wide variety of monitoring elements
including: (1) well design, (2) monitor well placement, (3) sample collection
methods, (4) sampling frequency, (5) sample preservation and handling, (6) se-
lection and preservation of constituents for monitoring, (7) sample analysis,
and (8) interpretation of water quality data. A discussion of these monitor-
ing elements is presented in the following paragraphs with detailed informa-
tion provided throughout the text.
-------
SECTION 2
SUMMARY
HYDR06EOLOGIC CHARACTERIZATION
A program designed to characterize the hydrogeology of an oil shale tract
prior to designing a groundwater quality monitoring program should include a
proper suite of geophysical logs and appropriate aquifer testing methods.
This report discusses these subjects and presents recommendations for their
use in the design and implementation of groundwater quality monitoring pro-
grams for MIS retorting areas.
Geophysical Methods
Several logs were evaluated in this study to determine their overall ef-
fectiveness in providing environmentally pertinent and reliable hydraulic
data. Those logs evaluated include:
Temperature Velocity
Caliper Sonic (acoustic)
Gamma-ray Density
Spinner Electric
Radioactive tracer Seisviewer.
With the exception of the seisviewer log, all the logs listed above were
found to be comparable to each other in cost. Accordingly, recommendation of
geophysical logs is based on effectiveness in obtaining reliable hydraulic
data.
The following log suite is recommended for its utility for hydrogeologic
characterization: temperature, caliper, sonic, and electric logs. Of more
limited value and receiving secondary, or lower, priority ranking are gamma-
ray, velocity, density, and spinner logs. The radioactive tracer and seis-
viewer logs are not recommended for obtaining hydraulic data for the design of
groundwater monitoring strategies at oil shale development sites.
-------
Hydraulic Methods
Tempo's previous study (Slav/son, 1980a) indicated a need for aquifer
testing and recommended that selected exploration and core holes be converted
to serve as testing wells. Existing wells can be conditioned or new wells
constructed to be of sufficient size to accommodate pumps for aquifer testing.
Four general methods of hydraulic testing procedures have been evaluated
and are classed as follows:
Drill stem tests
Dual packer tests
Long-term pump tests
Single packer tests.
Review of the testing procedures, equipment costs, and utility of the result-
ing data has produced the following priority ranking:
1. Dual Packer Jests provide specific hydrologic data at a minimal
cost when multiple tests are conducted in a single borehole.
Down-hole test equipment assembly allows for pumping, injection
tests, and discrete water quality sampling.
2. Long-Term Pump Tests produce the most representative data on
boundary conditions and flow patterns and are especially effec-
tive for determining regional groundwater conditions. Long-
term pump tests should be carefully planned and positioned to
provide maximum data per test because their use is limited by
the rather large expense of implementation.
3. Single Packer Tests provide horizon specific data similar to
the dual packer method. However, for each test, the packer
must be inserted and removed from the borehole. This labor
intensive activity can significantly increase the cost of data
acquisition.
4. Dri11 Stem Tests are commonly run during drilling operations.
They are of value when single, well-defined aquifer systems are
penetrated. However, when multiple aquifers are encountered
during drilling, interpretation of data resulting from drill
stem tests becomes extremely difficult. Drill stem tests are
therefore not recommended for determining hydraulic parameters
in complex hydrologic environments.
SAMPLING METHODS
The objective of a groundwater monitoring strategy in the oil shale re-
gion where MIS retort development could be selected as the mining methodology
is to (1) provide baseline groundwater quality data, (2) detect and measure
-------
groundwater flow within the abandoned retort interval, and (3) detect changes
induced by waste residuals (e.g., spent shale, retort water) within the aban-
doned retort zone. Compilation of baseline data and accurate evaluation of
the latter two aspects require collection of representative groundwater qual-
ity samples. However, a number of factors can influence the representative
nature of the groundwater samples collected. These factors include well de-
sign, sample collection methods, and sample handling procedures.
Well Design
The Upper and Lower Aquifer zones present in the Piceance Basin, Colo-
rado, are composed of numerous layers, each of which can possess variable wa-
ter quality and quantity characteristics. Since numerous wells are open or
perforated over the entire Upper or Lower Aquifer interval, the water quality
data collected from these wells represent a composite of all penetrated lay-
ers. On the other hand, a layer exhibiting greater hydrostatic head than ad-
jacent layers can influence portions of the well bore, resulting in collection
of a water quality sample that represents the high head layer and not a com-
posite of the entire open interval. Under both of these conditions, baseline
water quality data collected may not be adequately measured in detail, and for
operation/abandonment phase monitoring, groundwater flow through abandoned re-
torts may not be adequately represented. Furthermore, any trace constituents
or potential contaminant present may be sufficiently reduced below detection
limits due to the composite nature of the well design if mixing does occur.
A network of multiple completion wells is the recommended approach for a
groundwater monitoring program near the retort fields. Multiple completion
well design will enable the collection of representative data from each of the
intervals potentially affected by the oil shale retorting operation. The sug-
gested specifications for this type of well are:
Steel casing and polyvinyl chloride (PVC) well construction mate-
rial. Although the structural properties of PVC may preclude its
use as a casing material, the inert characteristics of PVC make
it ideal as a well construction material. PVC is also inexpen-
sive when compared with other materials.
The diameter of the PVC should be large enough to accommodate a
submersible pump. The recommended diameter and wall thickness
of the PVC is 6 inches OD and schedule 40 (19/64 inch),
respectively.
Each well of the multiple completion should be completed in a
different interval using cement grout to prevent the interconnec-
tion of different intervals.
Wells should be developed thoroughly, i.e., fresh water circu-
lated in the well bore, to remove any traces of drilling fluid or
other materials that may affect water quality samples.
It appears that wells completed over the entire Upper or Lower Aquifer
are suitable for groundwater monitoring in areas removed from the retort
8
-------
field(s). This open type of well design will provide general information on
the regional water quality and does not require the finer levels of completion
necessary for wells close to the retort field(s). It is recommended that the
same specifications suggested for the multiple completion wells be utilized
for the more regional wells if samples are to be collected via a submersible
pump.
The recommended specifications presented above are designed to allow for
sampling with a submersible pump. Although pumping samples is the best ap-
proach from a technical standpoint, there are some distinct trade-offs with
respect to the construction costs associated with the larger diameter wells.
There are also some significant trade-offs with respect to sampling costs.
The approximate costs for the well development are:
Approximate Cost
Design per Well (dollars)3
Large Diameter (6-inch) Well
Upper Aquifer single completion 18,000 - 20,000
Lower Aquifer single completion 35,000 - 38,000
Multiple completion 53,000 - 58,000
Smaller Diameter (2-5/8-inch tubing strings) Wells0
Dual completion (i.e., two completion strings 35,500 - 38,000
with one open over the entire Upper Aquifer
and one open over the entire Lower Aquifer
Multiple completion 39,000 - 44,000
Notes:
aThese costs include drilling, development, casing material, etc. in 1980
dollars.
Submersible pump can be utilized for sample collection.
Bailer can be utilized for sample collection.
These cost data show that large-diameter single and multiple completion wells
are more expensive than smaller diameter dual and multiple completion wells,
respectively. In addition, the cost for an entire groundwater monitoring pro-
gram would be substantially higher and equal to the cumulative cost of all
wells in the system. The approximate costs of an entire groundwater monitor-
ing program, including sampling, are presented in Table 1 of this section.
-------
TABLE 1. WELL CONSTRUCTION AND SAMPLING COSTS FOR DEEP AQUIFER WELLS (1980 dollars)
Item
Well Construction
Sampling Costs
Capital Requirements
Operational Requirements
(Quarterly)
Labor (quarterly)
Five-year Total (including
construction of 12 monitoring
well sites)
Fixed Submersible
Pump
53, 000-58 ,000a
61,800-79,800
200-400
135-200C
704,500-787,800
Portable Submersible
Pump (USGS)
53,000-58,000
55,000-60,000
1,400-1,700
11, 200-14, 000d
943,000-1,072,000
Bailing
(Tract C-a)
35,500-38,000
8,000-10,000
200-400
135-200°
440,700-478,000
Swabbing
(Tract C-b)
39,000-44,000
N/Ab
16,000-18,000
3,500-4,300e
858,000-974,000
Notes:
aAssumes similar well construction for fixed pump as with portable pump.
Tract C-b contracts swabbing rig, thereby eliminating capital requirements.
cAssumes sampling eight wells per day.
Assumes sampling one well per day.
eAssumes sampling three wells per day.
-------
Monitor Well Placement
One of the goals of hydrogeologic characterization efforts is to allow
description of groundwater flow patterns within and near a retort field. The
purpose of this description is to locate monitoring wells so as to sample flow
through and from the retort field area. Monitor wells should be located as
follows:
1. Near retort field (within a few hundred feet) and within the
field
2. Oriented downgradient of the MIS retorts along fracture lines
and major axis of anisotropy as defined by geologic testing
program
3. Accessible for sampling equipment.
Construction of new wells may be required for operation/abandonment moni-
toring. Wells constructed for hydrogeologic testing may not be appropriately
located for inclusion in the monitoring program.
Sample Collection Methods
Sampling of deep aquifer wells on Federal Lease Tracts C-a and C-b re-
viewed in this study is accomplished by bailing and swabbing, respectively.
Although these techniques obtain the desired results of collecting a sample,
there is some question as to the representative nature of the sample col-
lected. Some factors contributing to the problem of collecting a representa-
tive sample using these techniques follow.
Problems associated with bailing are:
t The water column chemistry can become stratified due to varia-
tions in water quality and hydrostatic head in the different lay-
ers penetrated by the well. Although this is a function of well
design, nonrepresentative samples will be bailed from this well
if the samples are collected inconsistently with respect to
depth. Water quality data are more representative if samples are
collected consistently adjacent to the water-producing intervals.
The water present in the well casing above the open, or perfo-
rated, section can be isolated from the aquifer water. Samples
collected from this portion of the well will be nonrepresentative
of the aquifer water chemistry.
Small deviations in the sample collection depth can significantly
affect the data when a bailer is being employed. The potential
magnitude of this effect is apparent from the profile sampling
data presented in this report.
11
-------
These potentially negative influences can be alleviated if correct bail-
ing procedures are exercised. The recommended procedure for bailing groundwa-
tsr samples is as follows:
1. Use a flow-through type bailer (e.g., Kemmerer sampler). Bail-
ers that are open at the top and sealed at the bottom do not
have this flow-through characteristic and will generally be
filled with the first water encountered in the well (i.e., wa-
ter near the static water level).
2. Compile well completion data. Of particular importance is the
well diameter, depth to aquifer, aquifer thickness, and total
depth.
3. For shallow wells with very slow groundwater movement, estimate
the well volume from the well completion data and extract at
least one well volume previous to sample collection. For both
shallow and deep wells with rapid groundwater movement, select
a sampling point adjacent to the aquifer.
4. Consistently sample from the same depth and adjacent to the
aquifer during every sampling effort.
5. Measure temperature, specific conductance, and pH in the field.
If these guidelines are followed, bailing is a very effective method for
collecting groundwater quality samples. In addition, bailing is the most
cost-effective approach (see Table 1).
Swabbing a well is a more representative sampling technique than bailing
in that a well volume can be removed prior to sample collection. However,
this technique is very expensive to employ and presents a potential for
contamination.
The following problems are associated with swabbing:
There is high potential for introducing organics into the sample
when oil-field equipment is used. Care must be taken to clean
the swabbing equipment thoroughly.
The amount of water swabbed from a well is difficult to deter-
mine, and can result in obtaining inconsistent and nonrepresenta-
tive samples. If possible, the discharge should be carefully
measured to provide the necessary data for collecting consistent
and representative data.
Swabbing may accelerate plugging of perforations in the well.
t Swabbing is extremely expensive and time-consuming.
Due to these factors, swabbing should not be employed as a sampling
method.
12
-------
For the deep wells to be utilized for monitoring modified in situ re-
torts, pumping is the recommended sampling approach from a technical stand-
point. Pumping allows a greater portion of the aquifer to be sampled,
minimizes the effects well casing or water stratification may have on the sam-
ple representativeness, and reduces the potential for missing or delaying the
observation of mobile pollutant constituents. In addition, a submersible pump
can be fixed in the well or be used as a mobile unit, alternatives which can
be very beneficial to a sample collection program. However, on a cost-effec-
tive basis, the fixed submersible pump is suggested for deep aquifer wells
(see Table 1).
The following procedure is recommended for collecting a representative
sample from a well when using a submersible pump:
1. Compile well construction data, including well diameter, total
depth, and perforated interval, or aquifer interval in an open
well.
2. Measure static water level and estimate well volume.
3. The pump intake should be placed approximately 5 feet above the
open, perforated, or screened aquifer interval.
4. The discharge rate should be maintained at a moderately low
rate to prevent excessive drawdowns in the aquifer and well, as
well as minimizing turbulent mixing in the annulus.
5. At least one well volume should be extracted from the well be-
fore sampling.
6. The parameters most easily monitored in the field are specific
conductance, pH, and temperature. These parameters should be
measured continuously throughout the pumping period. Continu-
ously monitoring these parameters is particularly important for
infrequently sampled monitor wells.
7. A sample should be collected only after the field parameters
have stabilized for a period of time. The data provided in the
text indicate that conductivity is the most representative pa-
rameter of infusion of aquifer water in the well bore or cas-
ing. -However, it is suggested that all of the parameters
(i.e., pH, temperature, and specific conductance) be utilized
to determine representative aquifer water to prevent premature
sample collection due to the failure of field apparatus.
8. The sample should be collected as close to the well head as
possible to avoid potential contamination, precipitation of
solutes, and the loss of dissolved gases.
In addition to providing consistency with respect to pump placement,
field measurements, etc. among the different sampling dates, these recommenda-
tions also provide a means for establishing the sampling protocols for each
13
-------
well. This is an important aspect in that the data presented in Section 3 in-
dicate the duration of pumping required for an equilibrated discharge may vary
substantially from well to well. Therefore, the individual testing of each
well is critical to representative sample collection. In addition, these pro-
tocols should be updated periodically for each well, particularly for wells
with large open intervals.
In addition to the technical aspects, there are some cost considerations
that need to be evaluated. Table 1 provides the approximate sampling costs
for four different sampling methods and corresponding well design necessary
for the sampling tool. The sampling methods evaluated were a fixed submersi-
ble pump, portable submersible pump, bailing, and swabbing. For comparison
purposes, the costs for each sampling method were developed under a quarterly
sampling frequency of 12 Upper and Lower Aquifer wells for a 5-year period.
Based on the data presented in Table 1, it is apparent that the bailing
method is the best approach from a cost perspective. The portable submersible
pump and swabbing methods are very expensive compared with the bailing method
and, therefore, are not recommended. Although the fixed submersible pump
clearly has some economic trade-offs when compared with bailing, there are
some technical advantages to using this approach and the fixed pump should not
be ruled out. The data comparing samples collected by bailing with samples
collected by pumping for deep aquifer wells indicate that more representative
samples are collected via a pump. Therefore, it is recommended that each
method be evaluated according to the type of well design and the overall moni-
toring strategy. It appears that the bailing method works well for the "near-
retort" type of well designs (i.e., wells with fine levels of completion),
whereas a fixed submersible pump provides better results in wells that are
completed over a large interval.
Sampling Frequency
Proper selection of well sampling frequency is a function of potential
pollutant mobility, and when hard data are not available, the selection is of-
ten made by trial and error. Shallow groundwater systems commonly display re-
sponse to seasonal or otherwise cyclic events of recharge and infiltration of
dissolved constituents from the surface. Regional pumping patterns can also
affect the variability of water quality in both deep and shallow wells. Such
variability would necessitate relatively greater sampling frequencies.
The aquifers to be monitored for the impacts of abandoned MIS retorts are
relatively deep and not subject to great variability from recharge events.
Such influence of cyclic events is usually attenuated during slow passage
through the aquifer. Hence, a somewhat low sampling frequency is appropriate.
Another consideration is the sequence of events leading to abandonment,
namely, mine-retort operation, termination of retorting, termination of dewa-
tering, and recovery of aquifer water levels in the mine-retort area. During
the operational phase, particularly when dewatering is appreciable, no re-
leases would be anticipated from the MIS retorts. Thus, low-frequency sam-
pling (e.g., annual) would be adequate. If dewatering is via wells (rather
than strictly from the mine itself), the dewatering wells (sampled
14
-------
individually) may be an acceptable location for sampling. Any groundwater
flow within the retort field during this dewatering phase would be dominated
by and directed toward the dewatering wells. Thus, any appreciable groundwa-
ter movement in the retort interval would be effectively sampled by these
wells.
During the time from cessation of dewatering through stabilization of wa-
ter levels, the groundwater system would be in a state of flux and rapid
changes in water quality may occur. During this period, more frequent sam-
pling is recommended. Initially, monthly sampling is appropriate to establish
patterns of temporal variability. This frequency can then probably be dimin-
ished to semiannual and then perhaps to annual as time trends are established.
Several years may pass before these low frequencies are appropriate.
Sample Preservation and Handling
Delayed receipt of samples at the analytical laboratory and incorrect
preservation techniques can significantly adversely affect sample chemistry.
To prevent any potential sample modification, the following sample preserva-
tion and handling procedures are recommended:
Sample volumes, preservatives, and containers should be selected
according to the EPA-recommended procedures presented in Methods
for Chemical Analyses of Waters and Wastes (U.S. Environmental.
Protection Agency, 1979).
The samples should be filtered in the field through a 0.45-micron
filter before preservation.
Data on past water quality trends should be consulted to detect
any anomalous data during the sampling effort.
Specific conductance, pH, and temperature should be measured in
the field at the time of sample withdrawal. This also applies to
oxidation-reduction potential and dissolved oxygen determina-
tions, if desired.
Accurate field notes should be maintained for future data evalua-
tion. These notes should include: specific times and dates the
activities were performed, water levels, source of sample,
weather conditions, well completion data, sample collection
method, field observations, reason for sampling, field measure-
ments, problems encountered, and the sample collector's identity.
t The samples should be shipped each day from the field to the ana-
lytical laboratory via commercial plane or bus. Both methods are
reliable and inexpensive, and provide reasonable assurance
against prolonged sample storage. If the samples cannot be
shipped and received at the laboratory within 24 hours, on-site
analytical facilities should be provided.
15
-------
t The chain of custody for the sample should be recorded and be as
limited as possible to prevent excessive sample handling, which
can result in shipment and analysis delays. Individuals should
be designated both in the field and at the laboratory to maintain
adequate quality control with respect to sample handling and
analysis activities.
If these procedures are followed, sample handling and preservation tech-
niques should not affect the analytical results.
Selection and Preservation of Constituents for Monitoring
Recommended monitoring constituents for general water quality, major in-
organics, organics, and trace metals are given below.
Sample preservation and handling requirements for these water quality pa-
rameters are dictated by the nature of the constituents to be analyzed. For
the recommended constituents, the holding times listed below are recommended
by U.S. EPA (1974). Bottle requirements (plastic versus glass) are also pro-
vided in this reference. Filtering of samples immediately after collection is
recommended with addition of chemical preservatives in the field at the time
of collection or addition of preservatives to sample bottles prior to initia-
tion of field activities. Preservation techniques include:
Preservative
General water quality constituents
Total dissolved solids
(filterable residue)
Conductance
pH
Alkalinity
Major inorganics
Calcium, magnesium, potassium,
and sodium
Bicarbonate
Carbonate
Chloride
Cool, 4°C
Cool, 4°C
Determine on site
Cool, 4°C
Nitric acid to pH < 2
Cool, 4°C*
Cool, 4°C*
None required
Maximum
holding
time
7 days
24 hours
6 hours
24 hours
6 months
24 hours*
24 hours*
7 days
(continued)
* Assumed same as alkalinity.
16
-------
Preservative
Major inorganics (continued)
Nitrate
Sulfate
Fluoride
Ammonia
Phosphate
Organics
Dissolved organic carbon
Kjeldahl nitrogen
Trace metals
Arsenic
Selenium
Vanadium
Molybdenum
Mercury
Cool, 4°C
Cool, 4°C
Cool, 4°C
Cool, 4°C, sulfuric acid
to pH < 2
Cool, 4°C
Cool, 4°C, sulfuric acid
to pH < 2
Cool, 4°C, sulfuric acid
to pH < 2
Nitric acid to pH < 2
Nitric acid to pH < 2
Nitric acid to pH < 2
Nitric acid to pH < 2
Nitric acid to pH < 2
Maximum
holding
time
24 hours
24 hours
7 days
24 hours
24 hours
24 hours
24 hours
6 months
6 months
6 months
6 months
38 days
(glass
container)
The short holding times listed here will be difficult, if not impossible,
to accomplish in the remoteness of the oil shale region unless on-site labo-
ratory facilities are developed. Such an approach is recommended for the
following:
t Conductance
pH
Alkalinity
t Carbonate
Bicarbonate
t Chloride
17
-------
Ammonia (electrode method)
t Fluoride (electrode method).
Since it may not be feasible to meet the listed holding time requirements
for many of the constituents listed (e.g., IDS, nitrate, sulfate, phosphate,
DOC, and Kjeldahl nitrogen), it is recommended that testing be initiated so
that more suitable holding times for the waters in question can be defined and
the nature and significance of errors evaluated.
Sample Analysis
Recommendations for sample analysis are as follows:
1. Routine monitoring of recommended constituents listed in the
preceding discussion of sample preservation and handling
2. More extensive sample collection and analysis (such as unique
indicators discussed in Slawson, 1980a, Section 10) should the
routine sampling program indicate an impact of MIS retorts on
groundwater quality
3. Use of standard analytical methods.
The constituents listed in the preceding discussion of sample preserva-
tion were selected for routine monitoring because high levels are expected
should materials leach from MIS retorts. In addition, constituents include
those which allow data checks (TDS-conductivity, cation-anion balance, etc.)
to be performed as a quality control measure. Should this routine monitoring
program indicate an impact of MIS retorts on groundwater quality, more exten-
sive analysis of samples is recommended. This analysis should include the
sets of possible unique indicators presented in Slawson, 1980a, Section 10.
This recommended list of constituents includes fewer constituents than
the analysis sets of presently implemented monitoring programs, such as out-
lined in Slawson, 1980a, Section 9. This shortened list should allow detec-
tion of groundwater quality impacts due to MIS retorts while economizing on
analytical needs.
Other sets of constituents, such as various organic fractionations and
stable isotope ratios, need to be evaluated further, particularly the inter-
pretation of such data with regard to indicating the impact of oil shale by-
products. Standard analytical methods, such as presented by U.S. EPA (1974)
or in Standard Methods (American Public Health Association, 1976), should be
emp 1 oyecT
Interpretation of Uater Quality Data
The purpose of interpreting water quality data is to define quality
trends, identify new pollution problems or regions of improvement, and assess
the effectiveness of pollution control activities. To ensure the utility of
the water resource information collected, data analysis procedures include
18
-------
(1) checks on data validity and (2) methods of presenting the resulting infor-
mation so it is useful for environmental description or control purposes.
Data checking procedures include:
Cation-anion balance
TDS-conductivity comparison
Conductivity-ion comparison (meq/1)
t Diluted-conductance method.
Data presentation and interpretation are key aspects of monitoring for
environmental detection and control. Several methods are available for orga-
nization and presentation of water quality data. These include tabulation and
graphical tabulation of appropriate water quality criteria or standards, pro-
viding a format for screening data, and identifying important sites or pollu-
tant constituents. Presentation of ionic concentration as milligrams per
liter or mi 11iequivalents per liter and segmentation of contributing compo-
nents, such as total and noncarbonate hardness or phenolphthalein and methyl
orange alkalinity, are useful techniques for data correlation and evaluation.
Further discussion of data analysis procedures is provided in Section 3.
19
-------
SECTION 3
HYDROGEOLOGIC CHARACTERIZATION METHODS
Much descriptive information and data have been published on the geologic
and hydrologic characteristics of the oil shale regions of Colorado, Utah, and
Wyoming. These studies, however, have been largely regional in scope, leading
to a generalized focus on developmental groundwater quality monitoring plans,
rather than environmental protection site- and source-specific orientations.
The goal of this study has been to develop support information that will pro-
vide a procedure for obtaining valid groundwater quality data to provide an
evaluation and decision-making framework for design of monitoring programs to
protect the environment and water quality at specific development sites. This
study is intended to be a planning document that will provide a technical
basis and a methodology for the design of groundwater quality monitoring pro-
grams for industrial oil shale developers and the several governmental agen-
cies concerned with environmental planning and protection. The Piceance Creek
Basin of Colorado, where the richest oil shale deposits lie and where it is
expected that most future leasing and industrial development will occur, is
discussed in this study. The general procedures and framework for environmen-
tally sound hydrogeologic characterization, however, are valid for other oil
shale regions.
Most of the hydrogeologic characterization methods described in this
study will be employed during the initial exploration/resource evaluation
phase of industrial development. Some methods will be employed during the
mine development phase, while others, such as sample collection for ongoing
water quality monitoring, will be conducted over the entire life of the proj-
ect, including the postclosure period.
To plan, design, and conduct a hydrogeologic characterization program as
a basis for designing a groundwater monitoring strategy, a general understand-
ing of basin hydrogeology is necessary. The following subsections describe
the Piceance Basin hydrogeology.
GENERAL BASIN HYDROGEOLOGY
The area contains three important aquifer systems: the Lower Aquifer,
the Upper Aquifer, and the alluvial aquifers. The Lower Aquifer occurs in the
Parachute Creek Member below the Mahogany Zone, and the Upper Aquifer is above
the Mahogany Zone (see Figure 2). The alluvial aquifer system occurs in the
stream valley bottoms.
20
-------
9.000
I
^v. Lower a
3.000
024 6 KILOMETRES
VERTICAL EXAGGERATION X 21
DATUM IS MEAN SEA LEVEL
Figure 2. Geologic section through Piceance Basin along north-south line between
Tracts C-a and C-b (Weeks et al., 1974).
-------
Lower Aquifer
The Lower Aquifar is bounded generally on the top by the Mahogany Zone
and on the bottom by the shales of the Garden Gulch Member. Porosity is
mostly secondary, resulting from fracturing and jointing of the marlstone and
oil shale of the lower Parachute Creek Member. Porosity also results from the
solution of the evaporite minerals in the saline section at the base of the
Parachute Creek Member. Removal of these soluble minerals by groundwater has
created a zone of high permeability (known as the leached zone) at the top of
the saline section. The saline section below the leached zone still contains
its original salts. Because of the high electrical resistivity of the salts,
which characterizes this zone on geophysical logs, it is called the "high re-
sistivity" (HR) zone. Inasmuch as both the high-kerogen-content oil shales
and the saline minerals of the HR zone are rather ductile, the HR zone has ex-
perienced little fracturing and has a low permeability. Because of these
characteristics, in the center of the basin the HR zone forms the lower con-
fining stratum.
The fracture-solution of this confined aquifer results in heterogeneous
hydraulic characteristics. In general, transmissivity increases with the sol-
uble mineral content from the margins to the center of the basin. The degree
of fracturing, resulting from deformation, increases toward the structural
axis of the basin, and northwest along the axis. Weeks et al. (1974) esti-
mated that the average transmissivity varies from 130 ft^/day near the south-
eastern corner of the basin, to 670 ft2/day in the area between Yellow and
Piceance Creeks. They estimated the storage coefficient to be on the order of
10"4 and the specific yield to be 10~1. Well yields of 200 to 400 gallons per
minute (gpm) are typical.
Upper Aquifer
The Upper Aquifer is separated from the Lower Aquifer by the Mahogany
Zone. Although no interaquifer response was observed during vertical perme-
ability tests, Weeks et al. (1974) have concluded that considerable movement
of water between the aquifers does occur. They base this conclusion on the
fact that the water level in the two aquifers rarely differs by more than 100
feet over the 1,200-foot head drop of the two aquifers across the basin.
The Upper Aquifer zone is composed of the Parachute Creek Member above
the Mahogany Zone and the Uinta Formation. The lower portion of the Uinta
Formation is divided by numerous tongues of the Green River Formation. Al-
though the primary porosity of the sandstones is greater than that of the
marlstones, the sandstone porosity has been decreased by precipitates from
groundwater, while fracturing has increased the permeability of the marl-
stones, which are more susceptible to fracturing than the Uinta sandstones.
The sandstones, therefore, tend to form confining layers for the marlstone
aquifers. The Upper Aquifer is generally confined but is unconfined in many
locations, depending on the relationship of the water level and the lithology.
Strata containing nahcolite (NaHC03) solution cavities, which occur in the
southern part of the basin, should form transmissive layers.
22
-------
The transmissivity (T) varies with saturated thickness, degree of frac-
turing, degree of solution, and location of wells with regard to fractures.
Calculated T values range from 8 to 1,000 ft2/day. The saturated thickness,
degree of solution, and transmissivity increase toward the basin center.
Weeks et al. (1974) considered representative values to be 70 fWday around
the rim, 130 ft2/day in the area around the center, and 270 ft2/day in the
center.
Porosity ranges from 10 percent to 1 percent. It is highest in the cen-
ter, where solution cavities are present, and least around the edges. The
calculated storage coefficient is on the order of 10"J, indicating confined
conditions. The total storage is probably somewhat less than that in the
Lower Aquifer due to lower saturated thickness and porosity.
Since the difference in water level between the two aquifers is rarely
more than 100 feet, the potentiometric map of either the Upper or Lower Aqui-
fer should not differ greatly. The potentiometric configuration is determined
by the transmissivity distribution and the recharge and discharge characteris-
tics. Recharge occurs around the rim of the basin, the gradual infiltration
of snowmelt in the spring probably being the major source. The downward po-
tential difference between the two aquifers around the rim of the basin indi-
cates that most of the recharge is to the Upper Aquifer and that the Lower
Aquifer is recharged by leakage from the Upper Aquifer through the Mahogany
Zone. The water migrates toward the center of the basin, where it discharges
to Piceance and Yellow Creeks at some locations. Here, the head of the Lower
Aquifer is higher than the Upper, and Lower Aquifer discharge is through the
Upper Aquifer.
Alluvial Aquifers
Alluvial sediments line most of the major stream valleys and are usually
saturated at their base. They are thickest along Piceance and Yellow Creeks.
Near the confluence with the White River, there may be 100 feet or more satu-
rated alluvium underlying Piceance Creek. All of these aquifers follow the
slope of their stream valleys. They are recharged in their upper reaches from
streams and from snowmelt. In the lower sections, they are recharged from the
deep aquifers and, in turn, discharge to the streams, maintaining the base
flow.
The hydraulic conductivity of these unconsolidated alluvial deposits is
high, reflected in transmissivities of 2,700 to 20,000 fWday. Their un-
consolidated nature also results in high specific yields, on the order of 20
percent. In spite of these favorable aquifer parameters, the alluvial aqui-
fers are not desirable areas for large-scale water development because of the
small total storage and boundary effects created by the aquifer morphology.
In addition, withdrawal from the aquifers is sure to affect the stream base
flow adversely, and with it agricultural interests, wildlife habitat, and ex-
isting water rights allocations.
23
-------
GEOPHYSICAL METHODS
Phase I study efforts are documented in Slawson (1980b) and summarize
geophysical methods that may be appropriate in defining the hydrogeologic
characteristics in oil shale environments. This comprehensive review includes
a wide range of geophysical well-logging techniques available through major
logging companies. The utility of these geophysical tools for defining hy-
draulic properties in typical oil shale stratigraphy was not addressed in the
Phase I study. This appraisal was conducted as part of the following Phase II
efforts.
Suites of geophysical logs run during the post-leasing exploration stud-
ies on the Federal oil shale tracts in the Piceance Basin have been reviewed.
Log suites for oil shale Tracts C-a and C-b are given in Tables 2 and 3, re-
spectively. These tables show that while similar suites of logs are run on
both of the Colorado tracts, specific logs are emphasized. For example, the
engineering production (spinner) logs perform well and are commonly run on
Tract C-a but seldom, if ever, on Tract C-b. Sonic logs are used extensively
on Tract C-b but only infrequently used on Tract C-a, where three-dimensional
velocity logs provide much of the same acoustic information. Use of alternate
logging tools reflects, in part, individual log response, the information de-
sired from their interpretation, the preference of the geophysical program
coordinator, the logging service company selected to perform the work, and
specific data-gathering requirements externally imposed on the exploration ef-
fort. Therefore, the most commonly run logs indicated in Tables 2 and 3 may
not reflect the most appropriate suite for defining the hydrogeology in any
one oil shale region.
Exploration studies on the Federal tracts are primarily interested in re-
source characterization. Defining the hydrogeologic framework, while impor-
tant to the mine design, is initially of secondary importance. In the Phase
II studies, Tempo reevaluated the geophysical exploration data with definition
of the hydrogeologic framework as a primary focus. Following a review of
these geophysical data and discussions with the major well-logging companies,
a suite of logs has been selected to evaluate the hydrologic characteristics
of test holes in an oil shale environment. This does not imply that a single
suite of logs would be best suited for all boreholes. Unique borehole condi-
tions must be dealt with on a site-specific basis. However, it is instructive
to select a suite of geophysical logs and evaluate their effectiveness in de-
fining the hydrogeologic framework in an oil shale environment. The following
suite of logs has been selected for this purpose:
Temperature log 3-D velocity log
t Caliper log Sonic/acoustic log
t Gamma-ray log Density log
Spinner log Electric log
Radioactive tracer log Seisviewer log.
24
-------
TABLE 2. GEOPHYSICAL DATA COLLECTION, TRACT C-a
Well designations
Gulf-Standard core holes
Monitor holes
Geophysical logs
1 2-3 4-5 6 7 8 9 10 11 12 13 14 15 1
ro
tn
Schlumberger
Dual Induction Laterolog X X
Compensated Neutron Formation Density X X
Borehole Compensated Sonic-Gamma Ray X
Engineered Production (Spinner-Temp)
Continuous Directional
Birdwell
Electric
Gamma-Ray Density Neutron
Three-Dimensional Velocity X
Temperature X X
Spinner X X
Caliper (only) X
Seisviewer X
Nuclear (Ganma-Ray-Neutron)
Density X
Ganma-Ray Density
Continuous Directional Inclinometer
X
X
X
X
X
X
XXX
XXX
XXX
X X
X
X X X X X X X
X X
X X X X X X X
X X X X X X X
X X X X X X X
X X X X XX
XX XXX
XX XXX
X
X X X X
X X X X
X X X X
X X X X
X X X X
X X X X
-------
TABLE 3. GEOPHYSICAL DATA COLLECTION, TRACT C-b
ro
01
Geophysical logs
Schl umber ger
Borehole,
Compensated Sonic
Laterolog
Formation Density
Nuclear
Formation Density
Temperature
Birdwell
Three-Dimensional
Velocity
Electric
Density
Nuclear
Caliper
Temperature
AT-1 AT-la AT-lb AT-ld SG-1
X XXX
X XXX
X XXX
X X
X X X X X
X
X
X
X
X
X
Well Designation
SB-la SB-8 SB-9 SG-10 SG-11 SG-17 SG-18 SG-19 SG-20 SG-21 Cb-1 Cb-2 Cb-3
XX XXXXXX
XX X X X X X
XXX XXXXX
xxxxxxxxx
X
X
X
X
X
X XXX
-------
A discussion of each of these logs has been developed for its specific
use in defining hydraulic parameters of interest. As with all geophysical
studies, conjunctive use of the individual logs is important in improving the
accuracy of the interpretation. In addition, alternative data sources, e.g.,
water production tests and computer analogs that complement the geophysical
record, are used wherever possible.
In compiling the information on current logging methods and sonde instru-
mentation, the logs from four major logging service companies were evaluated:
Birdwell Division of Seismography Service Corporation, Schlumberger Well Ser-
vices, Dresser Atlas, Inc., and Welex, A Halliburton Company. Data developed
during the study was drawn from interviews with logging company personnel, in-
formation sheets and catalogs provided by the companies, and a review of logs
run in wells on Federal Oil Shale Tracts C-a and C-b. Cost data for logging
services was taken from the most current Rocky Mountain price schedules for
the respective companies. Current prices may vary from those quoted in the
text.
Temperature Log
Principle of Operation--
Temperature logs were run on nearly every test hole during exploration
efforts on the Federal oil shale tracts. Temperature logs are made by passing
a temperature electrode down a cased or an uncased hole. Temperature logs use
a sonde with a resistance-type thermocouple or a wire calibrated to correlate
resistance variations with temperature variations. In the former, a junction
of two dissimilar metallic conductors is housed in a protective cage. An
electromotive force is inducted at the junction when the conductors are main-
tained at different temperatures. This force is measured and recorded on
strip charts or a magnetic tape at the surface. In the latter, the sonde uses
a length of platinum wire that rapidly assumes the temperature of the borehole
fluid. Variations in the temperature of the wire produce changes in resis-
tance that are detected at a bridge circuit in the sonde. These signals are
transmitted to a recording device at the surface.
The diameter of temperature sondes range from 1 to 3-5/8 inches and can
be run down boreholes 2 to 20 inches in diameter. The temperature tool is
used in a wide variety of borehole environments including water, mud, oil, or
air.
Two passes of the temperature sonde should be recorded for each test hole
studied. Both runs are made down the borehole. The first measurement should
be made immediately after pulling the drill string and before natural circula-
tion becomes established. The second run should be made at the end of the
logging program. If the drilling fluids have been well circulated, the first
run will provide an indication of the natural geothermal gradient, which can
be used as a reference to compare anomalies from the second pass. Temperature
anomalies show up at varying times after circulation has ceased, depending on
the thermal conductivity of the formation penetrated, the flow rates within
the well bore, and the diameter of the well.
27
-------
On the oil shale tracts, temperature logs provide indications of fluid
entry and exit from the well bore. In subsequent hydraulic studies they can
be used to locate formation waters leaking through casing, which could create
a contamination problem for water quality evaluations. During retorting, tem-
perature logs can be used to detect and monitor excursion events.
Figure 3 is a computer plot of a typical temperature log from Federal Oil
Shale Tract C-a. The log shows the effect of cooler formation water entering
the well bore through small permeable zones between 580- and 850-foot depths.
This water flows down the borehole, depressing the natural geothermal gradient
to a depth of 1,420 feet. Below 1,420 feet, the sharp increase in fluid tem-
perature suggests that the cooler waters have entered a "thief zone" and are
no longer depressing the temperature of the borehole fluid.
Cost Data-
Costs for running temperature logs are computed based on per foot depth
and operation charges. Minimum costs per test hole are based on 2,000 feet of
logged hole. Current price schedules for the four major logging service com-
panies are given in Table 4.
Evaluation-
Temperature logs are useful in providing indications of fluid movement in
well bores and are essential in establishing baseline temperature data. Such
information is utilized in subsequent pollutant migration evaluations or geo-
chemical studies.
Temperature anomalies found on Tract C-a are primarily the result of well
developed flow patterns from the Upper to the Lower Aquifer systems. These
conditions are favorable for deducing hydraulic data from temperature logs.
Tract C-b wells show less anomalous conditions, with many plots reflecting the
natural geothermal gradient of the area. These logs are less instructive.
Caliper Log
Principle of Operation
Caliper logs provide a continuous record of the variation in the diameter
of the uncased drill hole. Several sonde configurations are available, e.g.,
two-, three-, four-, and six-arm devices. The average diameter of the hole is
described by the tips of the arms of the device, which, when extended, contact
the sides of the drill hole. The independent action of each arm, when grouped
into pairs of opposed arms spaced 120 degrees apart, provides a direct mea-
surement of up to three specific borehole diameters. These can be recorded
simultaneously on a strip chart, with or without the calculated average hole
diameter.
The caliper log is run by lowering the sonde to the bottom of the test
section, actuating the arms, and pulling the tool out of the hole. It is com-
monly run with a temperature device or other logging tools.
28
-------
60 t-
55
ol
tc
3
< 50
at
a.
45
40 -
I I I 1 I i 1 I
I I I I I I I I 1 I I I I I I I I I I I I I
I I t II I
_L
I .... I
500 600 700 800 900
1,000 1,100
DEPTH (feet)
1,200 1,300 1,400 1,500 1,600
Figure 3. Computer plot of a typical temperature log from Tract C-a.
-------
TABLE 4. COST SCHEDULE FOR TEMPERATURE LOGS (dollars)
Depth
Company/Service
Birdwellb
Temperature
Differential temperature
per foot
0.22
0.22
minimum3
440.00
440.00
Operation
per foot
0.19
0.19
minimum3
380.00
380.00
Total
minimum3
820.00
820.00
Schlumberger0
High resolution 0.26
temperature
Dresser Atlas
Differential temperature 0.28
Welex6
Precision temperature log 0.25
520.00 0.21
560.00
500.00 0.20
420.00 940.00
750.00 1,310.00
400.00 900.00
Notes:
aAll the service companies have a 2,000-foot minimum.
Birdwell Rocky Mountain Price Schedule, June 1980.
GSchlumberger Rocky Mountain Price Schedule, October 1979.
Dresser Atlas Rocky Mountain Price Schedule, July 1980.
Welex Rocky Mountain Price Schedule, January 1980.
Caliper logs are primarily used to determine the volume of the drill hole
and thus the annular space between the casing and the well. They are useful
in identifying and permitting the correlation of nonround boreholes from well
to well. In hydraulic testing, caliper logs are useful in the selection of
competent beds required for setting packers. In general, determination of ac-
curate borehole diameters is essential for quantitative interpretation of pro-
duction engineering (spinner), electric, acoustic, density, and radiation
logs. In oil shale stratigraphy, they are useful in locating soft, friable,
or fractured zones, which are associated with porous and permeable beds.
The caliper sonde ranges in diameter from 1-5/8 to 3-5/8 inches for the
three- and the six-arm tool, respectively. The smaller tool can be operated
in a 3- to 30-inch-diameter borehole, while the larger tool requires a minimum
hole diameter of 6 inches. They operate equally well in air-, mud-, oil-, and
water-filled holes.
30
-------
Cost Data
Service company costs for running the caliper log is computed based on a
per foot depth and operation charge. Minimum charges, based on 2,000 feet of
logged hole, are given in Table 5.
TABLE 5. COST SCHEDULE FOR CALIPER LOGS (dollars)
Depth
Operation
Total
Company/Service per foot minimum3 per foot minimum3 minimum5
Birdwell
Caliper (3-arm) 0.22 440.00 0.19
Caliper (6-arm) 0.22 440.00 0.17
380.00
340.00
820.00
780.00
Schl umber ger
Caliper (all) 0.26 520.00 0.21
Dresser Atlas
Caliper (4-arm) 0.26 520.00 0.21
Welex6
Caliper (4-arm) 0.20 400.00 0.18
Notes:
A 1 1 r*S*WB>*4>«*« >* AM*M * M 4 A «* L* «>1 A 4 O C\f\f\ £ f*.fH± * 4 M 4 *> I*M
420.00 940.00
420.00 940.00
360.00 760.00
Birdwell Rocky Mountain Price Schedule, June 1980.
Schlumberger Rocky Mountain Price Schedule, October 1979.
Dresser Atlas Rocky Mountain Price Schedule, July 1980.
eWelex Rocky Mountain Price Schedule, January 1980.
Evaluation
In the dense, tight, oil-shale stratigraphy, fracture patterns control
the secondary porosity and permeability within the formation. These fractured
areas lead to zones of weakness in the borehole that may be subject to caving
or raveling. Caliper logs are designed to detect these out-of-gage portions
of the drill hole and therefore provide indirect information on porosity and
permeability in oil shale environments. To evaluate this relationship, pump
tests and spinner logs were compared to caliper logs. Below the static water
level in the Federal oil shale tract wells, permeable zones correspond well
with borehole enlargement due to caving (see Figure 5, page 41); i.e, lower
permeabilities were found throughout in-gage sections of the holes, while
31
-------
large washouts, in general, corresponded to zones of higher permeability. Al-
though the comparative evaluation is only qualitative, caliper logs are useful
in directing hydraulic test programs to potential zones of permeability for
further injection or production testing.
Gamma-Ray Log
Principle of Operation--
Gamma-ray logs measure emissions from natural radioactive materials found
in all rocks. When the gamma rays emitted from the formation penetrate the
sonde detector, usually a scintillometer, an electrical pulse is produced and
transmitted to the surface recorder through electrical cables. The gamma-ray
log is thus a curve relating depth to the intensity of natural radiation. Be-
cause clays and shales are considerably more radioactive than carbonates,
i.e., limestone or dolomite, and sandstone, this geophysical tool is espe-
cially useful in "fingerprinting" lithologic sequences that are correctable
across well fields.
The gamma-ray sonde measures 1-5/8, 2-3/4, and 3-5/8 inches in diameter,
depending on the company and tool selected, and can be used in a 2- to 15-
inch-diameter borehole. The gamma-ray tool operates in all test hole environ-
ments and is effective in cased and uncased holes. In cased holes, it is
combined with a casing collar locator for depth control while measuring com-
plementary parameters and providing information on the cased lithology. In
uncased holes, it can be combined with temperature, density, caliper, and
other types of neutron logs.
Cost Data--
Service company costs for gamma-ray logging are divided into depth and
operation charges with minimum fixed prices per hole. Price schedules for the
four major logging companies are given in Table 6.
Evaluation
The gamma-ray log is run separately or in combination with other logging
tools on nearly every test hole on the Federal oil shale tracts. It is pri-
marily used for lithologic correlation between tract wells and for depth con-
trol. Detailed core analysis for tract wells provides a better source for
lithologic information, however, and supersedes the data from gamma-ray logs.
The shale correction factor determined from the gamma-ray log is the key pa-
rameter in the petroleum industry but is not useful to the oil shale industry.
Spinner Log
Principle of Operation
The spinner, or engineering production, log measures vertical flow in the
borehole. The sonde consists of a propeller-type blade mounted to rotate
about a vertical axis. Rotation of the blade is measured in counts per minute
as a magnetic coupling passes a fixed reference point on the shaft, sending an
32
-------
TABLE 6. COST SCHEDULE FOR GAMMA-RAY LOGS (dollars)
Depth Operation
Total
Company/Service per foot minimum3 per foot minimum3 minimum3
Birdwellb
Schlumberger0
Dresser Atlas
Welex6
Notes :
0.22
0.26
0.26
0.20
440.00
520.00
520.00
400.00
* O AAA _ff..~.J.
0.19
0.21
0.22
0.18
.
380.00
420.00
440.00
360.00
820.00
940.00
960.00
760.00
Birdwell Rocky Mountain Price Schedule, June 1980.
cSchlumberger Rocky Mountain Price Schedule, October 1979.
Dresser Atlas Rocky Mountain Price Schedule, July 1980.
eWelex Rocky Mountain Price Schedule, January 1980.
electrical pulse to a surface recording station. Counts per minute are con-
verted to flow past the sonde based on hole diameter, blade size, configura-
tion etc. Measurements can be recorded with the sonde in a fixed position or
while it is being lowered into or pulled out of the borehole. When recordings
are made in the fixed position, the vertical flow rate in the hole must be
sufficient to overcome the mechanical friction of the tool. This minimum flow
rate will vary depending on the configuration and general condition of the
sonde, the size of the blade, and the diameter and degree to which the hole is
in gage. For example, experience has shown that a sonde with a 4-inch diame-
ter blade inserted in a 5-inch hole requires a flow rate of approximately
5 ft/min to overcome mechanical friction, and give an accurate measurement in
a fixed position. To minimize the effect of friction and measure small flow
rates, the sonde is moved up and down the well at a constant rate.
Unlike temperature, caliper, or gamma-ray logs discussed earlier, the
spinner log is a qualitative rather than a quantitative tool, requiring care-
ful calibration for each test hole. Calibration charts can be developed by
plotting counts per minute (cpm) versus logging speed. Figure 4 shows a cali-
bration plot for a hypothetical 5-inch-diameter hole. Data for construction
of the plot were gathered as follows:
t A gaged section of the borehole is selected, based on the caliper
log
t Preliminary up- and down-hole spinner measurements are made to
ensure that there is no vertical flow in the test section
33
-------
Three or more passes up and down the test section are made at
varied logging speeds and the cpm readings recorded for each pass
Counts per minute versus logging speed plot is constructed as
shown in Figure 4.
UJ
D
oc
UJ
a.
V)
t-
O
o
i r
i i i f
THRESHOLD OF
MECHANICAL FRICTION
I I I I I
DOWN-HOLE MEASUREMENT
UP-HOLE MEASUREMENT
I I I I I I I
10 20 30 40 50 60 70
LOGGING SPEED (ft/min)
80
90
100
110
Figure 4. Spinner log calibration plot.
Calibration plots can also be made for specific out-of-gage borehole con-
ditions if significant fluid production is suspected from a given strati -
graphic horizon. In this case, a static test section with similar borehole
characteristics can be used for construction of the calibration plot. In gen-
eral, sections with high rugosity produce turbulent flow in the well bore and
are extremely difficult to accurately calibrate.
Following construction of the calibration plot(s), a single run of the
spinner tool should be sufficient to determine vertical flow velocity in the
well. However, if the sonde is moved in the same direction and at approxi-
mately the same rate as the borehole, fluid mechanical friction of the tool
will not be overcome and inaccurate flow measurements will result. This can
be overcome by recording flow rates while moving the tool both up and down the
hole at a constant rate. Comparison of the two velocity versus depth logs
would show the velocity and direction of fluid movement more clearly.
Cost Data-
Costs for running a spinner survey are calculated based on the depth, op-
eration expenses, and the number of passes that are made up and down the hole.
These costs are given in Table 7.
34
-------
TABLE 7. COST SCHEDULE FOR SPINNER SURVEYS (dollars)
Depth
Operation
Total
Company/Service
per foot minimum3 per foot minimum3 minimum3
Birdweir
Spinner survey0 0.22 440.00
Schlumberger
Continuous flowmeter6 0.27 540.00
Second pass (in 0.22 440.00
combination)
Dresser Atlas
Spinner Flolog 0.28 560.00
Additional runs 0.19 380.00
Welex9
Spinner" 0.20 400.00
350.00
740.00
620.00
790.00
1,280.00
1,060.00
750.00 1,310.00
550.00 930.00
0.20
400.00
800.00
Notes:
aAll service companies have a 2,000-foot minimum.
bBirdwell Rocky Mountain Price Schedule, June 1980.
cIncludes one recorded run down and one recorded run up. For additional
recordings at different logging speeds, add $0.07/ft, $105.00 minimum.
Schlumberger Rocky Mountain Price Schedule, October 1979.
elf more than one descent is made into a well with the same tool, each
descent is considered a separate service and charged at the single service
rate.
fDresser Atlas Rocky Mountain Price Schedule, July 1980.
9Welex Rocky Mountain Price Schedule, January 1980
Available in limited areas.
Evaluation
Semiquantitative information can be developed from spinner surveys on a
site-specific basis when calibration plots are carefully constructed. How-
ever, these data are dependent on the hydraulic head relationship and there-
fore on the dynamic flow characteristics of the permeable beds penetrated and
interconnected by the well bore. For example, quantitative flow data can be
derived from spinner logs run in boreholes that intercept two permeable zones
with sufficiently different hydraulic heads to allow flow from one zone to
35
-------
another. This condition exists on Tract C-a, where water flows in response to
potential differences from the upper to the lower permeable zones. The flow
measured, however, does not necessarily reflect the true ability of an aquifer
to produce or accept fluid from the borehole, but rather provides information
on the existing flow system and provides lower limits of permeability and wa-
ter production. Likewise, if two highly permeable beds are interconnected by
a well and have nearly equal hydrostatic heads, the spinner survey provides
little information on the aquifer hydraulics since no flow would occur in the
well. This is perhaps one reason why spinner surveys are not as useful nor as
commonly run on Tract C-b wells as on Tract C-a wells.
Radioactive Tracer Log
Principle of Operation
The radioactive tracer sonde consists of an ejector that extrudes a
1, 1^
short-lived radioactive source (1, 1) ^n^0 the borehole and one or two
detecting elements. If a single-element sonde is used in logging, the radio-
active source is emitted and the detecting element is moved through the source
to determine its location in the borehole. After a short period of time, the
detecting element is again moved up or down the hole to locate the source, and
from the time-distance relationship the flow rate in the well can be calcu-
lated. With two detecting elements at fixed distances on the sonde, the
source material is ejected and detected at the same time, and the tool does
not have to be moved, thus reducing dispersion of the source and increasing
the peakedness of the log trace, hence providing greater accuracy in locating
the radioactive material in the well. With this type of tool, the source
ejector can be located at either the top or bottom of the sonde to measure
flow up or down the borehole.
The borehole instrument comes in 1- and 1-5/8-inch diameters and can be
run in 1-1/2- to 12-inch-diameter wells. It will operate in all fluid-filled
holes.
Cost Data-
Service charges for running radioactive tracer logs are given in Table 8.
These include standard per foot and operating costs, as well as radioactive
material ejector fees.
Evaluation
The accuracy of the radioactive tracer log depends on the peakedness of
the source-detecti on-versus-depth plot. This is primarily a function of know-
ing where the source is in the borehole. While tracer logs do not have the
mechanical friction problems inherent in the spinner tool, extremely low bore-
hole velocities provide time for diffusion of the source material and spread
of the radiati on-versus-depth plot, thereby limiting the accuracy of the mea-
surements. Also, turbulence associated with higher flow rates tends to dis-
perse the source material, especially in permeable areas where rugosity is
often a significant characteristic of the borehole.
36
-------
TABLE 8. COST SCHEDULE FOR RADIOACTIVE TRACER LOGS (dollars)
Depth
Operation
Company/Service
Total
per foot minimum3 per foot minimum3 minimum0
Birdweir
Radioactive tracer profilec
_d
.e,f
Schlumberger
Radioactive tracer5
Second run
Dresser Atlas9
Tracelog
Additional runs
Welex1
Radioactive tracer^
0.22
0.27
0.22
0.28
0.19
0.20
440.00
540.00
440.00
560.00
380.00
400.00
0.19
0.14
380.00
740.00
520.00
750.00
550.00
280.00
820.00
1,280.00
1,080.00
1,310.00
930.00
680.00
Notes:
aAll service companies have a 2,000-foot minimum.
bBirdwell Rocky Mountain Price Schedule, June 1980.
cRadioactive material ejector charge: $150.00 for the first ten stations and
$0.11 per station thereafter.
dSchlumberger Rocky Mountain Price Schedule, October 1979.
eWith radioactive tracer logging, an added charge of $112.00 per ejection of
radioactive material is applied when a down-hole ejector tool is used.
Radioactive material is charged at cost plus 10 percent handling charge.
^Dresser Atlas Rocky Mountain Price Schedule, July 1980.
Tracer dump bailers $130.00 per run.
nWelex Rocky Mpuntain Price Schedule, January 1980.
^Radioactive material not included in price.
A significant disadvantage of the tracer log is the inherent danger in
handling and the consequence of losing the radioactive source material
(thereby contaminating the well). This is especially acute when working with
shallow-water supply wells or in areas where groundwater may be transmitted
directly into underground mine works. Attempts have been made to minimize
this problem by using radioactive substances with a relatively short half-life
half-life = 8 days). In general, the potential danger versus the
37
-------
qualitative or semiquantitative information gained does not warrant the use of
radioactive tracer logs in groundwater wells.
Laboratory or supplier preparation of the source material requires time,
which can result in delays in the field. If this method is used, careful
planning must be made to coordinate drilling schedules and running other logs.
Three-Dimensional Velocity _L_o_g_
Principle of Operation
Birdwell's single-receiver velocity sonde provides a record of the com-
plete acoustic wave train as propagated along the fluid-borehole boundary of
the well. The total wave train is displayed as variable density, black lines
(legs) on a strip chart and includes the congressional, shear, and boundary
waves. The sonde contains a magnetostrictiye-type transmitting transducer
that generates pulses at a rate of 20 per minute. The ceramic receiving
transducer (a barium titanate crystal) converts the signals transmitted along
the borehole to electrical impulses that are transmitted to a receiver at the
surface and recorded.
The three-dimensional (3-D) velocity log is used in fracture studies, po-
rosity determinations, cement bond evaluations, and in the study of dynami-
cally determined elastic properties of rocks. In the latter, congressional
and shear waves are used in the calculation of elastic moduli (shear, bulk,
and Young's) and in determining Poisson's ratio. Elastic properties are use-
ful in oil shale mine design and are also used extensively in other types of
construction projects.
The sonde diameter varies from 1-3/4 to 3-3/4 inches and can be utilized
in test holes from 3 to 18 inches in diameter. The tool requires a fluid for-
mation boundary to transmit the acoustic wave train; water, mud, or oil medi-
ums are acceptable.
Cost Data--
Service company costs for running 3-D velocity logs, or equivalent, are
given in Table 9. Welex's fracture-finder microseismogram log provides forma-
tion information similar to Birdwell's 3-D velocity log. The cement bond/
variable density log is Schlumberger's closest equivalent to the 3-D velocity
log, but is specialized to determine the effectiveness of the cement seal in
the casing-formation annulus and does not give comparable information.
Dresser Atlas did not have a 3-D velocity log listed in its wireline service
catalog.
Evaluation
The 3-D velocity log provides valuable information on the elastic proper-
ties of rocks useful in mine design, and it is one of the few down-hole geo-
physical tools that provides a complete record of the acoustic wave train.
However, in hydrology studies where porosity determinations are of primary im-
portance, shear and boundary waves are not required. Variation in the
38
-------
TABLE 9. COST SCHEDULE FOR 3-D VELOCITY LOGS (dollars)
Depth Operation
_ Total
Company/Service per foot minimum3 per foot minimum3 minimum3
Birdwellb
3-D velocity 0.29 580.00 0.25 500.00 1,080.00
Schlumbergerc
Cement bond/ 0.26 520.00 0.25 500.00 1,020.00
variable density log
Welexd
Fracture-finder 0.27 540.00 0.23 460.00 1,000.00
Microseismogram log
Notes:
aAll service companies have a 2,000-foot minimum.
Birdwell Rocky Mountain Price Schedule, June 1980.
cSchlumberger Rocky Mountain Price Schedule, October 1979.
eWelex Rocky Mountain Price Schedule, January 1980.
interval transit time (At) of the compressional wave along the fluid-forma-
tion boundary provides the At value needed for porosity calculations. This
travel time is a function of the rock and fluid properties in the borehole as
well as the distance between the detectors. While the detector spacing in the
sonde is fixed, the overall travel distance depends on the rugosity of the
well bore. Smooth in-gage sections produce the shortest travel distances,
while washouts or out-of-gage sections produce longer travel distances for a
fixed set of receivers. The 3-D velocity log is not designed to compensate
for this variation in travel distance and thus will introduce error into the
At values and hence the porosity measurements calculated from these data.
Therefore, in hydrology studies where porosity determinations are of primary
interest, specially designed, compensated acoustic logs are recommended. This
type of log is discussed below.
Acoustic Log
Principle of Operation
The acoustic, or sonic, sonde consists of two sections. The upper sec-
tion houses the electronic equipment necessary to control and activate the
transmitting transducers that convert electrical impulses to acoustic pulses.
Pressure waves created by the acoustic pulses radiate out from the sonde, are
refracted through the formation, and return to the borehole instrument through
the drilling fluid. The lower section contains both transmitting and
39
-------
receiving transducer component in a rigid, slotted metal sleeve. The sleeve
is specially designed to separate acoustic energy transmitted through the in-
strument from signals received from the formation. For a single compensated
sampling point, At is computed through selectively combined time signals
from the receiving transducer array.
Changes in borehole diameter and misalignment of the instrument axis have
signification implications on the accuracy of the acoustic sonde. For limited
variations in borehole diameter or instrument misalignment, multiple trans-
ducer arrays have been developed to ensure accurate measurement recordings.
This is accomplished through a surface panel capable of combining and averag-
ing signals from two transducer arrays, inverted with respect to one another,
for the same borehole interval.
The diameter of the sonde varies from 3-3/8 to 3-3/4 inches, depending on
the tool selected. It is run in fluid-filled, open holes ranging from 6 to 18
inches in diameter at logging speeds of 30 to 80 ft/min.
Measurement of interval transit time is the primary purpose of the com-
pensated acoustic log. Interval transit time may be used to determine poros-
ity using the following equation:
- . (1)
'
Atf -
where: $ is porosity (dimension less)
At is interval transit time (usec/ft)
Atma is matrix interval transit time (usec/ft)
Af is fluid interval transit time (usec/ft).
Service companies provide automatically computed and recorded porosity
values given the desired fixed matrix and fluid velocities using the above re-
lationship. The utility of the porosity measurements for hydrology studies
and their correlation to permeable zones has been evaluated for a test section
located on Federal Oil Shale Tract C-b.
The test zone includes 800 feet of oil shale stratigraphy penetrated by
Tract C-b Well 32x-12 (see Figure 5). Water production from pump and spinner
tests and borehole enlargements from caliper logs are presented with a poros-
ity analog developed from Equation 1. The test zone was selected for its geo-
physical logs, varied lithology, and water production data. Stratigraphic
features included in the section are as follows:
220 feet at the base of the Uinta Formation (±800 to 1,020 feet)
Top of Parachute Creek (±1,020 feet)
Four Senators Zone (±1,100 feet)
40
-------
240.0
100
80
I
PRODUCTION ZONES
FROM PUMP/SPINNER TESTS (gpm)
BOREHOLE ENLARGEMENT
IN EXCESS OF 4 inches
(FROM CALIPER LOG)
<*
o
3
cc
£ 40
20
125
V///////A
BED1
800
18.0
20-0
BED 2 BED 3
180.0
38.0
74.0
50.0
BED 4 6 6 BED? BEDS
220.0
64.0
I
50.0
68.0
BED 9 10 11 BED 12
900
1,000
1,100
1,200
DEPTH (feet)
1,300
1,400
1,500
1,600
Figure 5. Acoustic porosity analog and aquifer production zones for
Tract C-b Well 32x-12.
-------
A-groove (±1,310 feet)
Mahogany Zone (±1,400 feet)
B-groove (±1,500 feet)
Top part of R-6 Zone (±1,520 to 1,600 feet).
The porosity analog was developed from interval transit times taken from
a Birdwell acoustic/borehole compensated log. The matrix interval transit
time, taken from a graph of oil shale yield versus time developed by Birdwell,
was set at 59 usec/ft. Varying this parameter shifted the porosity axis (y-
axis of the analog plot) but did not affect the relative magnitude of the po-
rosity values. As can be seen in Equation 1, decreasing the matrix interval
transit time will increase the porosity values when the other variables are
held constant. The fluid interval transit time was set at 198 usec/ft, an
average value for fluids in oil shale test holes (oral communication with
Mr. Asher Atkinson, Rocky Mountain Regional Manager for Birdwell Division).
Increasing this parameter increases the denominator of the porosity equation,
thus decreasing porosity values. Again, the shift in the axis does not affect
the relative magnitude of the calculated porosity values.
Porosity values for the upper part of the test hole (between 400 and 800
feet; not shown on Figure 5) are uniform, averaging about 32 percent void ra-
tio. These values appear to be high and are probably the result of a rela-
tively low average At^ value (held constant in Equation 1) for the tlinta
Formation and the uniform, but oversized, borehole diameter. Below a depth of
880 feet, the caliper log shows the hole returning to gage (10-3/4 inches) and
more variation in the porosity is observed. The prominent spike between
Beds 1 and 2 is a washout of probable high porosity that was too large for ac-
curate measurements, even with the averaging of signals from the transducer
arrays. Narrower washout features (shown at the base of the y-axis) are for
the most part eliminated from the porosity analog through transit-time signal
averaging. Some of the features may represent solution cavities that cause
high rugosity, which, with a continuous matrix framework, will transmit the
acoustic energy as if through solid rock.
The correlation between permeable production zones and the porosity ana-
log is complex. High water production from Bed 9 (between 1,393 and 1,450
feet) corresponds to a relatively wide band of high porosity values. The ap-
parent porosity appears to be a combination of the rich oil shale beds (Mahog-
any Zone) and true secondary porosity created by solution breccia zones and
fracture breccia, or "rubble" beds. The rich grades of oil shale tend to in-
crease At, thus increasing the calculated porosity when Atma remains constant.
The three prominent porosity peaks within and slightly above Bed 9 correspond
to washout zones on the caliper log and solution or breccia horizons on the
lithologic log. Here, partings and solution cavities must contribute signifi-
cantly to the void space in the rock matrix. Based solely on the relatively
low porosity calculations, Bed 7 (between 1,222 and 1,247 feet) cannot be ex-
pected to produce the large quantities of water shown in Figure 5. The
caliper log for Bed 7, however, shows three narrow washout zones, two corre-
sponding to fracture "rubble" breccia horizons. In addition, the rock
42
-------
fracture and partings log shows a large number of major fractures within the
bed. In this case, the permeability may be created by partings that are not
large enough to significantly increase the void ratio of the matrix, or the
partings may be oriented vertically and do not influence the speed of the
acoustic waves. An alternate explanation could be that solution cavities are
interconnected through an otherwise consolidated matrix. Bed 5 (between 1,135
and 1,145 feet) is producing from a narrow, highly fractured zone, with no
core recovery found within the bed (represented by a spike in the porosity
analog).
In general, the porosity analog shows high porosity values for the entire
test section. This is probably due to the relatively low fixed matrix travel
time for the varying grades of oil shale.
Cost Data-
Service company price schedules for running acoustic/sonic logs are given
in Table 10. These costs are broken down into depth and operation charges.
Evaluation-
Porosity calculations from acoustic/sonic log interval transit times
should be considered semiquantitative and used with an understanding of the
parameters that interact to yield these data. The matrix interval transit
time, held constant in constructing the porosity analog, can vary signifi-
cantly with a change in oil shale yield from 10 to 35 gal/ton. This could
cause a large error in the porosity calculation. Fluid interval transit time,
held constant in Equation 1, will also vary with temperature, pressure, and
amount of dissolved salts in the well fluid. However, these parameters pro-
duce less change in Atf in shallow borehole conditions and can generally be
disregarded. In addition, the interval transit time can be affected by ex-
treme borehole rugosity, as shown in Figure 5, even with the compensating re-
ceiving arrays of the acoustic sonde.
Of the parameters discussed above, changes in the grade of the oil shale
are believed to produce the largest single variation in the computed porosity.
Utilizing Fischer analysis to determine oil shale grade, and hence an approxi-
mate interval transit time, more quantitative porosity calculations can be
made by varying At«a with depth in Equation 1. Unfortunately, Fischer analy-
ses for Well 32x-12 and most of the other test holes on the Federal tracts are
confidential information and therefore were not available for study. It may
be of interest to tract developers who have access to Fischer analysis to cal-
culate porosity values varying At^a with depth and compare this analog with
water production in the well bore.
Density Log
Principle of Operation
The density sonde consists of a gamma-ray source (usually cesium-137),
two gamma-ray detectors, a caliper arm used to force the source/detector
against the well bore, and electronic equipment required to transmit data to
43
-------
TABLE 10. COST SCHEDULE FOR ACOUSTIC/SONIC LOGS (dollars)
Depth
Operation
Company/Service
per foot minimum3 per foot minimum3
Total
minimum3
Birdweir
Acoustic/borehole 0.29 580.00 0.25
compensated
Schlumberger0
Sonic/borehole 0.29 580.00 0.27
compensated
Dresser Atlas
Borehole compensated 0.20 580.00 0.27
acoustilog-caliper
Welex6
Compensated acoustic 0.27 540.00 0.23
velocity
500.00 1,080.00
540.00 1,120.00
540.00 1,120.00
460.00 1,000.00
Notes:
aAll service companies have a 2,000-foot minimum.
Birdwell Rocky Mountain Price Schedule, June 1980.
GSchlumberger Rocky Mountain Price Schedule, October 1979.
Dresser Atlas Rocky Mountain Price Schedule, July 1980.
eWelex Rocky Mountain Price Schedule, January 1980.
the surface panel. The source and detectors are shielded with heavy metal to
ensure that the signal received is primarily from gamma rays that have trav-
eled through the formation.
The count rate of gamma rays reaching the detectors is inversely propor-
tional to the number of electrons per unit volume of the formation between the
source and detectors. Therefore, the number of gamma rays per second reaching
the detector is a function of the bulk density of the formation. A compensat-
ing effect of the sonde is the short and long spacing of the detector relative
to the gamma-ray source, which reduces error caused by borehole rugosity, and
a perturbation created by the change in density of the mud cake relative to
the formation on the borehole wall.
The density log is primarily used to measure formation porosity. Logging
service companies provide automatically computed and recorded porosity values
from the compensated bulk density measurements. The relationship used to cal-
culate porosity is as follows:
44
-------
where: <|> is the porosity
Ama is the density of the formation matrix
ib is the bulk density measured by the logging tool
fcp is the density of the formation interstitial fluid.
The utility of these porosity data and their correlation with permeable
zones has been evaluated for a test section on Tract C-b Well 32x-12. This is
the same section used to study calculated porosity values from the acoustic
log.
Figure 6 shows the density porosity analog, aquifer production zones from
pump/spinner tests, and borehole enlargements (washouts) from a caliper log in
a format similar to Figure 5. This analog was developed using 2.52 gm/cc as
the fixed matrix density and 1.00 gm/cc for the interstitial fluid density.
The matrix density was derived from a graph developed by Birdwell relating oil
shale yield in gallons per ton to matrix density in grams per cubic centime-
ters. The value of 2.52 gm/cc represents the extrapolated density of oil
shale rock with a yield of zero gallons per ton. This is an equivalent den-
sity value to the matrix transit time used in computation of the acoustic po-
rosity analog. Like the acoustic porosity analog, varying the numerical value
of the matrix density does not alter the relative calculated porosity values,
it simply shifts the porosity (y-axis) of the plot, other parameters held con-
stant. Thus, increasing the matrix density will increase the porosity for a
given bulk density reading.
In general, the density-derived porosity analog appears to reflect secon-
dary porosity and its associated permeability more closely than the acoustic
analog. Nearly all the poorly consolidated fracture/rubble zones, indicated
by washouts on the caliper log, or zones of poor core recovery have been re-
corded as porosity peaks on the analog. These peaks correspond with beds of
high water production and suggest alternate horizons that should be considered
for inclusion in the permeability testing. For example, the prominent poros-
ity peak beween Beds 8 and 9 (Figure 6) should have been included in a packer
permeability test as it appears to have the potential of producing a signifi-
cant quantity of water. Smaller, less prominent peaks between Beds 1 and 2
and Beds 11 and 12 should also have been considered for inclusion in the hy-
drology testing program.
In the dense, tight, oil shale rocks, secondary porosity (vuggy solution
cavities or fraction zones) produces the principal groundwater flowpaths. In
sections where secondary porosity exists, a density or neutron porosity analog
should read higher than the acoustic porosity analog. The difference between
the two porosity values has been defined as the secondary porosity index
(SPI). This index exists because acoustic logs ignore vuggy solution porosity
since a continuous path for the acoustic energy exists through the solid for-
mation matrix. In comparison, density or neutron logs respond to bulk-volume
45
-------
100
80
I
AQUIFER PRODUCTION ZONES
FROM PUMP/SPINNER TESTS (gpm)
BOREHOLE ENLARGEMENT
IN EXCESS OF 4 inches
(FROM CALIPER LOG)
240.0
220.0
180.0
(VI
O
i
DC
O
O.
60
12.5
18
40
v///////\ E2) Y/////A
20
BED1
800
900
1.000
1.100
1,200
DEPTH (feet)
1,300
1,400
1,500
1,600
Figure 6. Density porosity analog and aquifer production zones for
Tract C-b Well 32x-12.
-------
porosity. For secondary fracture porosity, the bulk-volume porosity added by
the fracture system is small unless the zone is extensively rubblized, and the
SPI will not provide useful information.
Porosity analogs (density and acoustic) for the test section in Well 32x-
12 were computed with equivalent matrix characteristics so that the SPI could
be evaluated. Comparison of Figures 5 and 6 shows that the acoustic porosity
values are, in general, higher than the density porosity values. This rela-
tionship is more clearly shown in Figure 7 for Section 1 of a Birdwell elastic
property log for Well 32x-12. The computed porosity values in Figure 7 will
not correspond with Figures 5 and 6 because porosity in Figure 7 was calcu-
lated with apparent sandstone unit parameters as follows: matrix density,
2.62 gm/cc; fluid density 1.00 gm/cc; matrix interval transit time, 192 usec/
ft. However, the same general trends occur when the acoustic porosity is
greater than the density porosity. This is an anomalous situation, for poros-
ity calculations from density logs should represent the total matrix porosity
and be greater than the acoustic porosity. It appears that for the rich oil
shale rock, the large volume of organic material included in the matrix in-
creases the bulk density readings and thus reduces the calculated porosity
more than it affects the transit travel times used in the acoustic porosity
determinations. Porosity from density measurements are larger than acoustic
porosity in breccia zones (washouts on the caliper log), where secondary po-
rosity is extremely high (see Figure 7). Hence, the SPI values (shaded areas
in Figure 7) correspond to production test beds rather well and suggest where
additional packer permeable tests might have been run, i.e., shaded zone above
Bed 7.
A porosity analog was computed from bulk density measurements taken in
Tract C-a, Well CE-705A. Apparent limestone unit parameters (^ equal to
2.69 gm/cc, and if equal to 1.00 gm/cc) were used in the porosity calcula-
tions. Figure 8 shows the resulting analog along with a spinner survey for
th.e same section. The spinner survey was constructed so that the step-like
incremental change in water production or intake was positioned at the first
increase in slope of the log trace for water production and at the base of the
slope for thief zones. For this log presentation, water production zones will
be located down-hole from the step-wise increase in the spinner log or up-hole
from a step-wise decrease, given the established flow direction down-hole.
Qualitative evaluation of these logs shows a partial correlation between water
production/thief beds and high porosity values. However, a nearly perfect
correlation (except for Zone R-6) is found when porosity values are compared
to rich oil shale zones shown at the base of the y-axis. Again, it is well
known that the porosity analog is strongly influenced by oil shale grade.
Cost Data-
Cost information from four major logging companies that run formation
density logs is given in Table 11.
Evaluation
Bulk density measurements taken from the density log can be used directly
for cross-correlation of wells or test holes throughout the exploration phase
of an oil shale mine development program. Porosity analogs developed from the
47
-------
-p.
00
DENSITY
POROSITY ::; 1M. n
-4f-^-H---^
Figure 7. Birdwell elastic properties log for Well 32x-12, Tract C-b.
-------
617
POROSITY (x 10-2)
§ 8
1111 I 1111 I II1111111 11 III
CUMULATIVE WATER PRODUCTION
OR INTAKE (gpm)
8
S
-------
TABLE 11. COST SCHEDULE FOR DENSITY LOGS (dollars)
Depth
Operation
Company/Service
per foot minimum3 per foot minimum3
Total
minimum3
Birdweir
Density/borehole 0.27 540.00 0.23
compensated
Schlumbergerc
Formation density 0.29 580.00 0.27
Dresser Atlas
Compensated 0.29 580.00 0.27
densilog-caliper
We lex6
Compensated density log 0.27 540.00 0.23
345.00
540.00
540.00
885.00
1,120.00
1,120.00
460.00 1,000.00
Notes:
aAll service companies have a 2,000-foot minimum.
Birdwell Rocky Mountain Price Schedule, June 1980.
°Schlumberger Rocky Mountain Price Schedule, October 1979,
Dresser Atlas Rocky Mountain Price Schedule, July 1980.
eWelex Rocky Mountain Price Schedule, January 1980.
density logs can be used to define the hydrogeologic framework; however, these
data should be considered semiquantitative and used in conjunction with other
geophysical logs, i.e., caliper, fracture, lithologic, etc.
In constructing porosity analogs from density data, oil shale grade will
affect the porosity calculations. This is shown in Figure 8, where rich oil
shale zones correspond to high porosity values, and is a direct result of the
method used in constructing the analog. The matrix density, held constant in
computer routines used by logging companies to construct porosity analogs, can
vary 19 percent with a change in oil shale grade from 10 to 35 gal/ton. This
would produce a change in porosity of up to 25 percent if values of the other
parameters in Equation 2 are held constant, reflecting, in part, a real change
in the primary porosity.of the oil shale with pore spaces filled with less
dense organic material. This type of primary porosity would not serve as a
conduit for groundwater and therefore would not correlate with permeable zones
important to hydrogeologic studies. In an attempt to illuminate the effect of
oil shale grade on porosity calculations, another analog was developed for the
same test section in Well CE-705A and is shown in Figure 9. This porosity
analog was constructed with bulk density measurements taken from the same
50
-------
IS
-5
ft)
POROSITY (x 10-2)
& 8
g
CUMULATIVE WATER PRODUCTION
OR INTAKE Igpm)
S 8 8 S S 2 g S 8
&>
-5
-5
«j.
X
O.
ft>
o
O
O
I/)
O
o>
m
a
D
m
o.
z
S
3D
m
3
0)
-5
C
-5
I
o>
o
n-
o
i
CD
O
cn
_ l^l
. i
8
-------
Birdwell density/borehole compensated log used to construct Figure 8. In Fig-
ure 9, matrix densities were varied with depth based on Fischer analysis and
on the relationship of oil yield to specific gravities of Colorado oil shale
developed from a nearby test hole on Tract C-a. A problem in constructing the
analog developed in a few cases where bulk density measurements were found to
be higher than corresponding matrix densities based on the Fischer analysis.
In these cases, the numerator of Equation 2 became negative and negative po-
rosity values resulted. These values were set equal to zero in the computer
routine used to calculate the analog. An explanation for this phenomenon may
lie in errors, nonrepresentative Fischer analysis (2-foot varied lithologic
sections described by a single analysis), or calibration errors in the density
log. In addition, a discrepancy was noted in the density values for the vary-
ing grades of oil shale. Birdwell plots shows density varying from 2.49 to
1.66 gm/cc with a corresponding change in oil shale grade from 2 to 80 gal/
ton. A table developed by the Department of Energy shows density varying from
2.66 to 1.58 gm/cc for the same change in oil shale grade. This latter range
of densities was used to set matrix values for construction of Figure 10.
Comparison of Figures 6, 7, and 8 with Figure 9 show marked differences.
Figure 9 appears to provide a more realistic range of porosity values but
shows little correlation with water production from the spinner log. The gen-
erally high porosity values correlating with rich oil shale zones have been
eliminated, leaving isolated porosity peaks. Unfortunately, alternate logs
instructive in evaluating these peaks (caliper, fracture, and lithologic,
etc.) were not available for review; thus, the utility of Figure 9 could not
be fully determined. In theory, porosity analogs developed by varying the ma-
trix density to reflect the true lithologic conditions should provide a better
measure of porosity and should lead to correlation methods to equate permeable
and porous zones in the oil shale stratigraphy. Additional analogs should be
developed to evaluate this tool in defining the hydrogeologic framework.
Electric Logs
Principle of Operation-
Electric logs measure the electrical properties of the formation and
drilling fluids that penetrate the borehole wall. These properties include
electric potential and resistivity or, conversely, conductivity. The electric
log is primarily used for the construction and correlation of stratigraphic
and structural cross sections and in delineating permeable beds.
Multiple-track log presentations, including measurements of electric po-
tential and resistivity/conductivity, are commonly used. The dual-induction
laterolog discussed here consists of a correlation log, including spontaneous
potential, resistivity, and conductivity measurements on a log scale of 2
inches per 100 feet, and a detail log (5 inches per 100 feet) developed from
deep- and medium-reading induction devices and a shallow-investigation, fo-
cused resistivity tool. The detail log is recorded on a logarithmic grid
along with a standard spontaneous potential curve. Portions of the correla-
tion and detail logs from Tract C-a Well CE-705A are shown in Figures 10 and
11, respectively. The three types of electric logs (spontaneous potential,
induction, and focused current resistivity) are usually run simultaneously.
52
-------
1,000
in
O
r~
i
ui
o
fO
I
O
O
rd
i.
t-
M
O
c
O
(O
"cu
O
O)
en
u.
-------
MEDIUM INDUCTION LOG
DEEP INDUCTION LOG
8 §
§ 5
5
I
2
in
o
r~.
i
LU
o
n)
i
O
CO
i.
Dl
O
O
si
U
OJ
LO
1.000
1.050
1,100
1,150
1.190
DEPTHS
(O
O)
S-
3
01
-------
Spontaneous potentialThe naturally occurring electric potential of a
formation penetrated by a borehole is called the spontaneous potential, self-
potential, or simply SP. It is generally printed on the left track of the log
as shown in Figures 10 and 11. Two phenomena (electromechanical and electro-
kinetic) are thought to produce the potential current recorded in the SP log.
The amplitude of the current is the cumulative effect of these phenomena tak-
ing place between the drilling fluid and the formation. For an SP current to
be recorded, the well must be filled with a conductive fluid that can provide
electrical continuity between the SP electrode and the formation. Further-
more, if this conductive fluid and the formation water have essentially equal
resistivities the SP currents will be quite small and the log trace rather
featureless. The existence of the SP current is also dependent on a certain
minimum permeability that will allow ion migration between the drilling fluid
and the formation.
The electromotive forces (EMF) of electrochemical origins are believed to
be the largest contributor to the SP deflection. These are generated by dif-
ferences in solution concentration between the drilling fluid and the forma-
tion water. For example, if the salinity of the drilling fluid is lower than
that of the formation water, electric current flows into the formation oppo-
site to the permeable zones, producing a negative (left) deflection on the SP
log. Conversely, if the drilling fluid has a higher salinity than the forma-
tion water, a positive (right) deflection is recorded. Thus, the SP log is
theoretically useful in the detection of permeable beds and in defining the
location of their boundaries. This phenomenon may contribute to the shaded SP
response shown in Figures 10 and 11. These negative deflections appear to de-
fine permeable beds in the Lower Aquifer system. The spinner log indicates
significant fluid loss from the borehole that corresponds to the three upper
deflections. However, no change in vertical flow velocity is found opposite
the lowest stratigraphic SP deflection, nor do similar log anomalies up-hole
indicate a change in water production on the spinner log. This may be caused
by limitations inherent in the spinner log measurements or it may reflect rel-
atively low permeability of the beds. Whether the beds are permeable or not,
there is no direct relationship between the magnitude of the SP deflection and
the permeability of the formation, nor is there any direct relation to poros-
ity. Supplemental information from alternative geophysical logs is required
for an accurate interpretation in this case.
The electrokinetic portion of the SP log is generated when the drilling
fluid (an electrolyte) flows through a porous, nonmetallic medium (the mud
cake) into the formation. The EMF is primarily produced opposite permeable
formations where the pressure differential is maximum. Flow from the well
bore into the formation produces a negative (left) SP deflection, and flow
from a bed to the borehole produces a positive (right) deflection. The magni-
tude of the recorded potential is related to the velocity of the flow, resis-
tivity of the electrolyte in the mud cake or formation, as well as several
other factors. In general, the SP deflection generated by this electrofiltra-
tion is small and commonly considered negligible except for special situations
that are comparatively rare.
InductionThe induction sonde consists of several receiver and transmit-
ter coils.Constant intensity, high-frequency, electromagnetic waves are
55
-------
emitted from the transmitter coils, inducing secondary currents in the forma-
tion from the alternating magnetic fields set up by these waves. The eddy
currents flowing through the formation produce their own magnetic fields that
generate signals in the receiver coils. These induced signals are essen-
tially proportional to the conductivity of the formation or inversely propor-
tional to the resistivity. Variations of the transmitter/receiver coil
spacing in the sonde produce deep- and medium-reading tools.
Focused-current res istiv ityThe focused current sonde consists of a cen-
tral electrode symmetrically surrounded by additional pairs of interconnected
electrodes. The potential difference of the surrounding (guard) electrodes is
maintained at zero to focus the formation current into a thin sheet, which
flows horizontally into the borehole wall. Focused-current devices provide
better resolution than conventional resistivity tools in thin to moderately
thick, highly resistive beds. Focusing sondes are available for use in deep,
medium, and shallow depths of investigation.
The separation of deep and shallow resistivity measurements, whether in-
duction- or focused-current-derived, is an indication of invaded or permeable
zones. This separation occurs when the resistivity of the drilling fluid and
the water in the invaded bed are sufficiently different to alter the resistiv-
ity of that bed near the borehole. Thus, if the resistivity of the drilling
fluid is greater than that of the formation water, the shallow investigation
tool should read higher than the deep-reading device.
In Figure 10, separation of the short-normal and induction-resistivity
reading opposite the four negative SP deflections cannot be determined since
both are off the linear scale. The logarithmic grid used in Figure 11 shows
generally high resistivity values for the same section with no discernible
separation between the deep- and shallow-reading tools. This suggests that
invasion is so deep that it extends beyond the limits of the deep-reading
tool. Water chemistry data indicate that a salinity difference exists between
the borehole fluid and formation water, and the spinner survey shows a large
quantity of water intake for this zone.
An order-of-magnitude change in the resistivity occurs below a depth of
about 1,150 feet. Fischer analysis and lithologic records show no significant
change in the oil shale stratigraphy at this depth, suggesting that the reduc-
tion in resistivity (increase in conductivity) can be related to more saline
formation water below the Lower Aquifer system.
Cost Data--
Several combinations of resistivity log presentations are available from
logging service companies. Two representative resistivity logs have been se-
lected from each service company. The cost schedules for running these logs
are given in Table 12.
Evaluation--
In general, shallow, fresh-water aquifer test holes in oil shale strati-
graphy provide a poor working environment for electric logging devices. The
56
-------
TABLE 12. COST SCHEDULE FOR VARIOUS RESISTIVITY LOGS (dollars)
Depth
Company/Service
Birdwellb
Induction electric
FS guard log
Schl umber gerc
Induction electrical
Dual -inducti on laterolog
Dresser Atlas
Induction electric
Dual-induction focused
We lex6
Induction electric log
Dual-induction guard log
per foot
0.25
0.25
0.29
0.30
0.29
0.30
0.26
0.27
minimum3
500.00
500.00
580.00
600.00
580.00
600.00
520.00
540.00
Operation
per foot
0.24
0.24
0.25
0.25
0.25
0.25
0.24
0.23
minimum3
360.00
360.00
500.00
500.00
500.00
500.00
480.00
460.00
Total
minimum3
860.00
860.00
1,080.00
1,100.00
1,080.00
1,100.00
1,000.00
1,000.00
Notes:
aAll service companies have a 2,000-foot minimum.
Birdwell Rocky Mountain Price Schedule, June 1980.
°Schlumberger Rocky Mountain Price Schedule, October 1979.
Dresser Atlas Rocky Mountain Price Schedule, July 1980.
eWelex Rocky Mountain Price Schedule, January 1980.
highly resistive (clear water) drilling fluids and oil shale rock mask or dis-
tort the normal SP and induction/focused current resistivity tool response,
complicating the quantitative interpretation of these data. Specific condi-
tions or observed log responses that hinder the utility of these data are
given below:
t Spontaneous Potential
SP curves undergo gradual transition at bed boundaries in
highly resistive oil shale environments. Therefore, permeable
beds cannot be accurately located using the SP curve.
The highly resistive drilling fluids provide poor electrical
continuity between the SP electrode and the formation.
57
-------
Borehole fluids are used during drilling operations; there-
fore, the resistivity differences between formation waters and
drilling fluids is small, reducing the character of the SP
deflections.
Drilling muds are not always used during exploration studies;
therefore, the pressure differential caused by the mud cake
between permeable beds and the borehole may not develop, re-
ducing the electrokinetic component of the SP curve.
Fluid motion common in Tract C-a wells tends to mask the true
SP response.
Induction/Focused Current Logs
Resistivity logs have not produced a clear separation between
deep and shallow investigation tools for known permeable beds
in the oil shale environment.
-- Conductivity measurements have a high degree of uncertainty
in the nonconductive oil shale rock due primarily to instru-
ment sensitivity at the low end of the scale.
Invasion of permeable zones will be extremely deep due to min-
imal mud cake development, masking true formation water resis-
tivity determinations used in several alternative quantitative
log interpretation techniques (not discussed here).
Bearing in mind the difficulties of using resistivity logging devices in
oil shale environments, they are still useful in construction and the correla-
tion of cross sections. However, information derived from log interpretation
should be considered qualitative in nature.
Seisviewer Log
Principle of Operation
The seisviewer sonde consists of a transmitter and receiver transducer
mounted on a vertical axis that is rotated at a uniform rate during logging.
In this configuration, the transmitting transducer emits a narrow-band acous-
tic signal to the entire inside diameter of the borehole as the instrument is
lowered into the hole. The acoustic energy is transmitted through the drill-
ing fluid, strikes the fluid formation boundary, and is reflected back to the
receiving transducer. The amount of energy returned to the sonde is a func-
tion of the scatter caused by the physical properties of the borehole wall and
attenuation in the borehole fluid. This signal is subsequently sent to the
recording oscilloscopes at the surface via the wireline.
A flux-gate magnetometer is mounted on the vertical axis with the trans-
ducers and senses the earth's magnetic field. The sonde is then oriented by
magnetic direction, as shown at the top right of Figure 12.
58
-------
1,110
1,120
1,130
a.
LU
a
1,140
1,150
Figure 12. A portion of a seisviewer log for Tract C-b Well 32x-12.
59
-------
The diameter of the sonde is 3-3/8 inches and can be operated in a 4- to
12-inch borehole. It is run in fluid-filled (water, mud, or oil base), cased
or uncased holes.
The log presentation is an acoustic picture of the fluid formation bound-
ary as if the borehole had been vertically dissected and layed out flat. The
magnetic orientation of the log is given at the top of each log trace. Figure
12 shows a 40-foot section of a seisviewer log of Well 32x-12. Some of the
features of the borehole wall depicted on the log are the sinusoidal curves at
1,122 and 1,125 feet. These are low angle fractures dipping to the north and
south, respectively. Dark patches on the log are areas of weak signal return
and represent vugs or beds that have eroded during drilling or completion op-
erations. The sections from 1,115 to 1,120 feet and 1,145 to 1,150 feet are
in-gage, competent rock with a strong signal return. Water production test
zones 4 and 5 for Well 32x-12 are shown to the left of the log.
Cost Data
Birdwell Division is the only logging company of those reviewed that has
the facilities to operate a seisviewer. However, this logging tool has been
pulled out of general use and can only be obtained by special request to the
Birdwell office in Tulsa, Oklahoma. Although cost quotes were not developed
for the seisviwer, it will be expensive to obtain this type of borehole imag-
ery. A logging program involving several shallow oil shale exploration holes
would be required to bring the unit cost per log into a comparable price range
with other geophysical logging methods.
Evaluation
Seisviewer logs are used to define vugs, fractures, breccia zones, wash-
outs, and bedding planes in open holes. In the shallow, clear water, oil
shale exploration holes, resolution of these physical features is excellent
when the hole is close to the gage. These field data are extremely useful
when production testing directly follows the logging operation. In this case,
the seisviewer log can help guide the selection of hydrologic zones to be
tested and aid in the placement of the testing equipment. In Figure 12, for
example, water production for Bed 5 (between 1,135 and 1,145 feet) was re-
corded at 180 gpm, which was the third most productive zone tested in Well
32x-12. From the imagery in Figure 12 it appears that the upper test limit of
Bed 5 does not conform to the upper part of the production zone. Based on the
seisviewer log, the upper packer for a production test for Bed 5 should have
been set in the more competent rock at a depth of 1,125 feet. This would have
included the potentially permeable vugs and eroded bedding planes shown in the
log, providing a more representative picture of the production to be expected
from a mine shaft penetrating this horizon.
As indicated earlier, the seisviewer is no longer in widespread use. The
electronic equipment for this labor-intensive logging method is costly to run
and maintain. Viscous drilling fluids (mud or oil) and oblong or over-gage
borehole diameters attenuate the signal, thus reducing resolution. The log-
ging speed for high resolution is slow, about 5 ft/min, creating excessively
long logging runs for deep wells. A combination of these factors has limited
60
-------
the usefulness of this logging method in the oil industry, thus reducing its
overall utility and marketability. This is probably the primary reason why
Birdwell has elected to provide the service on a limited basis.
HYDRAULIC TEST METHODS
Geophysical methods of determining hydrogeologic parameters in oil shale
stratigraphy rely on direct or indirect measurements of the borehole wall, the
surrounding rock, and the formation fluid to deduce the hydraulic parameters
important to the development of an oil shale project. These methods, however,
do not include a significant class of direct hydraulic testing procedures that
provide detailed hydrology data through an evaluation of the response of the
test well to the injection and removal of fluids. Hydraulic test methods are
discussed in this subsection.
Well pump and injection tests range from simple, rather informal proce-
dures conducted during a period of a few hours, to sophisticated hydraulic
tests conducted over a continuous operational period of several weeks and in-
volving numerous observation wells. To simplify the profusion of methods,
testing procedures have been grouped into four general classes as follows:
1. Drill stem tests
2. Single packer tests
3. Dual packer tests
4. Long-term pump tests.
The groups are not intended to be inclusive, yet they provide a sufficiently
large range of testing methods to meet the needs of most oil shale development
projects. Each group is divided into three components including (1) test pro-
cedures, equipment, and costs, (2) analytical methods used to interpret test
data, and (3) remarks. Actual test data from the Federal oil shale tracts are
utilized wherever possible.
Review of the testing procedures, equipment, costs, and utility of the
resulting data has led to the following priority ranking of the four general
classes of tests:
Dual packer tests provide horizon-specific hydrologic data at a
minimal cost when multiple tests are conducted in a single bore-
hole. Down-hole test equipment assembly allows for pumping, in-
jection tests, and discrete water quality sampling.
Long-term pump tests of aquifer systems produce the most repre-
sentative regional data on boundary conditions and flow patterns.
However, these tests are expensive and should be conducted by
personnel knowledgeable in hydrologic principles.
61
-------
Single packer tests generate good-quality, bed-specific hydro-
logic information at about three to four times the cost of simi-
lar data gathered by dual packer tests (assuming a single-hole,
multiple-test application). Field operation and procedures are
simplified over the dual packer assembly.
t Drill stem test data is of limited value due to its nonspecific
nature, high cost relative to the data return, and difficult in-
terpretation. Drill stem tests are now seldom used in the devel-
opment of oil shale tracts.
Drill Stem Tests
Test Procedures, Equipment, and Costs
In conjunction with early exploration efforts, drill stem tests were rou-
tinely performed in core holes to define the hydrology of the Federal oil
shale tracts. This kind of test includes the "informal pump tests" conducted
on Tract C-a and the "jetting tests" performed on Tract C-b. Similar proce-
dures were followed for both testing methods with some minor differences in
the equipment utilized. Drill stem tests are performed in an open hole as
follows:
t The borehole is drilled to the desired depth
An air line is lowered down the drill stem to a point near the
bottom of the string
Air is blown or jetted through the air line, lifting the fluid in
the drill string to the surface
Discharge is measured as changes in water level through a Par-
shall flume or similar device and converted to a flow rate in
gallons per minute
Airlift pumping at a constant rate is maintained for a predeter-
mined length of time (2 hours for Tracts C-a and C-b)
Immediately following shutdown of the air compressor, a water-
level measuring device is lowered down the drill string (depth
sounder, or water-pressure recorder)
Recovery of the water level following shutdown is recorded
If an observation tube is installed in the well, both drawdown
and recovery-water-level measurements can be compiled.
On Tract C-a, drill stem tests were conducted after penetration of the
B-groove and at the bottom of the borehole. This testing program was intended
to provide hydrologic data on the Upper Aquifer system and on a combination of
the Upper and Lower Aquifer systems. The majority of these tests are de-
scribed in Rio Blanco Oil Shale Report (1974). On Tract C-b, from three to
62
-------
six drill stem tests were conducted on site wells and core holes. Test zones
included top of Parachute Creek, top of mining zone, base of mining zone, and
total depth of the hole. Raw data for these tests are given in C-b Shale Oil
Venture, 1974.
Equipment required to perform a drill stem test, in addition to that com-
monly available on drill rigs, include an air compressor of sufficient capac-
ity to overcome the pressure developed from the column of water within the
drill string and a water-level measuring device. A geologist or hydrologist
should be present to supervise the test.
Costs for each test are based on the total rig time, equipment cost or
rental, labor for supervision, and the number of tests conducted. Individual
tests should run from $600 to $750 for the short (4- to 5-hour) tests. Longer
tests are more expensive, depending on the amount of additional labor and rig
time involved.
Analytical Techniques
The time-recovery data compiled during the drill stem tests are used to
calculate transmissivity (T) and specific capacity. T is the rate at which
water will flow through a unit width of aquifer fully penetrating the satu-
rated thickness under a unit hydraulic gradient. T has dimensions of length
squared per unit of time because it represents flow through a vertical strip
of unit width. Specific capacity is yield per unit drawdown expressed in gal-
lons per minute per foot or gallons per day per foot.
Analytical methods for determining these parameters are derived from
Theis' nonequilibrium formula (Theis, 1935). A straight-line, or graphical,
solution for a modified Theis equation was discussed by Cooper and Jacob
(1946) and has been used by both tract developers to calculate T. A concise
description of this graphical solution is presented in Miller (1973). The
general method is as follows:
Time-recovery data are plotted on semi log paper, recovery (in
feet) on the arithmetic scale and time (in minutes) on the log
scale
The slope (AS) is determined by the change in water level
(recovery) through one log cycle of time
t Transmissivty (T) is then calculated from the following formula:
T - (264) (Q)
'
where: Q is the constant recovery (drawdown) discharge (gpm)
As is the slope (feet)
T is the transmissivity (gpd/ft).
63
-------
The success of the straight-line solution is based on the assumption that
the recovery time is long and the radius of the observation point to the pump-
ing (recovering) well is small such that the straight-line approximation coin-
cides with the Theis-type curve. This constraint is met within the first few
minutes of recovery (pumping) when measurements of the water level are taken
in the pumping well.
Relatively few, if any, of the aquifers in fractured oil shale strati-
graphy will conform to the basic hydrologic assumption of infinite extent in
all directions from the pumping well used by Theis to develop the flow equa-
tions. Geologic and hydrologic boundaries affect the slope of the time-recov-
ery (drawdown) plot. Impervious boundaries limit the flow of water to the
pumping well, causing a more rapid deepening on the cone of depression and
steepening the slope of the time-drawdown curve. Conversely, impervious
boundaries increase the rate of recovery and steepen the slope of the time-
recovery curve when calculated recovery (drawdown extended through the recov-
ery period minus residual drawdown) is plotted against time. Recharge
boundaries have the reverse effect on the slope in the straight-line solution.
Recharge water entering the well flattens the slope of the curve. Qualitative
evaluation of boundary conditions from the graphical solution are useful in
defining the hydrogeologic framework of the study area and in planning more
detailed hydraulic testing programs. A more detailed discussion of boundary
conditions on well hydraulics is given in Chapter 6 of Johnson (1975).
More sophisticated approaches are available to define T from confined-
aquifer, unsteady-state drawdown/recovery data. These include Theis1
straight-line recovery method (Theis, 1935), Theis1 curve-fitting method
(Jacob, 1940), and Chow's nomogram method (Chow, 1952). However, the addi-
tional time required to interpret the data from these methods is difficult to
justify in that the data sets are from thick, complex aquifer sequences that
are not adequately represented by the simplified models used to develop the
interpretational theory.
Remarks
Review of the drill stem test data submitted to the Area Oil Shale Super-
visor indicated that the "informal pump tests" provided ranges for T based on
the straight-line solution to the time-recovery data. Noting a change in
slope of the plot and the implicated boundary condition, T values were calcu-
lated using Jacob's method with Ac values derived from the primary and secon-
dary slopes of the graph (Figure 13). This is not consistent with standard
methods derived from the Theis nonequilibrium formula. The following is
stated in Johnson (1975), p. 118, regarding such an interpretation:
It should be pointed out in passing that calculation of the
transmissibility, T, of the water-bearing formation must be made
from the value of As corresponding to the slope of the first part
of the time-drawdown (recovery) graph. Beyond the point where a
change in slope occurs, a numerical value that may represent the
slope of the second part of the graph is of no significance in an-
alyzing the pumping (recovery) test data. No attempt should be
64
-------
made to use any such-value in either the Theis non-equilibrium or
modified non-equilibrium formulas.
Therefore, T values calculated in this manner have no theoretical basis and
can be extremely misleading to tract developers.
350
STATIC WATER LEVEL
INFORMAL RECOVERY TEST
264(240)/200
320 gpd/ft
264(2401/38
1,670 gpd/ft
240/210
1.1gpm/n
SOURCE: RIO BLANCO OIL SHALE PROJECT. HOLE C-7 PRESSURE
BOMB TEST: DEPTH 1.2001wt (WRIGHT WATER ENGINEERS:
AUGUST 19741
600
100
1,000
10,000
TIME SINCE PUMP OFF (minutes)
Figure 13. Jacob's straight-line solution for T.
Well completion reports (drill stem tests) for eight core holes on Tract
C-a show that boundary conditions usually affected the time-recovery plots
within the first 20 to 30 minutes of recovery. Without exception, impermeable
boundary conditions were indicated by these time-recovery curves. This is to
be anticipated in an aquifer where permeability is fracture-controlled because
of the low permeability of the unfractured matrix rocks. The tests should
have been conducted for a long enough period of time to observe if recharge
water had broken into the well in response to the head difference in the frac-
ture system created by pumping; thus the true nature of the boundary could
have been determined.
Raw data for 55 drill stem tests are given in Table 11 B-4, C-b Shale Oil
Venture (1974). These data have not been plotted to check the analytical pro-
cedures used to calculate T values.
A serious disadvantage of the drill stem test, and rendering less value
to the calculated parameters, is that T is obtained for the entire open por-
tion of the borehole and no zone-specific information is obtained. In
65
-------
addition, when combined (Upper and Lower Aquifer systems with differing pres-
sure heads that create production and "thief" zones) drill stem tests are con-
ducted, it is unlikely that the straight-line solution will adequately model
the well conditions from which T values are to be calculated. For these and
other reasons, drill stem tests on both Tracts C-a and C-b were discontinued
early in the exploration/data-gathering phase of development.
Single Packer Tests
Test Procedures, Equipment, and Costs--
Testing methods included here are single packer drawdown/recovery and in-
jection-pressure permeability tests. Test procedures for the former are simi-
lar to those discussed for drill stem tests except that a packer is lowered on
drill pipe to a point above the bottom of the hole (approximately 50 feet on
Tract C-b), water is lifted or jetted from the packed-off section, and water-
level measurements are compiled. The packer is then removed, the hole deep-
ened to the next zone of interest, and the test repeated.
Equipment for the packer test includes an air compressor, a string of
drill pipe, and a packer. Inflatable packers, as opposed to compression or
leather cups, are recommended because they seal better on rough walls or in
irregular shaped holes, reduce testing time, and are therefore more
economical.
Costs for running a single packer drawdown/recovery test requires rig
time to set the packer in addition to labor and equipment for a standard drill
stem test. The cost (in 1980 dollars) is estimated to be $1,800 to $2,000 per
test.
The injection-permeability test is run by drilling the borehole to the
desired depth, pulling the drill string, and seating the packer at the desired
depth above the bottom of the hole. The section is flushed out to remove
drilling fluid and water is pumped under pressure into the test zone. The
constant pump discharge (Q) and applied pressure (H2) are recorded. Follow-
ing completion of the test, the hole is deepened to the next test horizon and
the procedure repeated.
Pressure-permeability tests on Tract C-b were run in conjunction with
drawdown/recovery tests. The procedure varies slightly from the injection
test in that after the packer is set, a valve is opened to allow formation
fluid to flow into the drill pipe, thus reducing the hydrostatic pressure in
the test section. The valve is then closed and data on the pressure recovery
are recorded. A pump test is performed following recovery of the hydrostatic
pressure. The injection pressure test is then conducted by pumping water at a
constant rate into the test section and observing the pressure change in the
drill pipe. Commonly, several different injection rates are used during the
test.
Single packer injection-permeability tests require substantially more
equipment than pump tests, including a centrifugal test pump, a water meter
to measure injection flow rates, connection pipes, a swivel plug valve, a
66
-------
pressure gage and sub for the gage, etc. Further details on equipment re-
quirements and arrangement for testing are given in Bureau of Reclamation
(1977). In addition to the above equipment, a clear source of water is re-
quired for testing. This can be discharge from local wells or springs but
should be of equal or better quality (lower IDS) than the formation fluid in
the test zone. In arid areas this water may have to be trucked to the test
site and can become a substantial cost item.
The injection pump is the primary piece of test equipment. Tests are
usually run using the rig's mud pump. These multiple-cylinder-type pumps usu-
ally have a maximum capacity of from 25 to 30 gpm and provide acceptable test
results only when low permeabilities or short test sections allow development
of back pressure on the formation. In addition, since the fluctuating pres-
sure through this type of pump is difficult to read accurately, it is recom-
mended that a suitable centrifugal pump be obtained for testing.
Tests should be run for 20 minutes or longer with readings of injection
rates (gpm) and applied pressure (psi) taken at 5-minute intervals. Pressure
can be increased during the test to determine rock characteristics but, to
prevent blowouts or fracturing the borehole wall, it should not be taken too
high. As a general rule-of-thumb, safe pressure in consolidated rock is 0.5
psi per foot of depth from the ground surface to the upper packer.
Costs for the injection test vary with the availability of a suitable
injection fluid and the cost of obtaining the surface equipment. It is esti-
mated (in 1980 dollars) that $2,200 to $2,600 per test would cover the equip-
ment, operation, and labor costs incurred by a single packer injection
permeability test.
Analytical Techniques
T values can be calculated from a single packer drawdown/recovery test
using methods discussed under "Drill Stem Tests" in this section. Injection-
permeability tests are discussed in Ahrens and Barlow (1951). Figure 14 is a
reproduction from this report that shows the setup for the single packer per-
meability test. Parameters measured during testing are as follows:
1. Elevation of the ground surface at the test site (feet)
2. Radius of the hole, R (feet)
3. Length of the test section (the distance between the packer and
the bottom of the hole), A (feet)
4. Depth from ground surface to bottom of the hole (feet)
5. Distance of swivel above ground surface (feet)
6. Applied pressure of head, \\2 (psi or feet)
7. Steady flow into well at 5-minute intervals, Q (gpm)
67
-------
SWIVELC
GROUND SURFACE
ZONEI
BASE OF ZONE 1
ZONE II
((Ct+4)r(Tll+H.A)|
WATER TABLE
ZONE III
(Cs+4) r H
TOP OF IMPERMEABLE ZONE
LIMITATIONS: Q/i < 0.10. S > 5A. A > lOr
* Coefficient of penneeoility (ft/iecl under unit o/edient
- StMdy flow into will (eft)
- EffMtrw hMd - h, + h2 - L (ft)
- In ttttibonwMT note. dtairafaMwMnnmrtMri bottom of hot* in tnt»
ram* MtwMn mini md wmr tabte (ft)
Cu
K
Q
H
"1
hj - ApplM pranm « colter (ft); 1 pri - 2.31 fMt
L - H«d Ion in nip. d« to trteoon: foe qu»ntiti« C«« *«n 4 9pm in 1%" p.p«,
it nwy M ignorad (ft)
X - Pwemt of unmunMd itnta (X - H/Tu)
A Imgth of twt wetion (ft)
r - Ridiui of t«« note (ft)
Niducttvitv coefficient, unsettmtso bed
ductivity
fflcitnt, Mturattd bid
C>
U - Thfeknm of umnunud mmrW (ft)
S - ThickMiiofiniirandm«wial (ft)
Tu - U - 0 * H
O - Dimnc* from ground Hirfra to bottom of note (ft)
a - Sorfiot am of tnt wction (ft); in Miihod I tnt of «nM ptut MM of bottom;
Figure 14. Single packer injection test setup
(after Ahrens and Barlow, 1951).
68
-------
8. Nominal size of pipe (inches) and length of pipe (feet) between
swivel and packer
9. Thickness of saturated material above a relatively impermeable
bed, S (feet).
In addition to the these measurements, head loss in the drill pipe due to
friction (L), saturated bed conductivity coefficients (Cs), and a definition
of boundary conditions between Zones 1 and 2 are required to interpret test
results. Graphs required to determine these parameters and numerical examples
are provided in Ahrens and Barlow (1951).
Multiple pressure injection tests are performed in the same manner as de-
scribed above except that the pressure is applied in more than one essentially
equal steps. The applied pressure can be estimated by determining the maximum
safe pressure and dividing by the number of pressure steps desired.
Synthetic test results of multiple pressure tests for varying formation
conditions have been postulated in Bureau of Reclamation (1977). These are
given in Figure 15. Circled numbers on Figure 15 denote the following proba-
ble conditions:
1. Probably very narrow, clean fractures; laminar flow; low perme-
ability with discharge directly proportional to head
2. Firm, practically impermeable material; tight fractures; little
or no intake regardless of pressure
3. Highly permeable, relatively large open fractures indicated by
high rates of water intake and no back pressure (pressure shown
on gage due entirely to pipe resistance)
4. High permeability with open and permeable fractures containing
filling material that tends to collect in traps and retard
flow; turbulent flow
5. High permeability; contains fracture filling material that
washes out and increases permeability with time; fractures
probably are relatively large; turbulent flow
6. Similar to (4) but tighter fractures and laminar flow
7. Packer failed or fractures are large and have been washed
clean -- highly permeable; turbulent flow (test takes capacity
of the pump with little or no back pressure)
8. Fairly wide and open fractures filled with clay gouge material
that tends to pack and seal under water pressure (takes full
pressure with no water intake near end of test)
9. Open fractures with filling that tends to block and then break
under increased pressure; probably permeable; turbulent flow.
69
-------
I
o>
*
Z
100
PRACTICALLY
IMPERMEABLE
NO INTAKE
VERY PERMEABLE; TAKES
CAPACITY OF PUMP; NO
BACKPRESSURE
PACKER BROKE LOOSE;
TOOK CAPACITY
0010 50 1001
PLUGGED TIGHT WITH NO
MEASURABLE INTAKE AT
MAXIMUM PRESSURE
EFFECTIVE DIFFERENTIAL PRESSURE (psi)
Figure 15. Plots of simulated, multiple pressure,
permeability tests (after Bureau of
Reclamation, 1977).
Tract C-b developers used a technique presented by Homer (1951) to ana-
lyze the pressure-recovery data from the single packer tests. This method is
essentially the same as Jacob's straight-line solution except that pressure in
psi is plotted against time on semilog paper instead of water levels in feet.
A drawdown analysis presented by Odeh and Jones (1965) on Tract C-b was
used to analyze the multiple-pressure, single packer injection tests. Al-
though developed primarily for formation evaluation from oil and gas wells
flowing at variable rates, this technique has had wider application. Analysis
of field data is conducted as follows (for greater detail, see Odeh and Jones,
1965):
Production in barrels per day is plotted on regular graph paper
versus time in appropriate units (minutes)
Average flow rates for specific time increments are calculated
70
-------
The change in pressure, Ap, (original formation pressure minus
flowing bottom-hole pressure) is determined and divided by the
average flow rate (qn) for each increment, Ap/qn
The sunmation of the different flow rates divided by the last
flow rate is calculated as a function of time from the following
expression and plotted against Ap/qn
n-1
~ X A Qi logCVti)
Hn 1=0
where: qn is the last flow interval (bpd)
q-j is the itn flow interval (bpd)
tn is the total flow time (minutes)
t^ is the flow time for each change in rate (minutes)
The slope (m) of the resulting straight-line plot is determined
T is calculated from the formula T = 7.06 u/m (where u is the
viscosity of the fluid in centipoise).
T values and permeability for single packer tests in Well SG-17 were cal-
culated as described above. Computer plots from the analysis are given in the
C-b Shale Oil Venture (1979).
Remarks
Single packer tests have performed well in the oil shale stratigraphy on
the Federal tracts. Analytical methods for data interpertation are readily
available.
Detailed information was compiled for Tract C-b, borehole SG-17, where 40
single packer tests were performed. These data provided a composite picture
of horizontal transmissivity through the lithologic section penetrated by the
well. These data were the primary input parameters for a computer model spe-
cifically designed for the Tract C-b mining and reinjection program. As such,
the accuracy of these parameters is extremely important to the oil shale proj-
ect. These computer-derived permeabilities are not consistent with values for
the same test sections presented to the area oil shale office in February of
1975 (C-b Shale Oil Venture, 1975). In addition, test results would be more
easily evaluated if they were presented in generally accepted water supply
units (gpd/ft2) rather than Darcy units adopted in petroleum engineering.
The primary drawback in using the single packer test method is that it is
very costly. Setting up the pump for injection and the "round trip" for the
rig to set and remove the packer is time-intensive. Because the tests are run
71
-------
prior to completing the well or core hole, geophysical logs useful in direct-
ing the hydrologic program by defining test beds cannot be utilized. These
drawbacks are in part overcome through hydraulic testing using the dual packer
method described below.
Dual Packer Tests
Procedures, Equipment, and Costs
Dual packer tests have been run on Tract C-b and are referred to as
"mini-pump tests" in the C-b Shale Oil Venture (1979).
In general, the test procedure is to drill the borehole to its final
depth. The drill string is then removed and geophysical logs can be run in
the open hole at this point if they are part of the overall testing program.
The dual packer assembly is lowered to the bottom of the borehole and testing
proceeds upward through the zones of interest. The packer assembly is set
straddling the test zone and the desired test(s) are run. The packers are
then deflated and moved up the hole to the next test horizon.
The equipment utilized in dual packer testing includes the packers, a
submersible pump, a multipurpose valve, and pressure transducers. The strad-
dle packers should be gas-inflatable so they can be deflated and reinflated
without requiring a return to the surface for redressing. This allows testing
of all zones during one trip into and out of the hole. A submersible pump
should be installed between the packers so that water samples and pump test
data can be collected. The multiple-purpose valve installed between the pack-
ers and above the pump provides access to the packed-off zone for fluid injec-
tion and can be sealed off during pump testing. Pressure transducers
installed above, below, and in the packed-off zone are used to measure pres-
sure changes and detect packer failure. Surface equipment is be similar to
that described for the single packer test.
In 1978, the U.S. Geological Survey (USGS) developed a custom packer as-
sembly for hydrologic testing and hydrofracturing by modifying a production
injection packer manufactured by Lynes, Inc., of Houston, Texas. This equip-
ment was tested in the Piceance Basin. Study results are documented in U.S.
Geological Survey (1978).
The USGS tests show that the dual packer assembly requires from one-
quarter to one-third less time than a standard single packer assembly for the
same hydrologic test because several tests can be performed on one round trip.
Costs are cut in nearly direct proportion to the time saved, resulting in
costs of about $500 for a 4- to 5-hour pump test and about $650 for an injec-
tion test (if water is trucked to the test site).
Analytical Techniques-
Dual packer tests on Tract C-b were conducted in 1975 in twin holes SG-1
and SG-1A. Equivalent test zones with rich oil shale beds were isolated in
each well with straddle packers and pump and injection tests performed. Semi-
confined, unsteady-state conditions described by Hantush and Jacob (1955) were
72
-------
used to model the aquifer. Solutions for the unsteady-state flow have been
described by Walton (1962) and Hantush (1956). These analytical methods are
discussed below.
Walton's method is a curve-fitting procedure from which transmissivity,
storage coefficient, hydraulic resistance of a semipervious layer, and leakage
factor of the water-bearing stratum can be determined. The reasoning used to
develop the solution is similar to Theis' method except there are several type
curves instead of one. This family of curves can be drawn from data published
by Hantush (1956) or found in Walton (1962).
The analytical procedure of Walton is as follows:
A family of type curves is developed on double-logarithmic paper
Drawdown versus time is plotted on double-log paper of the same
scale as that used for the family of curves
Observed data is superimposed over the family of type curves and
the best fit is found keeping the x- and y-axes parallel
A match point on the superimposed observed data sheet is selected
and the four corresponding parameters are read
These values are substituted into the appropriate equations and
the hydrologic parameters of interest calculated.
Hantush 's Method I (Hantush, 1956) solution uses the inflection point of
the time-drawdown data plotted on semilogarithmic paper. To determine the in-
flection point, the steady-state drawdown (maximum drawdown) is required and
should be known through direct observation or by extrapolation. This method
uses data from a single observation piezometer. The solution is developed as
f o 1 1 ows :
A plot on semilogarithmic paper of drawdown versus time (time on
the logarithmic scale) is prepared and the best fit curve is
drawn through the plotted points
t Determine the value of the maximum drawdown by extrapolating the
plotted points through time
Calculate the inflection point (Sp) on the curve using the for-
mula (see Hantush, 1956),
Sp = -- Ko(r/L)
4irkD
where Q is the discharge
k is the hydraulic conductivity
D is the saturated thickness
73
-------
r is the distance from the pumping well to the observation
well
L is the leakage factor of the water-bearing layer
Ko is the Bessel function
Read the value of time (tp) that corresponds to Sp
t Determine the slope of the best fit curve at the inflection point
(ASp) by the change in slope over one log cycle that includes the
inflection point, or by the tangent to the curve at the inflec-
tion point.
Substitute the values at Sp and ASp in the formula,
^OJP. . er/L Ko(r/L) ,
ASp
and determine the value of r/L by extrapolation from tables in
Hantush (1956)
t Transmissivity (kD) is then calculated using the equation,
L er/L.
4irkD
and a table of values for e~x (Hantush, 1956)
The storage coefficient (S) can then be calculated using the fol-
lowing equation:
s , 4kD(tp)
2rL
Hydraulic resistance (c) of the semipervious layer is then found
from the relation, c = L2/kD.
Injection permeability tests can be analyzed using the method of Odeh and
Jones (1965) described earlier. An alternative injection test is presented in
Ahrens and Barlow (1951) for steady flow conditions. Figure 16 is a diagram
of the test setup and equations used to calculate the permeability coefficient
(K). Measurements taken during testing are the same as those for a single
packer test (see page 67) with the following exceptions:
3. Length of test section, A, is the distance between the packers
(feet)
4. Depth, D, is measured from the ground surface to the uppermost
part of the lower packer.
74
-------
SWIVELt
GROUND SURFACE
CurH
BASE OF ZONE I
ZONE!
T
A
2Q
(Ctr)(Tu-(-H-A)
ZONE II
WATER TABLE
2r
K-Q/CirH
TOP OF IMPERMEABLE ZONE
ZONE III
LIMITATIONS: O/l < 0.10. S > SA, A > 10 r; in Mithod II, thidcnm of Mdl patter mutt to > 10 r.
K - Cotfflotnt of pwmMbility (ft/Mel undar unit gradiant
Q - StMdy flow into mil (eft)
H - EffactiM hMd - h1 + h2 - L (ft)
h, - In tut abo»a watar tabla. dlit»nc« tomaan tim* tat bottom of hok In tm
totow wit»f ahte I ft); dimnM hmmin innivil md «mr tibt» Iftl
hz - AppHad pranura at collar (ft); 1 pai
I Haad Ion in pipa dua to friction; for quantitio Ian than 4 gpm in 1V pipa,
it may to ignored (ft)
X - Pareantof unMuratad itrata (X - H/Tu)
A - LanothoftaitMctionlft)
r - RadHnoftaathola(ft)
Cu - Conductinty ooaffldant. umaturatad tod
Cl - Conductivity coafficiant. uturatad tod
U - TMcknan of unanuratad matarial (ft)
S - TmcknanofianintadmnariaKft)
Tu * U - O + H
D - DManca from ground iiirfaea to bottom of hola (ft)
a - Surfaca ana of tan Mcbon (ft); in Mathod I ana of wall pkn ana of bottom:
mMathodllanaofwaH
Figure 16. Dual packer steady flow injection test
(after Bureau of Reclamation, 1951).
75
-------
Remarks
Dual packer tests were conducted in only two holes, SG-1 and SG-1A on
Tract C-b. In each of these holes a single, interconnected horizon was iso-
lated and tests run without moving the packers. This testing method did not
utilize the primary economic advantage of the dual packer assembly, namely,
the ability to run several tests from one round trip in the borehole.
Analysis of the pump test data from the same section using Walton's
method shows large variations in T values. This variation could be caused by
inaccuracies in the water level, pressure measures (pressure measurements are
only accurate to ±1/4 foot), or significant leakage through the semipervious
layer during testing, which makes a unique fit to the family of curves diffi-
cult. T values calculated by the Walton and Hantush methods show relatively
close agreement but are low in relation to other test results for the same
bed. The accuracy of Hantush's method depends on precision water-level mea-
surements and the estimation of the steady-state (maximum) drawdown. Fortu-
nately, an independent check of T, S, and L can be made by substituting these
parameters into equations presented by Hantush and Jacob (1955) and calculat-
ing drawdown and time values that should fall within the observed data points.
The equations utilized in this check are as follows:
s = 9-W(u,r/L)
4irkD
and
where s = drawdown in the observation piezometer a distance r from the
pumping well
kD = aquifer transmissivity
S = coefficient of storage
t = time since pumping started
and w(u,r/L) is the "well function" for a specific piezometer with distance r
from sampling well and leakage factor L.
Long-Term Pump Tests
Procedures, Equipment, and Costs
Long-term pump tests have been conducted on both Tracts C-a and C-b.
Procedures for performing this type of test are given in numerous hydrology
texts. Chapter 10, Bureau of Reclamation 1977 Ground Water Manual provides an
in-depth discussion of acceptable methods, instrumentation, and required
equipment for pump testing.
76
-------
Cost items are similar to those for a dual packer pump test (with or
without the packers) and include labor, operation, and equipment. Total costs
can range from $3,000 up to $10,000 for a more sophisticated long-term test
with multiple observation wells.
Analytical Techniques--
Long-term pump tests provide the most representative information on aqui-
fer characteristics and boundary conditions. Analytical methods used by tract
developers are similar to those discussed earlier and include curve fitting,
calculation, and straight-line solutions. These methods have been developed
for isotropic aquifers and therefore provide average values of the hydraulic
parameters in anisotropic systems. Little information is developed for the
maximum and minimum flow directions or rates that are important in mine design
and developing dewatering programs. Anisotropic aquifer solutions that ad-
dress these shortcomings are discussed below.
Fracture-controlled aquifers in oil shale stratigraphy are prone to ex-
hibit anisotropic flow with the principal axis parallel to the strike of the
primary fracture system. The shape of the drawdown cone for the Upper Aquifer
on Tract C-a, as defined by Weeks et al. (1974), is elliptical, indicating a
strongly anisotropic aquifer. Several solutions to unsteady-state flow in
confined or unconfined anisotropic aquifers have been presented by Hantush
(1966) and Hantush and Thomas (1966). Alternate analytical methods are used
based on available information for the anisotropic system. This information
can be grouped into three cases:
Principal direction of anisotropy known
Principal direction of anistropy not known
Drawdown ellipse for test well known.
Solutions for these cases will be discussed in turn.
Principle direction of anisotropy known (Hantush method)Geological and
geophysical surveys of Oil Shale Tract C-a evaluated surface fault and joint
systems. These data have been condensed into rose diagrams showing principal
and subset joint and fracture systems. Figure 17 shows surface joint strikes
from the outcrops in the vicinity of the mine development plan (MDP) area,
Tract C-a. The primary joint set ranges from N40-70°W with N52°W as the
average strike direction. Secondary and tertiary joint sets are also shown in
the diagram and both have a joint frequency of two to five relative to the
primary system. Figure 18 shows a rose diagram of photo!inear strikes within
the MDP area, Tract C-a, from work conducted by R.A. Hodgson (1979). The pri-
mary linear sets ranges from N45-75°W with N61°W as the average strike di-
rection. Alternate joint systems are also presented in Figure 18. These data
are in agreement with the surface geologic study and suggest the principal
anisotropic flow axis should be about N57°W. Assuming that these data accu-
rately define the principal direction of anisotropic flow (field data show
principal flow direction more to the east), and that information from at least
two groups of observation wells on different radial lines from the pumped well
77
-------
STRUCTURAL STRIKE
AND DIP
AVERAGE NUMBER
IN EACH 10° INCREMENT
12% 16% 20%
20% 15% 10% 5%
PERCENT IN EACH
10° INCREMENT
JOINT STRIKE
JOINT SET RANGE WTD. AVE.*
PRIMARY N40°-70°W NS2°W
SECONDARY N2a°-6(f>t N3S"»E
TERTIARY N10°-20°W N12°W
PERCENT OF TOTAL
JOINTS MEASURED
54
19
19
92
APPROXIMATE
RELATIVE
JOINT FREQUENCY
S
2
2
WTO. AVE. -WEIGHTED AVERAGE STRIKE (COMPASS OIRECTION)
OF ALL JOINTS WITHIN THE SET.
SOURCE. DATA FROM RIO BLANCO OIL SHALE COMPANY
Figure 17. Rose diagram of surface joint strikes in vicinity of MOP
area, Tract C-a (based on eight nearby outcrop stations).
78
-------
71,750 fe«t - TOTAL OF LINEAR LENGTHS WITHIN MAP AREA (A!
47,495 tem TREND NW (66.4%I
24.075 fmt TREND NE (33.6%!
PERCENT OF TOTAL LENGTH
IN EACH 5° INCREMENT
r
STRIKE (S)
LINEAR SET
PRIMARY
SECONDARY
SUBSET
SUBSET
TERTIARY
SUBSET
SUBSET
FOURTH
FIFTH
RANGE WTO. AVG. (Cl
N4S-7S°V»
NS-30°W
N20-30°W
NS-15°W
N6S-90°E
N80-90"E
N8fi-7S°E
NBO-aS°Y»
NSO-60°E
NS1°W
N19°W
N26°W
N12°W
N79°f
N86°E
N71°E
NS4°W
N58°E
*
PERCENT Of TOTAt
LINEAR LENGTHS
MEASURED ID)
22.8
22.7
19,2
5.9
49
10.8
10J3
9.5
7.4
APPROXIMATE
LINEAR LENGTH
FREQUENCY
(E)
6
S
4
1
1
2
2
2
1-2
30.9
NOTES
(A) MAP AHEA OF H0OSC FIGURE MI-I14
(81 REFERENCED FROM OHIO NORTH (J°W OF TRUE NORTH)
(Cl WEIGHTED 3V LENGTHS OF ALL L1NEAHS WITHIN THE SET OR SUBSET
(Ol PERCENT Of TOTAL LINEAR LENGTHS WITHIN MAP AREA (113 LINEARS WHOSE COMBINED LENGTH
I57I.970IM1I
<6I LINEAR SET PERCENTAGE COLUMN INDICATES APPROXIMATE RELATIVE LINEAR LENGTH FREQUENCY
FOR EVERY 1 foot Of LINEAR LENGTH IN THf FOURTH AND FIFTH SETS,«, 5, AND * Itti ARE IN THE
PRIMARY, SECONDARY, AND TERTIARY SETS. RESPECTIVELY
SOURCE R A HODGSON GULF a«0 197(1
Figure 18. Rose diagram of photolinear strikes within MOP area,
Tract C-a (data from R.A. Hodgson, Gulf R&D, 1979).
79
-------
is available, then the transmissivity parallel to the major flow axis (Tv),
minor flow axis (Ty), and the storage coefficient (S) can be determined (see
Figure 19). The procedure and equations developed by Hantush are as follows:
Isotropic methods (Theis, Chow, Jacob) are used on each of the
observation well rays to determine values for the effective
transmissivity (Te), S/TI, and S/T2, Te =
Parameters S/Ti and S/T2 are combined in Equation 3 to provide
values of a and subsequently in Equation 4 to yield Tx and Ty
a » II = cos2(e+an) + m sin2 (9+an) (3)
Tn cos29 + m sin29
where: Tn is the transmissivity in the direction (9+ot) with
the x-axis (Figure 19)
m is equal to
Tx/Ty (Te/Ty)2 . (4)
If an = 1, then Equations 3 and 4 can be combined:
m _ Te _ an cos2 9 - cos2 (9+an)
Ty sin2 (9+0^) - an sin2 9
and m can be calculated because 9, a, a, and Te are known.
Substituting m into Equation 4 provides values of Tx and Ty.
Values of T]_ and T£ can be found by substituting m, 9, and a
into Equation 6 and TI into Equation 3 to find 13:
Tl = Tx/(cos2 9 * m Sln2 9) ' (6)
S is determined from the relationship S/TI and S/T2 and should
be essentially the same.
Principal direction of anisotropy not known (Hantush method) If the
principal direction of anisotropy is not known and there are at least three
groups of observation stations on radial lines from the pumped well, then Tx,
Ty, and S can be determined for the aquifer system. Figure 20 shows the re-
quired observation wells and some of the parameters used in the solution. The
method presented by Hantush is as follows:
Isotropic methods are used to determine Te, S/TI, S/Tg, and S/T3
as discussed above.
80
-------
OBSERVATION WELLS
Figure 19. Illustration of parameters used by Hantush (1966)
(known direction of anisotropy).
Figure 20. Illustration of parameters used by Hantush (1966)
(unknown direction of anisotropy).
81
-------
S/T2, and 8/13 are combined in Equation 3 to determine 32
and 33. 9 can be calculated from the following equation because
ci2 and 33 are known.
tan(29) = "2 (a^l} sin2 a2 " ^2-D sin2 a3
(33-!) sin2 02 - (32'1) sin2 a3
t Substituting 33, 02, 9, and Tg into Equation 5 yields m, and TI,
T2, and 13 are found by substituting Tx, 9, m, a2> a3> ancl al
into the following formula:
T = T / cos2 (9+on) + m sin2 (9+on) (8)
1 1 A
S is then calculated from the relationship(s) S/TI, S/T2, and
5/13 and should be essentially the same value.
Equal drawdown ellipse known (Hantush-Thomas) Hantush and Thomas (1966)
have shown that if the effective transmissivity (Te), the length of the major
flow axis (32), and the length of the minor flow axis (bs) are known for an
anisotropic aquifer, then S, Tx, 3nd Ty C3n be calculsted. To utilize this
method, sufficient observstion ststions sre required such thst equ3l drswdown
ellipses csn be constructed 3bout the test well. Analytical 'methods and equa-
tions presented by Hsntush snd Thomas sre ss follows:
Isotropic methods sre used to determine Te 3nd S/t for each ray
containing observation well(s).
Te is substituted into the formul3(s) presented by Hantush (1966)
3nd drawdown (s) is calculated for any distance along a given
radii for the desired time.
s = __ w(u.)
4ir(Te)
from Equation 4, Tx snd Ty can be determined.
where u1 = r2S/4t(Tn)
r is the radius from the test well
t is the desired time
W(u') is the "well function"
From the s values, one or more equal drawdown ellipses are con-
structed and as and bs are determined (note: if there are suf-
ficient observation points, the equal drawdown ellipses can be
constructed from field data).
82
-------
Tx, Ty, and Tn are calculated using the following relationships:
Tx = as/bs (Te) (10)
Ty = bs/as (Te) (11)
(Te)
The "well function" of W(u') is found using Te, a specific
drawdown (s) ellipse, and a modification of Equation 9
(13)
Corresponding values of u1 are found in tables presented by Wal-
ton (1962) and S is computed from the following relationship:
u, s
4(Te) ts 4(Tn) t
Vertical hydraulic conductivity (leakage) for the Federal tracts has been
calculated through a computer solution for the Neuman-Witherspoon leaky aqui-
fer equation (Neuman and Witherspoon, 1969). If individual permeable zones
within the Upper or Lower Aquifer systems are being evaluated (single or dual
packer tests), leakage becomes a more important parameter and semi confined aq-
uifer conditions more accurately model field conditions. The analytical meth-
ods can be modified for semiconfined conditions by including the leakage
factor (L). This is accomplished by modifying Equation 3 as follows:
Tl _ /LA2 _ cos2 (9+On) + m sin2 (9+ctn)
n/ cos2 9 + m sin2
an=-^Mpl= - = ~ <15)
Tn \L
where Ln = Tnc (c is a constant). The procedure is the same as above except
that Equation 15 replaces Equation 3 and isotropic semiconfined methods are
used to calculate Te and S/Tn.
Remarks--
Anisotropic flow patterns controlled by major fracture systems have been
analyzed by tract developers using the R.E. Glover method and reevaluated by
Kaman Tempo using a technique described by Kruseman and Ridder (1976). This
analysis is discussed in Slawson (1980b). The long-term pump tests conducted
on Tract C-b, the analytical methods, and recommendations for test modifica-
tions are also documented therein.
EVALUATION OF MINE DEVELOPMENT DATA
The third category of methods to obtain hydrogeologic data is evaluation
of mine development data. Primary data sources contained within this category
83
-------
consist of those compiled from the existing monitoring program and ongoing
mine construction. The results of baseline data collection programs on Lease
Tracts C-a and C-b are presented in Sections 5 and 6 (pages 51-128) of a com-
panion report entitled Monitoring Grgundwater Quality; The Impact of In Situ
Oil Shale Retorting (EPA-600/7-80-132).Evaluation of the existing monitoring
program and data compiled therein is discussed in detail in Section 9 (pages
150-185} of that report.
As mine workings are developed, a perspective of the rock fracture and/or
solution cavity system(s) can be gained, which can greatly supplement the data
obtained in the two previous categories (geophysical methods and well testing
procedures). For example, detailed surface geological surveys and analysis of
photo!inear strikes have been used to define anisotropic conditions that will
affect long-term pump tests. These data are presented earlier in this sec-
tion. Additional information can be gained by examining and mapping fracture
surfaces encountered during mine development, observing and recording relative
amounts of water entering the mine in different zones and areas, and sampling
the quality of such water encountered.
The concept of mine development activities done in conjunction with hy-
drogeologic assessment offers a unique opportunity to conduct studies such as
the dewatering and reinjection programs conducted at Federal Tracts C-a and
C-b. Unfortunately, no record was kept of the quantities or quality of waters
transported during these programs and therefore assessment of their utility in
defining the hydrogeologic framework could not be made.
84
-------
SECTION 4
SAMPLING METHODS
This section addresses the sampling methods currently being utilized in
the oil shale region. Factors that influence the sampling methods are also
discussed. These factors include well construction, sample handling, and
preservation techniques.
WELL CONSTRUCTION FACTORS
The groundwater hydrology in the oil shale region can be significantly
affected by the stratigraphy and structure of the area. Therefore, it is im-
portant to develop a site-specific characterization of the hydrogeology prior
to the development of well specifications for a groundwater quality monitoring
program. The purpose of this characterization work is to identify intervals
of distinct water quality and hydraulic character. It is cost-effective to
coordinate this hydrogeologic analysis with the preliminary resource explora-
tion and evaluation efforts.
In addition to the hydrogeologic considerations, monitoring needs are an
important consideration. Each well should be located and designed according
to the objectives of an overall monitoring strategy. For instance, wells
needed exclusively for piezometric measurements require accessibility only for
water-level measuring instruments and should be designed with a minimum inner
diameter. The need to collect water quality data or to conduct pump or injec-
tion tests dictate a different well design, as do wells monitoring two or more
aquifers (i.e., multicompletion wells).
Preliminary site-specific characterization of the hydrogeology and objec-
tive analysis of the data requirements are essential to proper well design
procedures. If this type of approach is utilized, both costly and timely well
recompletion efforts will be avoided.
Discussed below are some aspects of well design and construction that
should also be considered prior to implementation of a groundwater quality
monitoring program.
Well Construction
Open Well or Perforated Over Entire Aquifer
This type of well construction is common in the oil shale region. When
the rock is well consolidated and competent, such in as the Lower Aquifer, the
85
-------
well is left open. In the Upper Aquifer wells, where tubing is usually per-
forated over the entire interval to maintain accessibility, caving is still a
problem when semi consolidated rock is intercepted by the well. Both types of
well construction are designed to monitor a single vertical interval, in this
case the entire Lower or Upper Aquifer.
Although this type of construction is commonly utilized for groundwater
quality monitoring in the Piceance Basin, there are some disadvantages asso-
ciated with the design. The regional hydrogeologic concept of a dual aquifer
system (i.e., Upper and Lower) separated by the relatively impermeable Mahog-
any Zone has resulted in this design. However, on a smaller scale, the
groundwater hydrology is more complicated. It has been shown that the deep
aquifer of the oil shale region is actually composed of numerous fractures and
cavities (i.e., secondary porosity) that will contribute variable water qual-
ity to a well completed over the entire interval. A sample collected from
this well may reflect the composite water quality of the entire section or the
water quality of a high head interval. In any case, the sample may not repre-
sent the true groundwater quality for a given aquifer. A potential pollutant
of low concentration present in this situation may become diluted below detec-
tion limits in a composite sample, or it may not be detected at all in samples
collected from an aquifer dominating the open section (i.e., a high head in-
terval). In addition, a well completed over the entire section may not pro-
vide any information on the source of the contaminant.
In addition to water quality considerations, the hydraulic characteris-
tics (e.g., transmissivity) of the different aquifer intervals are difficult
to determine with this type of well design. Although an aquifer test will
provide composite information on all of the aquifer intervals, the test proce-
dure, without elaborate and costly modification, would be inconclusive for
specific aquifer intervals. Furthermore, the interconnection of these differ-
ent aquifer intervals can result in the collection of water quality samples
from a layer exhibiting greater head rather than a composite including the ad-
jacent layers.
Multiple Completion Wells-
Multiple completion wells are designed to monitor more than one aquifer
interval (see Figure 21). The wells of this type in the Piceance Basin have
two to four tubing strings per well, each of which are perforated in a spe-
cific aquifer interval. Potential interconnection among the different aqui-
fers is prevented by the placement of cement grout in the annulus above and/or
below the perforated zone and in some cases by bridging plugs used in conjunc-
tion with cement. This type of well construction is designed to minimize the
problem of nondelineation of the vertical distribution of groundwater quality
and hydraulic characteristics exhibited by the different layers within the
Lower and Upper Aquifer zones.
Although this type of well construction provides for more representative
sample collection from the various horizons, there are some problems associ-
ated with the present design utilized in the region. These problems include:
86
-------
500
a.
UJ
Q
1,000
1,500
1.7101-
.''J-:-'';.rV:VV i;
J'J'i'-y.--/' 'I..';"-..*'
"''""'
f-rr
V: j
'-.I
8-5/8-inch CASING AT 156 feet CEMENTED
TO SURFACE
7-7/8-inch HOLE DRILLED TO 1,036 feet
6-3/4-inch HOLE DRILLED TO 1.710 feet
STRING No. 4: 2-3/8-inch tubing
OPEN-ENDED AT 550 feet
-TOP OF CEMENT 792 feet BY CBL
STRING No. 3: 2-3/8-inch TUBING CEMENTED
AT 1,040 feet PERFORATED
820 to 1,005 feet
'"'« j
/ V',*
^
ra
!.W»
If";;'.
£'
^r**
'&
i
..*.-;
^
i
i
.-:.-, ,;, -f-.y-:<.
<;;{ <;&>;;
w-i '- '/ '
K&| Sf!r:*r';^
l^^{ L?s-^
rlY/'i fe6ft!{l
i',~-.J t-'.J/. .".'-.''
i;v f^'f'«-
s.i*j fair" ('""' '''^
C'-j':{ li'Si1- ?''''''
Ui"'J'rtf' '^
i?3?* ii
^.;aSH
^r1:':-/::^.-*^
&6$j$fj(
i
1
1
r
rriTAi ncoTu
STRING No. 2: 2-3/8-inch TUBING CEMENTED
AT 1,501 feet PERFORATED
1,050 to 1,480 feet
STRING No. 1: 2-3/8-inch TUBING CEMENTED
AT 1,709 feet PERFORATED
1,530 to 1,680 feet
Figure 21. An example of multiple completion well, Tract C-b
Well S6-21.
87
-------
t From a technical standpoint, pumping is the preferred method for
assuring the collection of representative samples. Because the
diameter of the tubing strings in these multiple completion wells
will not accommodate a submersible pump, this type of well design
is not recommended in the groundwater monitoring network.
Although cement grout and bridging plugs are utilized, it is dif-
ficult to completely ensure that interconnection will not occur
between different aquifer intervals using these techniques. If
interconnection does occur, water quality samples collected from
the well may be nonrepresentative and costly recompletion efforts
may be required.
The structural properties of the small-diameter tubing strings
are, in some settings, insufficient at the depths required for
monitoring deep aquifers. Failure of a tubing string can result
in very expensive and time-consuming replacement.
It should be noted that the above referenced groundwater monitoring sys-
tem was derived from a well recompletion effort conducted on a tract in the
Piceance Basin. Many of the problems cited are due to the well design prior
to recompletion (e.g., the 2-5/8-inch tubing strings). It is strongly recom-
mended that future multiple completion designs be modified to allow for 6-inch
diameter wells. This aspect would not only provide for sample collection by
pumping but also significantly reduce potential failure of a well at depth.
To accommodate a submersible pump for sample collection, larger-diameter
boreholes are required for installation of the larger diameter, multiple com-
pletion wells. The borehole should be drilled large enough to accommodate the
casing, 6-inch-diameter well strings, and cement grout. These proposed well
specifications require an annulus of 10 to 12 inches. The cost implications
(in 1980 dollars) of this increase in diameter are substantial during the ini-
tial drilling operation, on the order of $26 to $30 per foot. Casing costs
are estimated (in 1980 dollars) to be in the range of $10 to $13 per foot.
Drilling, casing, and equipment costs can be obtained for comparison purposes
in Everett et al. (1976). In Everett et al. (1976), a methodology for updat-
ing the 1976 costs is provided.
Although the costs are substantially higher for the multiple, 6-inch well
design, the sampling approach is more effective compared to the smaller diame-
ter tubing strings. The common procedure for sampling these smaller diameter
wells is a bailer, which represents a passive method of groundwater quality
monitoring. The effectiveness of a well in providing baseline water quality
data and/or detecting potential pollutant excursions using bailing techniques
is dependent upon the location of the well and the hydrologic gradient. In
comparison to pumping, the passive nature of a monitoring program utilizing
bailing as a sample collection method requires an additional number of wells
to be incorporated in the network. The larger diameter wells allow samples to
be collected by pumping. Since pumping samples a larger cross-sectional area
of the aquifer, fewer wells are required in the monitoring program. The ac-
tive nature of this sampling approach will also allow the detection of any
-------
potential pollutants present in the zone of groundwater flow intercepted by
the pumping.
Well Size
The diameter of the monitoring well should be large enough to accommodate
the sampling tool. Where a submersible pump is to be utilized in deep aqui-
fers, the well diameter should be at least 6 inches. For shallower alluvial
wells, a 4-inch-diameter well is adequate to accommodate a submersible pump.
Wells from which water-level measurements are required need only be 1 inch in
diameter.
The diameter of the borehole into which the casing is placed must be at
least large enough for proper casing placement. It is recommended that the
borehole be at least 2 inches larger than the casing in the multiple comple-
tion wells to permit proper placement of the cement grout around the casing
and adjacent to the layers or aquifers that are to be sealed from the well.
The approximate costs of drilling, casing, grout placement, etc. are pro-
vided in the discussion on well construction above.
Annular Seal
The annular space consists of the area between the casing material and
the borehole. This space is unavoidable regardless of the drilling method or
casing installation. To prevent contamination of the well from surface drain-
age or from formations other than the aquifer to be monitored, this annular
space should be sealed.
The most common material used in providing an annular seal is cement
grout. Cement grout is a fluid slurry composed of a mixture of Portland ce-
ment and water. The ratio of water to cement for a suitable grout mixture is
5 to 6 gallons of water per 90 pounds of cement (Johnson Division, 1966).
Mixtures of more than 6 gallons of water to 90 pounds of cement should not be
used because the amount of shrinkage upon settling increases with water con-
tent, producing an inadequate annular seal. In addition, the water used for
the grout should be free of oil or other organic material, dissolved solids
content should be less than 2,000 mg/1, and the sulfate content should be kept
to a minimum.
The correct placement of the cement grout is equally as important as its
composition. To assure that the grout will provide a satisfactory seal
against potential pollutants from the surface or aquifers not to be incorpo-
rated in the monitoring well, implacement should be continuous with the cement
slurry introduced through a pipe 2 to 4 inches in diameter. Introducing the
grout through a pipe to the desired depth will prevent any gravitational sepa-
ration of the cement due to "free falling." This aspect is particularly im-
portant in the deep aquifer wells where the grout may have to be placed at
great depths to prevent aquifer interconnection.
89
-------
Casing Material
Well casing materials can play a critical role in a groundwater quality
monitoring well. The potential influence of the casing material on groundwa-
ter chemistry, the structural properties of the well, and economics of the
monitoring program are all important considerations. In general, the proper
selection of casing requires site-specific evaluation of the monitoring objec-
tives, groundwater characteristics, and the anticipated well specifications.
Some properties of various well casing materials that should be considered and
evaluated prior to installation are presented below.
Plastic Casing and Screens-
Plastic well casing is widely utilized in groundwater monitoring wells,
particularly at shallow depths. The most common type of plastic casing used
is polyvinyl chloride (PVC).
The primary advantages associated with PVC casing include:
t Nonconducting electrochemical reactions will not be a factor
affecting the groundwater quality
Inert resists chemical attack (with the exception of ketones,
esters, and aromatics (U.S. EPA, 1977)
Lightweight easy to handle and install
Inexpensive when compared to other casing materials (i.e., steel
and stainless steel) if the previous recommendations are fol-
lowed (i.e., 6-5/8-inch-diameter wells), the PVC should be at
least schedule 40 (19/64 inch) in thickness; with these specifi-
cations the PVC would cost approximately $2.50 to $3.00 per foot
(Everett et al., 1976).
The disadvantages associated with PVC casing include:
The structural properties of PVC may be inadequate at depth.
Given the well consolidated rock in the oil shale region, this
should not be a great problem provided the casing is installed
correctly and the pipe schedule is selected properly.
If PVC is cemented together, organic solvents will be introduced
into the groundwater system, resulting in anomalous trace organic
determinations. To alleviate this problem, it is suggested that
pressure joints be used for PVC connections.
PVC possesses a hydrophobic surface when initially introduced
into the groundwater system, causing trace organics to be ex-
tracted from the groundwater until equilibrium between the PVC
and the groundwater system is reached.
90
-------
Steel Casing and Perforated Tubing--
Steel casing and perforated tubing is widely used in the deep aquifer
wells of the Piceance Basin. The basic disadvantages of utilizing steel mate-
rials in a well are:
Steel casing and tubing materials are active conductors and will
be involved in electrochemical reactions, in most cases causing
the plating of iron (although iron can go into solution as well).
Steel materials can contaminate water quality samples collected
from the well through the introduction of trace metals derived
from the casing or tubing. Also, the sorption of trace metals or
organic constituents may occur due to metal oxides.
In the Piceance Basin, perforated steel tubing commonly has to be
replaced due to the corrosive groundwater environment, which is a
costly procedure. Excessive corrosion can result in nonrepresen-
tative samples being collected from the well.
The structural properties of the small, 2-5/8-inch perforated
tubing strings are, in some cases, not sufficient to withstand
deep aquifer conditions. In this situation, the tubing will fail
and accessibility to the well will not be maintained.
Steel materials cost approximately $1 to $1.75 more per foot than
PVC.
Steel materials can be more difficult to handle.
Many of these disadvantages are due to the restricted well diameters. If
the inner diameter of the wells were expanded to 6 inches, nonpumping sampling
techniques could be discontinued. In addition, the collection of groundwater
samples that reflect the effects of the steel casing material would be mini-
mized, provided the well is flushed prior to sample collection. Failure of
the wells can also be significantly reduced, if not completely eliminated,
with the increased diameter.
Stainless Steel Casing and Screens-
Stainless steel materials technically surpass any material for groundwa-
ter quality monitoring purposes. They are inert to all chemical reactions and
will not contaminate the groundwater environment. Furthermore, stainless
steel materials are structurally stable under any conditions if selected
properly.
The major detriment to installing stainless steel materials in monitoring
wells is the cost. Stainless steel screen generally costs between $25 and $35
per foot, significantly more than other casing materials. The advantages of
the stainless steel material do not compensate for the economics, particularly
when the disadvantages associated with the other materials can be mitigated if
correct sampling procedures are followed.
91
-------
Well Security and Protection
As with any well, proper procedures should be taken to ensure the protec-
tion and security of the monitoring well after installation. These procedures
will prevent the inadvertent or deliberate introduction of materials into the
well. Proper protection will also deny accessibility to small rodents. These
foreign materials can notably affect the groundwater quality data obtained
from the well, particularly if nonpumping sampling techniques are practiced.
Well security can be best acquired by placing a locking cap on the well.
If continuous monitoring equipment (e.g., Stevens Water Level Recorder) is em-
ployed, it should be protected as well. This can usually be done by welding a
metal box with hinges onto the well casing and installing a lock on the metal
box.
WELL DESIGN AND SAMPLING COSTS
Well Design Costs
Approximate costs for each well design are provided below. These costs
assume that 24 sites were selected for each well design. This assumption pro-
vides for a per-well distribution of the base costs that are accrued by a mul-
tiple drilling operation. Such base costs include capital requirements for
mobilizing drilling equipment to the region, contracting geophysical equipment
on a monthly basis, and delivery of materials (i.e., tubing strings, casing,
packers, etc.) to the site.
The costs provided
tributed among the Upper
sign would have 12 dual
Aquifer (see Figure 22),
to construct 2 wells per
one in the Lower Aquifer
for each approach since
below assume that the number of wells were evenly dis-
and Lower Aquifer. For instance, the Tract C-a de-
completion wells montoring both the Upper and Lower
whereas the U.S. Geological Survey (USGS) approach is
site (see Figure 23), one in the Upper Aquifer and
This distribution will assure comparable well costs
there are the same number of wells in each aquifer.
The approximate costs of each well design are as follows:
Design
USGS
Upper Aquifer well
Lower Aquifer well
Tract C-a
Dual completion well
Tract C-b
Multiple completion well
Approximate cost
per well
(dollars)
18,000-20,000
35,000-38,000
35,500-38,000
39,000-44,000
Approximate cost
per site
(dollars)
53,000-58,000
35,500-38,000
39,000-44,000
92
-------
DEPTH (ft)
8-5/8-inch SURFACE CASING
CEMENT
11-inch HOLE SIZE
25-foot CEMENT
(TO *850 feet)
HALLIBURTON SPEED E-LINE
BRIDGE PLUG
PACKER (TENSION)
CEMENT
4-1/2-inch LINER
6-3/4-inch BOREHOLE
HORIZONTAL SCALE: 1""1'
VERTICAL SCALE: 1" - 200'
1,800
Figure 22. Typical recompleted Upper Aquifer monitoring well
for Tract C-a (derived from Rio Blanco Oil Shale
Co., March 1979).
93
-------
UPPER
AQUIFER WELL
DISTANCE BETWEEN
WELLS = 100 feet
500-
1000-
x
a.
in
a
1500-
2000-
2500-
UINTA
FORMATION
UPPER
AQUIFER
~l
8-5/8-inch OD
STEEL SURFACE CASING
11-inchANNULUS
CEMENT
LOWER
AQUIFER WELL
-6-3/4-inch BORE HOLE
6-5/8-inch OD STEEL CASING
MAHOGANY
ZONE
GREEN RIVER
FORMATION
5-1/2-inch BORE HOLE-
LOWER
AQUIFER
HORIZONTAL SCALE. 1" > V
VERTICAL SCALE: t " 250'
Figure 23. USGS Upper and Lower Aquifer monitoring well design,
Sampling Costs
The approximate sampling costs for each well design and corresponding
sampling method are presented in the subsection that follows ("Sample Collec-
tion Methods"). In addition to the sampling methods currently being utilized
in the oil shale region (i.e., bailing, swabbing, and portable submersible
pump), a fixed submersible pump was analyzed as a sampling approach. For com-
parison purposes, the costs for each sampling method were developed under the
assumption of a quarterly sampling frequency of 12 Upper and Lower Aquifer
wells for a 5-year period. Sampling costs include the wages for personnel,
the materials utilized, and equipment capital requirements.
94
-------
The sampling costs and corresponding well design costs (derived from well
design cost data presented above) are given in Table 13. Based on the data
presented in Table 13, it is apparent that the bailing method is the best ap-
proach from a cost perspective. The portable submersible pump and swabbing
methods are very expensive compared to the bailing method and, therefore, are
not recommended. Although the fixed submersible pump is more expensive to
utilize than bailing, this method should not be ruled out due to the technical
advantages of the approach. Each method is discussed further in the discus-
sion of "Sample Collection Methods" (below).
The initial step in developing the well construction costs was to iden-
tify the specifications for each design. The costs for the drilling opera-
tions, geophysical logging, and materials relative to each of the design
specifications was provided by companies dealing in these areas. Three compa-
nies were contacted for each of these areas and average costs were developed.
These construction costs were then reconfirmed by the respective designers
(i.e., USGS, Tract C-a developers, and Tract C-b developers).
In most cases, the sampling costs were provided through correspondence
with USGS, Tract C-a developers, and Tract C-b developers. The costs that
were not included in this correspondence were developed in a similar manner to
the well construction costs.
MONITOR WELL PLACEMENT
The placement and design of monitoring wells is defined by the design of
the MIS operation, the site-specific hydrogeology, and by the potential mobil-
ity of the constituents from the MIS retorts. An earlier companion report
(Slawson, 1980a) examined proposed monitoring programs for Federal Oil Shale
Tracts U-a and U-b to identify information deficiencies and to develop a moni-
toring design program. Monitor well program designs are developed for differ-
ent aspects of the MIS mining operations. Specific examples are presented
that show monitor well placement for proposed and existing alluvial, Bird's
Nest, and Douglas Creek aquifers. Additional wells are identified in the sat-
urated zone of the Uinta and Green River Formations above the Bird's Nest aq-
uifer. Source specific monitoring systems for spent shale landfills include
observation wells and multiple completion wells, as well as geophysical and
unsaturated sediment monitoring devices. For greater detail on monitor well
placement, see Slawson (1980a), entitled Groundwater Quality Monitoring of
Western Oil Shale Development: Monitoring Program Development.
SAMPLE COLLECTION METHODS
Three predominant methods of sample collection are commonly utilized in
the oil shale region: bailing, swabbing, and pumping. Each of these sampling
procedures is discussed below with respect to advantages and disadvantages of
the methodology, as well as the approximate costs for initiating and conduct-
ing each procedure. Related issues, i.e., quality and custody control of the
water quality samples obtained, are discussed by Everett (1980) and Slawson
(1980b).
95
-------
TABLE 13. SAMPLING COSTS
Fixed Submersible Portable Submersible Bailing Swabbing
Item Pump Pump (USGS) (Tract C-a) (Tract C-b)
Well Construction 53,000-58,000a 53,000-58,000 35,500-38,000 39,000-44,000
Sampling Costs
Capital Requirements 61,800-79,800 55,000-60,000 8,000-10,000 N/Ab
Operational Requirements
(Quarterly) 200-400 1,400-1,700 200-400 16,000-18,000
Labor (quarterly) 135-200° 11,200-14,000d 135-200° 3,500-4,300e
Five-year Total (including
construction of 12 monitoring
well sites) 704,500-787,800 943,000-1,072,000 440,700-478,000 858,000-974,000
Notes:
ill CO!
Assumes similar well construction for fixed pump as with portable pump.
All costs in 1980 dollars.
a
Tract C-b contracts swabbing rig, thereby eliminating capital requirements.
cAssumes the sampling of eight wells per day.
Assumes sampling of one well per day.
eAssumes the sampling of three wells per day.
-------
Ba
Bailing involves introducing a hollow cylinder that is supported from the
surface into the well. Figure 24 portrays the features of a Kemmerer sampler,
a commonly used bailer. The cylinder can be tripped to close at any desired
depth thereby collecting a sample. The important aspect of the Kemmerer sam-
pler is that it allows the water to flow through the cylinder, thus permitting
samples to be collected from any depth. Samplers that are open at the top and
sealed at the bottom do not have this flow-through characteristic and should
not be used because the sampler is generally filled with the first water en-
countered in the well, i.e., the water near the static water level.
D510
m
J 1 I
II
II
2fK
II
I)
II
4c
n
M
n
n
_ i ^.
ch- i-ch *
<: '} dh
I
.St 9
U-^J h
"9 J
j$
Iv
m
o
spr P
spr
J^o :
ch Chain that anchors upper valve to upper interior guide
Rubber drain tube
Brass drain tube
Interior guide fastened to inner surface of sampler
Rubber tube
Jaw of release
Jaw spring
Lower valve
Messenger
Opening interior of drain tube
Pinch cock
Upper release spring operating on horizontal pin, one end of which fits into
groove on central rod
spr Spring fastened to lower internal guide and operating in groove on central
rod to provide lower release
Stop on central rod
Upper valve
Left: View of complete sampler with valves open
Top Right: Another type of construction of upper valve and tripping device
Bottom Right: Another type of construction of lower valve and drain tube'
Figure 24. Features of the modified Kemmerer bailer
(P.S. Welch, Limnological Methods, p. 200,
Figure 59).
97
-------
The major advantage of utilizing the bailing method is that it allows
samples to be collected from small-diameter wells that have relatively deep
static water levels, a situation that generally restricts the use of other
sampling methods. Bailing is also very simple to use and does not require a
large number of personnel for operation. It is also fairly inexpensive, with
capital requirements (i.e., bailer, winch, power source, truck, etc.) of
$8,000 to $10,000 (in 1980 dollars).
There are a number of potential problems associated with bailing water
quality samples. Extreme variations in the water quality data can be observed
when the depth selected for sampling is inconsistent. This is pronounced when
the well is completed in an aquifer possessing multiple permeable intervals,
which may contribute dissimilar water quality. The groundwater in such wells
can be stratified, resulting in noticeable vertical changes in water quality.
Schmidt (1977) attributes this stratification to the distinct water quality in
each permeable zone penetrated by the well. Schmidt (1977) also suggests that
variations in the composition of aquifer materials with depth and possible
differences in the sources of recharge can modify groundwater quality in wells
penetrating these intervals.
An additional problem is the nonrepresentative nature of the sample col-
lected. Water standing in the well bore above the screened or open interval
will be isolated and have little or no mixing with natural groundwater. This
stagnation effect is particularly pronounced in little-used or nonpumping
wells. Samples collected from this stagnant zone are not indicative of the
groundwater quality and will result in unrepresentative data. Factors con-
tributing to the unrepresentative nature of the samples collected from the
well bore include the introduction of unnatural constituents through the in-
teraction between the casing with the groundwater system, as well as foreign
material entering the well from the surface. Furthermore, changes in pH, and
subsequently in water quality, can be induced through the variations in pres-
sure and C02 dissolution in the well bore (Summers and Brandvold, 1967).
The magnitude of the vertical variations that can be observed during sam-
pling a well is shown in Figures 25, 26, and 27. These vertical profiles were
compiled by collecting samples from specific depths via a bailer and perform-
ing specific conductance and temperature measurements for each sample. The
results of the detailed water chemistry analysis for selected samples are
shown in Table 14.
The temperature and conductivity profile for Well GS-13 (well diagrammed
in Figure 28) show a declining level of conductivity over the approximately
375 feet of water standing in the well bore. The conductivity measurement ob-
tained near the static water level was 2,300 umho/cm, compared to a measure-
ment of 1,600 ymho/cm obtained near the bottom of the open interval (see Fig-
ure 25). Temperature measurements were much more uniform with depth.
Appreciable increases in conductivity with depth were also noted in Wells
D-17 and D-18 (Figures 26 and 27, respectively). In Well D-18, an order-of-
magnitude increase in conductivity was observed in a very small interval near
the bottom of the well. Above this level the conductivity was very stable.
The decline in conductivity with depth noted in Well GS-13 is also seen in the
98
-------
DEPTH (ft)
0
100
200
300
400
500
600
700
800
900.
POTENTIOMETRIC SURFACE 425 ft
I I I I I I I
I I I I
Figure 25.
16 17 18 19 20 21 22 23 13 14 15 16 17
SPECIFIC CONDUCTANCE
(Mmhm/cm @ 25° C) X 10?
TEMPERATURE (°C)
Variation in specific conductance and temperature
with depth, Upper Aquifer Well GS-13, Tract C-a.
99
-------
DEPTH
0
100
200
300
400
500
600
700
800
900
1000
(ft)
1100
POTENTIOMETRIC SURFACE 373 ft
I I
I
I I I
I
12 14 16 18 20 22 24 26
15 16 17 18 19 20
TEMPERATURE (°C)
SPECIFIC CONDUCTANCE
(Mmhot/cm @ 25° C) X 102
Figure 26. Variation in specific conductance and temperature
with depth, Lower Aquifer Well D-17, Tract C-a.
100
-------
DEPTH ft)
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
1300
1400
1500
1600
17OTI
-
-
-
_j
i
POTENTIOMETRIC SURF
* <,
\
I I I I I
I
I I I
10 20 30 40 50 60 19 20 21 22 23 24
SPECIFIC CONDUCTANCE TEMPERATURE (°C)
(nmhos/cm@25°C)X102
Figure* 27. Variation in specific conductance and temperature
with depth, Lower Aquifer Well D-18, Tract C-a.
101
-------
TABLE 14. VARIATION IN WATER QUALITY WITH DEPTH IN SELECTED DEEP AQUIFER WELLS, TRACT C-a
o
PO
Depth (feet)
Well GS-13
Constituent*
pH
Specific conductance
Total dissolved solids
Calcium
Magnesium
Sod i urn
Potassium
Bicarbonate
Carbonate
Sulfate
Chloride
Fluoride
Ammonia
Arsenic
Boron
Mercury
Selenium
Dissolved organic carbon
Note:
a
Constituent units are mg/1
450
7.3
1,940
1,409
67
118
281
0.8
775
<1
610
13.4
0.48
<0.1
0.01
0.29
<0.001
<0.01
28.6
except for
575
7.5
1,559
1,140
46
89
213
0.5
551
<1
466
11.4
<0.1
<0.1
<0.01
0.31
<0.001
<0.01
30.6
pH units
725
7.5
1,344
1,160
42
82
194
0.4
637
<1
262
35.7
<0.1
<0.1
<0.01
0.31
<0.001
<0.01
27.2
and specific
Well D-17
475
9.1
1,344
1,093
3.4
32
372
1.9
686
8.8
304
28.2
1.6
<0.1
0.01
0.31
<0.001
<0.01
24.1
conductance
875
9.0
1,790 2
1,174 1
3.9
30
417
2.6
898 1
13.1,
170
53.6
2.6
<0.1
<0.01
0.53
<0.001
<0.01
35.0
in pmhos/cm
990
8.6
,210
,524
5.0
24
557
2.3
,355
4.8
118
59.0
5.6
<0.1
<0.01
0.73
<0.001
<0.01
25.7
at 25°C.
Well
1,400
8.2
3,856
2,954
2.2
6.6
1,224
2.3
3,089
<1
97
47.2
9.3
<0.1
<0.01
0.79
<0.001
<0.01
12.4
D-18
1,500
8.1
64,794
37,839
2.8
6.8
16,816
14.2
45,682
<1
220
480
10.1
3.9
<0.01
0.87
<0.001
<0.01
21.0
-------
DEPTH (ft)
BRIDGE PLUG
CEMENT
1751
Figure 28. Well diagram of Upper Aquifer Well GS-13, Tract C-a.
103
-------
water chemistry data (Table 14). The largest changes with depth were observed
for the major inorganic ions. The trace constituents and DOC were generally
more stable with depth. The general ionic composition is fairly consistent at
all three depths in Well GS-13, although sulfate concentrations decreased with
depth to a greater extent than most other ions.
The increases in conductivity with depth observed in Wells D-17 and D-18
are also consistent with water chemistry analysis (Table 14). Most of the
conductivity increase in Well D-17 can be attributed to increased sodium bi-
carbonate concentration. Magnesium, carbonate, and sulfate levels decline
with depth. Thus, the salinity of the water increased and the ionic composi-
tion of the well water changed.
The variation in water quality with depth, as indicated in Figures 25,
26, and 27 and Table 14, demonstrate the importance of consistent sample col-
lection depths. This is a very critical area and essential to accurate data
interpretation. The depth selected for sample collection is of equal impor-
tance. It is obvious that a sample collected near the static water level or
in the cased section above the aquifer is unrepresentative of the groundwater
system. Therefore, it is recommended that samples be collected from the por-
tion of the well that is screened or open and adjacent to the aquifer.
To collect a representative sample using the bailer method and to assure
the above-cited effects are minimal, at least one well volume of water should
be evacuated from shallow wells in which the groundwater movement is very
slow. This is particularly important because stagnation effects can greatly
influence the water quality of these low-yielding wells. Once one well volume
has been removed, a representative sample can be obtained. In very-low-yield-
ing wells where the evacuation process has resulted in dryness, the well
should be allowed to recover prior to sample withdrawal.
Removal of one well volume is very impractical for eliminating stagnant
water in a deep well with the bailer method because of the quantity of water
obtainable from the well on each down-hole trip. Attempting to evacuate one
or more well volumes would be very time-consuming and inefficient for the
field personnel. As previously demonstrated, representative samples of the
natural groundwater system can be obtained by sampling adjacent to the open or
perforated section. Marsh and Lloyd (1980) have indicated that this is par-
ticularly true for wells that monitor aquifers where significant groundwater
movement is occurring. If this type of approach is to be utilized, the hy-
draulic characteristics of the aquifer to be monitored should be determined.
These characteristics can be best determined during the initial drilling oper-
ations. However, if postdrilling determinations are necessary, down-hole
flowmeter surveys and other geophysical methods can be beneficial.
Additional modification of the sample chemistry can occur when transfer-
ring the sample from the bailer to the sample container. Precipitation of
easily oxidized constituents by introducing atmospheric oxygen during the
transfer and altering the natural oxidation-reduction potential is of primary
concern. The loss of dissolved gases is also a potential problem during the
transfer process. To alleviate the effects of the sample transfer, sample
104
-------
contact with the air should be minimized. Furthermore, caution should be
taken when transferring the sample in order to prevent unnecessary agitation.
An additional consideration when sampling is the number of samples neces-
sary to determine the difference among data collected during different sam-
pling periods. To accurately address this consideration, the variability of
the data needs to be established. To identify this variability, four repli-
cate samples were collected from selected depths in four wells. The sampling
program involved the (1) collection of field data (i.e., specific conductance,
temperature, and pH) and (2) collection of samples for detailed chemical
analysis.
The results of the statistical analysis indicate that only one sample is
necessary for accurately determining a difference among data sets for the ma-
jor constituents (e.g., bicarbonate, sodium, TDS). This is also true for con-
stituents that are intermediate in concentration (e.g., sulfate, chloride).
For the constituents that are low in concentration (.e.g, calcium, magnesium,
fluoride, etc.), an extremely high number of samples appears to be required.
However, this aspect can be attributed to the statistical procedure and it is
therefore not necessary to collect more than one sample to characterize these
constituents.
In summary, the following procedure is recommended for collecting samples
from a well when using a bailer:
1. Use a flow-through type of bailer (e.g., Kemmerer sampler)
2. Compile well completion data; of particular importance is well
diameter, depth to aquifer, aquifer thickness, and total depth
3. For shallow wells with very slow groundwater movement, estimate
the well volume from the well completion data and extract at
least one well volume previous to sample collection.
4. For both shallow and deep wells, select a sampling point adja-
cent to the aquifer
5. Consistently sample from the same depth and adjacent to the aq-
uifer during every sampling effort
6. It is necessary to collect only one sample from the well
7. Measure temperature, specific conductance, and pH in the field.
Pumping
The use of a submersible pump is a common procedure for sample collection
from the alluvial wells on Tract C-b and the deep wells monitored by the USGS.
For this sampling method, the submersible pump is introduced to the desired
depth and a sample collected from the discharge line. A typical pumping appa-
ratus configuration is shown in Figure 29.
105
-------
110-V POWER
SUPPLY
ELECTRIC POWER CORD-
CASING-
WATER METER
DISCHARGE PIPE
-STATIC WATER LEVEL
CEMENT-
DRILLHOLE-
PUMP
INTAKE
- MOTOR
Figure 29. Typical pump apparatus configuration
(after Slawson, 1980b).
The use of the submersible pump for sample collection is the superior
sampling approach. Submersible pumps can be used to collect samples from any
depth provided the pump is properly selected and the well is conducive to
pumping. Submersible pumps can efficiently extract sufficient volumes of wa-
ter and eliminate the stagnant well bore water, thereby allowing representa-
tive sample collection. Extraction can be performed in a relatively short
period of time. Additional advantages of submersible pumps include:
Easy installation and withdrawal from a shallow well (although a
great deal of effort may be required for a deep well)
t Little maintenance is required
106
-------
They can be used as portable or fixed pumps; both Tract C-b and
USGS personnel have designed a truck-portable pump capable of ef-
ficiently servicing a suite of wells
The well discharge can be easily controlled and both very low or
very high discharges can be obtained
They have relatively little effect on the native aquifer water
quality (nitrogen and airlift methods, however, can substantially
alter the iron content and pH of the water and, therefore, are
not recommended).
Although these characteristics make the submersible pump the preferred
sampling approach, some aspects may preclude its use. The primary disadvan-
tage of the submersible pump is the minimum size requirements of the well an-
nulus. Submersible pumps generally require a minimum 4-inch-diameter well for
shallow sampling efforts and a 6- to 8-inch-diameter well for deeper sampling.
As most of the wells in the study area have 2-5/8-inch tubing in the deep
wells, the use of a submersible pump in the sampling program is limited.
An additional disadvantage of this approach is the capital requirement,
which can be significantly higher than other sampling devices. The Tract C-b
truck-portable pump, which is capable of sampling at depths up to 100 feet,
costs $10,000 to $12,000 in 1980. However, the USGS portable pump rig, which
is designed to be set at depths of 500 to 600 feet (i.e., Upper Aquifer
wells), requires an initial expenditure of $50,000 to $60,000. Rigs designed
to sample Lower Aquifer wells require a capability of pumping from depths of
1,000 to 1,500 feet and require an initial expenditure (in 1980 dollars) in
excess of $70,000, substantially more than the capital requirements for bail-
ing. In addition, the time required for pump placement and withdrawal at
these depths is about 7 to 12 hours, depending on the depth of the pump place-
ment (i.e., sampling Upper or Lower Aquifer).
As pointed out above, a mobile pumping rig for sampling deep wells (i.e.,
Upper and Lower Aquifers) is very time-consuming and expensive, particularly
on a frequent sampling basis. A more feasible approach may be a fixed pump in
each well and a mobile generator for a power source. This approach requires
substantially less manhours because the time for pump placement and withdrawal
is eliminated. The approximate initial expenditure for a submersible pump in
each well, capable of pumping from up to 1,600 feet at 40 gpm, is $6,600 to
$7,800 (in 1980 dollars). Additional expenditures of $950 to $1,980 (in 1980
dollars) are required for each well for discharge pipe, power line, etc., de-
pending on the depth of the pump.
To employ this sampling method efficiently, some expertise in the follow-
ing areas is required: pump placement, discharge rate, duration of discharge,
and representative sample collection. Although the submersible pump repre-
sents the most reliable method for collecting representative groundwater sam-
ples, incorrect procedures can produce inaccurate data. In addition, applying
consistent procedures during each sampling event allows better comparison
among the data collected on different dates.
107
-------
Before establishing the sampling procedures for each well (i.e., pump
placement, discharge rates, etc.), the wells need to be individually evalu-
ated. Samples for water quality analysis should be collected only after the
discharge has equilibrated. Additional factors to be considered during this
evaluation include local hydrogeology, well construction, and well location,
all of which can affect the time associated with obtaining a stable or equili-
brated well discharge.
The importance of testing wells individually prior to establishing sam-
pling protocols is apparent from the data presented in the following figures
and tables. These data were collected from three wells installed by the USGS
in the Piceance Basin. The typical well construction for these wells is shown
in Figure 23.
The objective of this sampling program was to evaluate the variability of
the water quality with pumping time. The types of data collected during this
survey included continuous pH, temperature, and specific conductance measure-
ments. Samples for detailed chemical analysis were also collected at selected
intervals.
As Figures 30 through 34 indicate, the conductivity can vary substan-
tially with pumping time. In the case of Well 75-1A (Figure 30), a very large
and rapid change in conductivity was initially observed. After approximately
one well volume, the conductivity stabilized and remained fairly constant for
the duration of the test.
2000 r
CUMULATIVE NUMBER OF WELL VOLUMES
1.0
1.91
I
USGS WELL 75-1A
800
I
I
10 20 30 40 SO 60 70
TIME SINCE START (mini
80
90
100
Figure 30. Variation in specific conductance with continued
pumping, USGS Well 75-1A, 1980.
108
-------
14,000
CUMULATIVE NUMBER OF WELL VOLUMES
0.5 1.0
COLORADO CORE HOLE NUMBER 3
2000
10 20 30
40 50 60 70
TIME SINCE START (min)
80
100
Figure 31. Variation in specific conductance with continued
pumping, US6S Colorado Core Hole #3, 1980.
70,000
60.000
50,000
t 40,000
30,000
20,000
10,000
CUMULATIVE NUMBER OF WELL VOLUMES
O.S
1.0
I
I
USGS WELL TH 75-18
I I I
I I I
10
TIME SINCE START (min)
100
140
Figure 32. Variation in specific conductance with continued
pumping, USGS Well TH75-1B, 1980.
109
-------
CUMULATIVE NUMBER OF WELL VOLUMES
0.5
1.0
I
I
USGSWELLTH75-1B
100
140
TIME SINCE START (min)
Figure 33. Variation of temperature of pumped discharge,
Well TH75-1B, 1980.
10.0
CUMULATIVE NUMBER OF WELL VOLUMES
1.0
1.01
9.0
pH
8.0
7.0
I
USGS WELL 75-1A
I
10 20 30 40 50 60
TIME SINCE START (min)
70
80
90
100
Figure 34. Variation in pH with continued pumping,
USGS Well 75-1A, 1980.
The conductivity data collected for Colorado Core Hole #3 (Figure 31),
steadily declined throughout the entire test. Conductivity values obtained
toward the end of the test were about 20 percent of the initial measurements.
Although more than one well volume was discharged from this well, the test was
obviously not -long enough for obtaining an equilibrated discharge.
An increasing trend in conductivity was observed for Well TH75-1B (Figure
32). During the test performed on this well, the conductivity was fairly sta-
ble at around 30,000 umho/cm, until approximately three-quarters of a well
volume had been discharged. At this point, the conductivity increased
abruptly to around 58,000 ymho/cm, where it stabilized for the duration of
the test.
The other constituents measured in the field also changed during the
tests. The temperature of the discharge of Well TH75-1B (Figure 33) initially
declined steadily and then appeared to increase slightly. The pH of Well
110
-------
75-1A (Figure 34) initially increased one pH unit and stabilized after about
10 minutes of pumping.
These patterns of changes in the constituents measured in the field are
also reflected in the water chemistry analysis (Table 15). For instance, the
large change in conductivity for Well 75-1A is repeated for several major in-
organic ions (potassium, sodium, bicarbonate, chloride, and sulfate), alkalin-
ity, TDS, and fluoride concentrations. Most of the trace constituents (arse-
nic, boron, mercury, and selenium) were largely unchanged for the duration of
the pumping.
The data collected during this survey and presented above point out the
need for the individual testing of each well. It is obvious that a sample
collected during the first few minutes of pumping and before conductivity has
stabilized will not be representative. It is also obvious that the extraction
of one well volume previous to representative sample collection is not a com-
pletely accurate rule-of-thumb, since the data for Colorado Core Hole #3 never
stabilized, even after more than one and one-fourth volumes had been
extracted.
In regard to pump location, it is recommended that the pump intake be
placed approximately 5 feet above the open, perforated, or screened aquifer
interval. The rationale for placing the pump in this location is as follows:
t A structurally unstable aquifer interval could fail due to the
excessive stresses created by the pump if it were placed directly
opposite the open, perforated, or screened interval
If the well is not developed properly, the pump can produce suf-
ficient turbulence in the aquifer interval to produce sand, etc.
If the pump is placed in the aquifer interval and the discharge
is too high, excessive drawdown may create cascading conditions
that can produce sufficient turbulence to modify easily oxidized
constituents
t Humenick et al. (1980) have pointed out that this pump location
significantly reduces the volume of water necessary for extrac-
tion before representative aquifer water is obtained.
Figure 35 (from Humenick et al., 1980) illustrates two wells. Well A,
with the pump intake 5 feet above the open aquifer interval, requires 12 gal-
lons of discharge before formation water is produced. For Well B, the pump
intake is 35 feet above the open interval and requires 77 gallons of discharge
before representative formation water is produced.
In short, the following procedure defining sampling protocols is recom-
mended for collecting representative samples from a well when using a submers-
ible pump:
111
-------
TABLE 15. WATER CHEMISTRY OF SAMPLES COLLECTED AFTER DISCHARGE OF VARYING WELL VOLUMES,
USGS WELLS, PICEANCE BASIN, 1980
Well volumes discharged
Well 75-1A
Constituent3
Total dissolved solids
Calcium
Magnesium
Sodium
Potassium
Bicarbonate
Carbonate
Sulfate
Chloride
Fluoride
Ammonia
Arsenic
Boron
Mercury
Selenium
Dissolved organic carbon
Note:
a
Constituent units are mg/1
0
1,176
4.3
32
429
1.2
944
<1
184
85.7
0.6
0.4
<0.01
0.3
<1
<0.01
10.1
except
1
816
13.8
54.9
235
0.4
695
5
149
10.7
0.2
0.3
<0.01
0.3
<1
<0.01
13.1
2
836
11.7
52.5
225
0.4
708
<1
144
12.3
0.2
0.2
<0.01
0.4
<1
<0.01
34.5
for pH units and
Core Hole # 3
0
7,148
1.8
24.6
2,910
21.3
3,496
39.5
229
2,236
0.4
5.6
<0.01
0.5
<1
0.01
26.3
1
3,276
3.2
29.3
1,320
13.6
1,719
14.1
228
852
0.2
2.3
<0.01
0.4
<1
0.01
24.0
specific conductance
Well TH75-1B
0
22,880
3.6
4.4
10,200
22.5
23,300
32.5
131
1,540
17.3
8.0
0.02
0.8
<1
<0.01
23.4
in iimhos/cm
0.2
22,400
2.0
2.2
9,900
22.1
23,640
<1
87.2
1,670
17.2
8.3
0.02
0.8
<1
<0.01
26.9
at 25°C.
1
45,220
1.5
2.8
19,650
32.5
48,560
<1
61.7
3,730
18.4
8.2
0.01
0.9
<1
<0.01
24.4
-------
12 gallons
77 gallons
DRAWDOWN (3 feet)
-STATIC WATER LEVEL
DRAWDOWN (17 feet)
PUMP INTAKE 5 feet
ABOVE OPEN HOLE
CONFINING LAYER
AQUIFER
PUMP INTAKE 35 feet
ABOVE OPEN HOLE
VOLUME OF STATIC WELL BORE WATER
VOLUME OF STATIC WATER DISCHARGED
BEFORE FORMATION WATER IS PRODUCED
Figure 35. Comparison of pump locations and the volume
of water necessary for extraction before
representative aquifer water is obtained
(modified from Humenick et al., 1980).
113
-------
1. Compile well construction data, including well diameter, total
depth, and perforated interval or aquifer interval in an open
well.
2. Measure static water level and estimate well volume.
3. The pump intake should be placed approximately 5 feet above the
open, perforated, or screened aquifer interval.
4. The discharge rate should be maintained at a moderately low
rate to prevent excessive drawdown in the aquifer and well and
minimize turbulent mixing in the annulus.
5. Extract at least one well volume from the well.
6. Continuously monitor and measure specific conductance, pH, and
temperature in the field throughout the pumping period. Con-
tinuously monitoring these parameters is particularly important
for little-used groundwater quality monitoring wells.
7. Collect the sample only after the field parameters have stabi-
lized for a period of time. Although the data indicate that
the conductivity is the most conclusive of aquifer water, it is
suggested that all of the parameters be monitored to indicate
representative aquifer water to prevent premature sample
collection.
8. Collect the sample as close to the well head as possible to
avoid potential contamination, precipitation of solutes, and
the loss of dissolved gases.
It is recommended that these protocols be recharacterized periodically
for each well, particularly for wells with large, open intervals. Once these
protocols are defined or redefined and consistency among items such as dis-
charge rate, time of collection, and pump placement is established, represen-
tative samples can be collected. However, to produce comparable data for
establishing water quality trends, these procedures (i.e., pump placement,
discharge rate, etc.) should be followed during each sampling effort.
Swabbing
The swabbing method is utilized by Cathedral Bluffs Shale Oil Company
(operators of Tract C-b) for sampling deep aquifer wells. This methodology is
a common procedure used in oil field operations and has been adapted for use
on Tract C-b as a sampling procedure. The swabbing technique involves intro-
ducing a swabbing cup into the well, which is supported from the surface by a
pipe, and removing a portion of the water from the 2-5/8-inch-diameter well.
The water extracted from the well is discharged through a line to the place
where the water quality samples are collected. This sampling method requires
a capitalization cost (in 1980 dollars) of approximately $50,000 to $55,000
and requires four personnel for operation. Tract C-b contracts the equipment
114
-------
and personnel for each sampling effort, performed quarterly, at a cost of
about $20,000.
The sampling approach applied by Tract C-b personnel is to completely
evacuate a well volume of water. After this evacuation process has been per-
formed, the well is allowed to recover for at least 24 hours. After this
period, the swabbing equipment is returned to the site and the process is re-
peated and a sample is collected. Parameters measured in the field include
specific conductance, pH, and temperature.
The advantages of swabbing are as follows:
Swabbing can be used where the depth to water is relatively great
and well diameters are relatively small
At least one well volume can be obtained from the well, allowing
for representative sample collection.
The disadvantages of this method include:
Difficulty in regulating the volumes of water obtained from the
well and the discharge rates
t Well contamination can occur when oil-field equipment is used for
deep aquifer sampling; also, there is a potential for cross-con-
taminating the samples
Very difficult to employ
Accelerated plugging of the piezometer perforations is a poten-
tial problem, particularly with the small diameter of the
piezometers
Consistent water quality sample collection is difficult to
achieve due to the vertical mixing of the well water upon extrac-
tion of the water; consistent swabbing depth during each sampling
effort would help alleviate this problem.
In general, the use of swabbing is not recommended as a sampling technique.
SAMPLING FREQUENCY
Defining an appropriate sampling frequency is a complex issue influenced
by location of sampling sites, monitoring goals, climatological factors, and
characteristics of groundwater flow. As a result, sampling frequency should
be defined on a case-by-case and likely trial-and-error basis. One of the key
factors is groundwater flow rate. If flow from a potential pollution source
to a monitoring well is expected to be on the order of decades (assuming a re-
lease occurs), then very frequent sampling does not seem warranted and perhaps
annual sampling for a few indicator constituents would suffice.
115
-------
The complexity of the hydrogeology of the oil shale region makes estima-
tion of groundwater flow rate difficult at best and the actual flow rates
highly site specific. Table 16 lists some estimates of travel time in the Up-
per Aquifer zone of the Piceance Creek Basin. The wide variation in results
reinforces the care needed in design of monitoring programs, as our under-
standing of the system is incomplete.
TABLE 16. FLOW RATES OF THE UPPER AQUIFER, PICEANCE CREEK BASIN,
ESTIMATED BY THREE STUDIES
Travel time
Flow velocity (years to
Study reference (feet per day) travel 1 mile)
Lawrence Berkeley Labs, 1978
(data from Weeks et al., 1974)
U.S. Atomic Energy Commission,
1972
Knutson, 1973
0.05
0.36 - 0.78a
11.7
300
20 - 40
1.2
Note:
aRange for representative gradient and maximum gradient cases.
SAMPLE HANDLING AND PRESERVATION
Proper methods of sample handling and preservation should be exercised to
minimize the changes in the geochemical environment from which the sample is
extracted. The chemical qualities of some samples can change within a few
hours or minutes following withdrawal. Other constituents can be preserved
and stabilized for a limited period of time, whereas still other constituents
have a shelf life of up to 6 months. In addition to sample time constraints,
the sampler should be aware of potential problems that may arise from improper
selection of sample volumes, containers, and preservatives, as well as inade-
quate field records and chain-of-custody preparations.
Field Data Collection
Parameters that should be measured in the field include pH, temperature,
and specific conductance. If dissolved oxygen and oxidation-reduction poten-
tial measurements are required, these should also be determined in the field.
Although some of these parameters have holding times of up to 24 hours (see
Table 17), it is recommended that these determinations be made in the field
with the appropriate apparatus to prevent inaccurate results from delay in re-
ceipt of the samples at the analytical laboratory. In addition to holding
times Table 17, derived from U.S. EPA (1974), contains information regarding
the recommended choice of preservatives and sample containers and volume re-
quirements for various constituents.
116
-------
TABLE 17. RECOMMENDATION FOR SAMPLING AND PRESERVATION OF SAMPLES
ACCORDING TO MEASUREMENT3
Parameter
Measured
Acidity
Alkalinity
Arsenic
BOO
Bromi de
Chloride
Chlorine
COD
Color
Cyanides
Dissolved Oxygen
Probe
Winkler
Fluoride
Hardness
Iodide
MBAS
Metals
Dissolved
Suspended
Total
Mercury
Dissolved
Total
Nitrogen
Ammon i a
Kjeldahl, total
Nitrate
Nitrite
NTA
Oil and grease
Organic carbon
PH
Volume
Required
(ml)
100
100
100
1,000
100
50
200
50
50
500
300
300
300
100
100
250
200
200
100
100
100
400
500
100
50
50
1,000
25
25
Container0
P,G
P,G
P,G
P,G
P,G
P,G
P,G
P,G
P,G
P,G
6 only
G only
P,G
P,G
P,6
P,G
P,G
P,G
P,G
P,G
P,G
P,G
P,G
P,G
P,G
G only
P,G
P,G
Preservative
None required
Cool, 4°C
HN03 to pH < 2
Cool, 4°C
Cool, 4°C
None required
Determine on site
H2S04 to pH < 2
Cool, 4°C
Cool, 4°; NaOH to pH 12
Determine on site
Fix on site
None required
Cool, 4°C; HN03 to pH < 2
Cool, 4°C
Cool, 4°C
Filter on site;
HN03 to pH < 2
Filter on site
HN03 to pH < 2
Filter; HN03 to pH < 2
HN03 to pH < 2
Cool, 4°C; H2S04 to pH < 2
Cool, 4°C; H,,S04 to pH < 2
Cool, 4°
Cool, 4°C
Cool, 4°C
Cool, 4°C;
HC1 or H2S04 to pH < 2
Cool, 4°C;
HC1 or H2S04 to pH < 2
Determine on site
Holding Timec
24 hours
24 hours
6 months
6 hours
24 hours
7 days
No holding
7 days
24 hours
None
4 to 8 hours
7 days
6 monthsd
24 hours
24 hours
6 monthsd
6 months
6 months
38 days (glass);
13 days (hard
plastic)
38 days (glass);
13 days (hard
plastic)"3
24 hours
24 hours6
24 hours
48 hours
24 hours
24 hours
24 hours
6 hours
(continued)
117
-------
TABLE 17 (continued)
Constituent
Volume
Required
(ml) Container'3
Preservative
Holding Timec
Phenolics
Phosphorus
500 G only Cool, 4°C; H3P04 to pH < 4; 24 hours
1.0 g CuS04/l
Orthophosphate,
dissolved
Hydro lyz able
Total
Total, dissolved
Residue
Filterable
Nonf ilterable
Total
Volatile
Settleable matter
Silica
Specific conductance
Sulfate
Sulfide
Sulfite
Temperature
Turbidity
50
50
50
50
100
100
100
100
1,000
50
100
50
50
50
1,000
100
P,G
P,G
P.G
P.G
P,G
P,G
P,G
P.G
P.G
P only
P,G
P.fi
P.G
P,G
P.G
P,G
Filter on site; cool 4°C
Cool, 4°C; H2S04 to pH < 2
Cool, 4°C; H2S04 to pH < 2
Filter on site; cool, 4°C
H2S04 to pH < 2
Cool, 4°C
Cool, 4°C
Cool, 4°C
Cool, 4°C
None required
Cool, 4°C
Cool, 4°C
Cool, 4°C
2 ml zinc acetate
Determine on site
Determine on site
Cool, 4QC
24 hours
24 hours6
24 hours6
24 hours6
7 days
7 days
7 days
7 days
24 hours
7 days
24 hoursf
7 days
24 hours
No holding
No holding
7 days
Notes:
aMore specific instructions for preservation and sampling are found with each procedure as
detailed in U.S. EPA (1974). A general discussion on sampling water and industrial wastewater
may be found in ASTM, Part 31, p. 72-82 (1976), Method D-3370.
bPlastic (P) or glass(G); for metals polyethylene with a polypropylene cap (no liner) is
preferred.
clt should be pointed out that holding times listed above are recommended for properly pre-
served samples based on currently available data. It is recognized that for some sample
types, extension of these times may be possible, while for other types, these times may be too
long. Where shipping regulations prevent the use of the proper preservation technique or the
holding time is exceeded, such as the case of a 24-hour composite, the final reported data for
these samples should indicate the specific variance.
^Where HN03 cannot be used because of shipping restrictions, the sample may be initially pre-
served by icing and immediately shipped to the laboratory. Upon receipt in the laboratory,
the sample must be acidified to a pH < 2 with HN03 (normally 3 ml 1:1 HN03/1 is sufficient).
At the time of analysis, the sample container should be thoroughly rinsed with 1:1 HN03 and
the washings added to the sample (volume correction may be required).
eOata obtained from National Enforcement Investigations Center, Denver, Colorado, support a
4-week holding time for this parameter in sewerage systems (SIC 4952).
If the sample is stabilized by cooling, it should be warmed to 25°C for reading, or tempera-
ture correction made and results reported at 25°C.
118
-------
Today, field studies are supported by some fairly precise, portable, ana-
lytical equipment that furnish accurate data, thus eliminating any effects
delayed sample shipment may have on the validity of water chemistry analysis
performed at the laboratory. Furthermore, these determinations can be easily
measured and provide valuable on-site information regarding aquifer character-
istics of aid in the collection of representative aquifer water.
Specific Conductance-
Specific conductance is a measure of the ability of a solution to trans-
mit an electrical current. In water samples, the specific conductance is an
indication of the concentration of dissolved solids (i.e., salinity). The
unit of measurement for specific conductance is the inverse of the resistivity
and is typically expressed in micromhos per centimeter.
Specific conductance is an important measurement that should always be
made in the field during sample collection. This parameter is very useful in
determining when aquifer water has been obtained and thereby aids in the col-
lection of representative samples. The recommended holding time is only 24
hours (U.S. EPA, 1979), which may present problems in obtaining accurate re-
sults from the analytical laboratory if sample shipment is delayed.
Temperature-
Temperature should always be measured immediately after sample with-
drawal. It is a very easy measurement to obtain, and the equipment used for
its determination should be accurate to within ±0.1°C to allow for future geo-
chemical evaluations of equilibrium thermodynamics. Also, field determina-
tions of aquifer water temperatures prevent inaccurate measurements due to the
modification of the sample temperature during sample preservation and
transportation.
PH-
Wood (1976) provides the following description for pH: "The pH of a so-
lution is a measure of effective hydrogen-ion concentration or, more accu-
rately, it is the negative logarithm of the hydrogen-ion activity in moles per
litre: pH = -log (H*)." The pH of an aqueous solution is typically con-
trolled by the disassociation of acids, bases, and hydrolysis. The pH of a
groundwater sample is further controlled by the carbonate system, including
dissolved carbon dioxide, bicarbonate, and carbonate ions.
The pH of an aqueous solution can be measured precisely and quickly with
mechanical instruments. Some researchers have found that pH is the best pa-
rameter for determining that representative aquifer water has been obtained
from a well (Brown et al., 1970; Wood, 1976; and Humenick et al., 1980). The
holding time for pH is 7 hours (U.S. EPA, 1979), however, which can affect
analytical determinations if sample shipment is delayed. Therefore, it is
highly recommended that pH be measured in the field at the time of sample
withdrawal.
119
-------
Dissolved Oxygen and Oxidation-Reduction Potential--
Geochemical evaluations of a groundwater system may require dissolved
oxygen and oxidation-reduction potential measurements. If this is the case,
these measurements should be conducted in the field at the time of sample col-
lection for accurate results. Particular care must be exercised during these
measurements to prevent atmospheric aeration of the sample during collection
and analysis. Many companies produce precise, portable, easily used, analyti-
cal equipment for these measurements.
Field Notes and Records. Sample Labels
The following notes and records for sample collection should be main-
tained for future data evaluation:
Time and date of arrival, sample collection, and departure from
the well site
The water level of the well
Description of the sample source, including well number and loca-
tion and the following additional information (if applicable):
Depth of sample collection (of critical importance for
bailing)
Duration of pumping previous to sample collection
Well volumes extracted previous to sample collection
-- Pump placement
Well data information pertaining to well construction and comple-
tion and the aquifer(s) or section of aquifer in which the well
is completed, including:
Length and depth to screened interval, open interval, and/or
casing interval
-- Well annulus
Total depth of well
Water quality data for specific conductance, pH, temperature,
water level, etc.
Sampling specifications, particularly the procedures previously
employed for sample collection, that wiVI establish consistent
sampling methods for each sampling effort, including:
« Pump placement
120
-------
Discharge rate
-- Time of sample collection, etc.
The type of sampling methodology utilized for sample withdrawal
Field observations pertinent to sample collection, including
color, sediment, turbidity, etc.
The reason for the sampling effort
The results of field determinations performed at the time of sam-
ple collection (e.g., temperature, specific conductance, pH, oxi-
dation-reduction potential, dissolved oxygen, etc.)
Any problems encountered in the field during sample collection.
The identity of the sample collector.
Sample labels should be prepared before the sampling effort and affixed to the
sample container. If possible, the information should be duplicated on the
sample container itself to prevent errors resulting from label detachment dur-
ing sample handling and shipment. In addition, waterproof pens should be used
by the sampler to prevent dissipation. The following information should be
included on the sample label and container:
t Time and date of sample collection (if multiple samples are to be
collected from the same well, the hierarchy or succession of the
samples collected should also be noted)
The well number and location
The preservative (if any) utilized
If the sample has been filtered in the field or been sent to the
analytical laboratory unfiltered.
Field Handling and Preservation Techniques
Preservation of samples through the use of techniques currently available
and easily applied in the field can only retard the chemical or biological
changes that take place after the sample has been withdrawn from the well.
Methods of preservation are relatively limited and are intended to: (1) re-
tard biological activity, (2) retard hydrolysis of chemical compounds and com-
plexes, (3) reduce volatility of constituents, and (4) reduce absorption
effects (U.S. EPA, 1979). In general, preservation techniques include pH con-
trol, chemical addition, refrigeration, and freezing. The following preserva-
tives are used to retard.sample changes after collection (U.S. EPA, 1977):
121
-------
Preservative
HgCl2
Action
Acid (H2S04)
Alkali (NaOH)
Refrigeration
Bacterial inhibitor
Applicable To
Metals solvent, prevents
precipitation
Acid (HN03)
Acid (H2S04) Bacterial inhibitor
Salt formation with
organic bases
Salt formation with vola-
tile compounds
Bacterial inhibitor
Nitrogen forms, phosphorus
forms
Metals
Organic samples (COO, oil
and grease, organic carbon)
Ammonia, amines
Cyanides, organic acids
Acidity-alkalinity,
organic materials, BOD,
color, odor, organic P,
organic N, carbon, etc.,
biological organisms
(coliform, etc.)
Containers used for sample collection should be selected for their non-
reactivity with the particular analytical parameter to be measured. Depending
on the constituent(s), the containers typically consist of either glass or
plastic. Table 17 provides the recommended sample container for the particu-
lar analysis of interest. In addition, it is generally advantageous to pre-
pare the sample containers with the appropriate preservative prior to sample
collection. This procedure may be very time-consuming since separate bottles
and chemical preservatives are required for certain parameters, which may re-
sult in several containers for each sample collected. However, this prepara-
tion will result in the elimination of laborious effort in the field during
sample collection.
Determination of dissolved concentrations will require the sample to be
filtered through a 0.45-micron filter prior to acidification. If the sample
is not filtered and acid is added as a preservative, much of the particulate
matter will be dissolved by the acid resulting in anomalously high concentra-
tions of dissolved constituents. It is also recommended that the sample be
filtered as soon as possible after withdrawal, preferably in the field. How-
ever, samples to be used for on-site temperature, dissolved oxygen, pH, and/or
oxidation-reduction potential measurements should not be filtered before these
determinations have been made.
After the sample has been filtered and preserved, the recommended proce-
dure is to place the samples on ice for further preservation during shipment.
The use of an ice chest is the preferred approach for sample shipment since
ice chests are easy to handle and are insulated such that a temperature be-
tween 0° and 10°C can be maintained for a limited period of time.
122
-------
Sample Shipment
The relative remoteness of the oil shale region can result in some delay
in receipt of the groundwater quality samples at commercial analytical labora-
tories. To evaluate the effects of this potential problem on the results of
chemical analysis, a testing program was initiated by Tempo. The program was
conducted in conjunction with Cathedral Bluffs Shale Oil Company personnel on
Federal Lease Tract C-b. The field effort involved the sampling of three al-
luvial wells with a portable, submersible pump. At each well, the samples
were collected after the field parameters (i.e., specific conductance, pH, and
temperature) had stabilized. The sample collected from each well was handled
as follows:
The sample was split three ways and preserved with EPA-recom-
mended preservatives (see Table 17)
Samples were refrigerated or cooled as recommended by EPA (see
Table 17)
The samples were then shipped to the analytical laboratory for
analysis in the following sequence:
First sample split was analyzed within 24 hours of sample
collection
Second sample split was analyzed 7 days after sample
collection
Third sample split was analyzed 15 days after sample
collection.
This sequence of sample analysis was intended to simulate circumstances
that can arise during field sampling efforts. The initial split represents
the optimum situation for sample shipment, i.e., immediately after sample col-
lection. The second sample split represents the situation where samples col-
lected during the week are shipped to the laboratory for analysis at the end
of the week. The third sample split represents either a lengthy field survey
resulting in a shipment of samples at the end of two working weeks or a sig-
nificant delay in the receipt of the samples at the laboratory due to shipping
problems.
The chemical analyses presented in Tables 18 through 20 represent samples
collected from alluvial wells A-6, A-9, and A-12, respectively. The constitu-
ents of the chemical analyses consisted of specific conductance, pH, total
dissolved solids (TDS), bicarbonate, carbonate, chloride, ammonia, sulfate,
nitrate, and dissolved organic carbon (DOC). The constituents chosen were
based on the following EPA holding-time recommendations:
123
-------
TABLE 18. CHEMICAL ANALYSIS OF SAMPLES TAKEN FROM ALLUVIAL WELL A-6 FOR THREE DIFFERENT
TIMES OF ANALYSIS
Within 24 hours
Constituent Repl 1 Repl 2 Repl 3
Specific
Conductance* 1.270 1,260 1,280 1
pHb 7.8 7.75 7.8
Bicarbonate 712 706 710
Carbonate <1 <1 <1
Sulfate 189 208 195
Chloride 16.2 15.8 15.8
TDS 888 880 884
Ammonia O.I O.I O.I
Nitrate 0.20 0.20 0.20
Fluoride 0.33 0.34 0.33
DOC 16 14 14
Notes:
Measured 1,350 in the field, units umhos/cm
Measured 7.5 in the field, pH units.
7 days
Mean
.270
7.8
709
<1
197
15.9
884
O.I
0.20
0.33
15
at 25°C.
Std.
Dev.
8.2
0.02
2.5
0
7.9
0.2
3.3
0
0
0
1
Repl Replicate Sampling; Std. Dev. Standard Deviation.
Repl 1
1,220
7.70
706
<1
249
12.4
918
O.I
0.20
0.40
46
Repl 2
1,210
7.80
702
<1
233
12.9
912
0.1
0.20
0.40
52
Constituent units
Repl 3
1,205
7.70
699
<1
243
12.8
920
O.I
0.30
0.41
49
mg/1.
Mean
1,212
7.73
702
<1
242
12.7
917
0.1
0.23
0.40
49
Std.
Dev.
6.2
0.05
2.9
0
6.6
0.22
3.4
0
0.5
0
2.4
Repl 1
1,250
7.70
693
<1
234
12.8
928
O.I
0.20
0.39
63
Repl 2
1,260
7.80
699
<1
242
12.6
920
0.1
0.20
0.40
61
15 days
Repl 3
1,255
7.79
699
<1
240
12.9
920
0.1
0.20
0.40
56
Mean
1,255
7.76
697
<1
239
12.9
923
O.I
0.20
0.40
60
Std.
Dev.
4.1
0.05
2.8
0
3.4
0.1
3.8
0
0
0
2.9
-------
TABLE 19. CHEMICAL ANALYSIS OF SAMPLES TAKEN FROM ALLUVIAL WELL A-9 FOR THREE DIFFERENT
TtMES OF ANALYSIS
Within 24 hours
Constituent Repl 1 Rep] Z
Specific
Conductance9 1,110 1,100
pHb 7.97 7.99
Bicarbonate 482 486
Carbonate <1 <1
,_. Sulfate 299 296
£> Chloride 7.22 7.59
IDS 784 784
Ammonia <0.1 <0.1
Nitrate 1.0 0.9
Fluoride 0.20 0.18
DOC 87
Notes:
a
Specific conductance measured 1,
pH measured 7.6 in the field, pH
Repl -- Replicate Sampling; Std.
Repl
1,095
7.
486
<1
305
7.
776
<0.
1.
0.
10
170 in
units.
Dev.
Std.
3 Mean Oev .
1,102
85 7.
485
<1
300
6 7.
781
1 <0.
0 0.
19 0.
8
the field
Standard
6.2
94 0.06
1.9
0
3.7
47 0.26
3.8
1 0
97 0.04
19 0
1.3
Repl 1
1,050
7.93
489
<1
315
4.90
796
<0.1
1.0
0.19
34
, units pmhos/cm at
Deviation.
7
Repl 2
1,055
7.90
493
<1
322
5.32
800
<0.1
0.9
0.21
42
25°C.
Constituent units
days
Repl 3
1,055
7.82
492
<1
320
5.30
804
<0.1
1.0
0.20
37
mg/1.
15 days
Mean
1,053
7.88
491
<1
319
5.17
800
<0.1
0.97
0.20
38
Std.
Dev.
2.4
0.05
1.7
0
2.9
0.19
3.3
0
0.04
0
3.3
Repl 1
1,085
8.05
492
<1
355
5.20
812
<0.1
1.0
0.19
45
Repl 2
1,110
8.05
489
<1
325
5.65
804
<0.1
0.9
0.20
51
Repl 3
1,115
7.98
492
<1
329
5.26
816
<0.1
1.0
0.20
50
Mean
1,103
8.03
491
<1
330
5.37
811
<0.1
0.97
0.20
49
Std.
Dev.
13.1
0.04
1.4
0
4.1
0.17
5.0
0
0.04
0
2.6
-------
TABLE 20. CHEMICAL ANALYSIS OF SAMPLES TAKEN FROM ALLUVIAL WELL A-12 FOR THREE DIFFERENT
TIMES OF ANALYSIS
CD
Within 24 hours
Constituent Repl 1 Repl 2 Repl
Specific
Conductance* 1,410 1,400 1,400
pHb 7.8 7.85 7.
Bicarbonate 598 596 596
Carbonate <1 <1 <1
Sulfate 398 418 402
Chloride 9.71 10.10 9.
TDS 1,052 1,040 1,056
Ammonia <0.1 <0.1 <0.
Nitrate 0.30 0.20 0.
Fluoride 0.20 0.18 0.
DOC 15 15 19
Notes:
a
Specific conductance measured 1,350 in
bpH measured 7.5 in the field, pH units.
Repl Replicate Sampling; Std. Dev.
3 Mean
1,403
8 7.82
597
<1
406
31 9.71
1,049
1 <0.1
20 0.23
18 0.19
16
the field,
Std.
Dev.
4.7
0.02
1.0
0
8.6
0.32
6.8
0
0.05
0.01
1.9
Repl 1
1,385
7.81
596
<1
469
7.22
1,116
<0.1
0.20
0.23
37
units umhos/cm at
Standard Deviation.
7
Repl 2
1,390
7.86
606
<1
447
7.31
1,116
<0.1
0.20
0.20
35
25°C.
Constituent units
days
Repl 3
1,375
7.85
602
<1
450
8.17
1,100
<0.1
0.30
0.19
34
mg/1.
15 days
Mean
1,380
7.84
601
<1
455
7.57
1,111
<0.1
0.23
0.21
35
Std.
Dev.
4.1
0.02
4.1
0
9.7
0.43
7.5
0
0.05
0.02
1.3
Repl 1
1,405
7.83
609
<1
472
7.56
1,136
<0.1
0.20
0.20
57
Repl 2
1,395
7.84
596
<1
465
7.56
1,156
<0.1
0.20
0.20
56
Repl 3
1,400
7.85
602
<1
451
7.47
1,056
<0.1
0.20
0.20
55
Mean
1,400
7.84
602
<1
462
7.53
1,116
<0.1
0.20
0.20
56
Std.
Dev.
4.1
0
5.3
0
8.9
0.04
43.2
0
0
0
0.8
-------
EPA Recommended
Constituent Holding Time
Specific conductance 24 hours
pH 6 hours
Total dissolved solids 6 months
Bicarbonate 24 hours
Carbonate 24 hours
Chloride 7 days
Fluoride 7 days
Ammonia 24 hours
Nitrate 24 hours
Sulfate 7 days
Dissolved organic carbon 24 hours
Trace metals 6 months
Due to the long shelf life of the trace metals (i.e., 6 months), these con-
stituents were not incorporated in the chemical analysis.
The constituents that display changes in concentration during the holding
periods are specific conductance, pH, sulfate, chloride, and OOC. TDS concen-
trations also vary somewhat between holding periods. The other constituents
either had concentrations below detection limits (e.g., ammonia and carbonate)
or maintained fairly uniform concentrations throughout the entire 15-day pe-
riod (e.g., bicarbonate, nitrate, and fluoride).
In every sample, the specific conductance and pH data differ somewhat be-
tween the field and the analyses performed after 24 hours. After this period,
the next two analyses indicate that the pH remains fairly constant. The spe-
cific conductance does vary slightly during the latter two analyses. The TDS
data do not reflect the trend observed for specific conductance but instead
generally increase slightly during the 15-day period.
The sulfate and chloride concentrations were also variable, particularly
between the 24-hour analysis and the 7-day analysis. In general, sulfate in-
creased in concentration during the holding periods, whereas the chloride con-
centration decreased during the same interim. In some instances, such as the
decrease in chloride concentration during the initial 7-day period for the
sample collected from Well A-6, the changes were considerable.
The constituent that displayed the most appreciable variability was DOC.
In all three well samples, the DOC concentration increased significantly over
the 15-day period. However, the polyethylene sample containers may have con-
tributed somewhat to this trend. It has been demonstrated that polyethylene
can contribute contaminating organics to the sample and affect the DOC
concentrations.
127
-------
Summers (1972) has demonstrated that changes in pH, specific conductiv-
ity, and the carbonate-bicarbonate system are all indicative of sample aging.
Typically, the following reaction will control these changes: CaCOs + h^O +
C02 -» Ca£ + 2HC03. Changes in the C02 concentration can also influence these
constituents. The data collected by this survey do not reflect appreciable
increases in the bicarbonate concentration during the 15-day period. However,
in that the greatest increase in pH occurred during the first 24 hours, per-
haps the samples achieved equilibrium before the first analysis by the labora-
tory. If such were the case, a significant increase in the bicarbonate would
require immediate analysis for detection and probably could not be observed
after a period of time, in this case 24 hours.
The data also indicate that chloride and sulfate may be the most sensi-
tive parameters with respect to sample holding-time considerations. Although
the EPA-recommended holding time for these constituents is 7 days, it is ap-
parent from the data that the most significant changes in the concentrations
of these constituents occurred during the first 7 days. Due to the potential
contamination of the DOC by the polyethylene sample container, the effects of
the holding times on this constituent are inconclusive.
It is apparent that correct and quick sample shipment is critical for ac-
curate analytical results. In the oil shale regions, time constraints may
prevent field personnel from delivering the samples to the laboratory, partic-
ularly if the sampling effort extends beyond 24 hours. In this case, trie most
efficient procedure is to ship the samples via commercial bus or plane. This
procedure is very inexpensive (on the order of $2 to $10 per ice chest) and
will eliminate unnecessary trips to the analytical laboratory by field person-
nel. Furthermore, transportation by these methods is very reliable and pro-
vides reasonable assurance against changes in sample chemistry due to pro-
longed sample storage.
Chain of Custody
The typical chain of custody in the oil shale region includes the sam-
pler, the individuals involved in the transportation, and the individuals
handling the sample at the analytical laboratory. The proper procedures that
should be followed during this chain of custody are:
Include as few people as possible in the chain of custody.
Collect, preserve, and ice the samples according to the recom-
mended procedures
Label each sample container according to the recommendations pre-
viously presented indicating the analysis required on the sample
labels if preparations have not previously been made
Maintain a field notebook or logbook during each survey and store
it in a safe place, with all entries signed by the individual
responsible for the field effort
128
-------
Assign complete responsibility for the collected samples (includ-
ing those delivered by field personnel) to the individual con-
ducting the effort, including overseeing all transportation
activities (including timely delivery of the samples to the bus-
line or airline facility and their receipt at the analytical lab-
oratory) and maintaining a record of these activities as follows:
time and date of deliveries, method of transportation, and the
individual(s) performing the transportation.
Furthermore, effort should be made to have the laboratory performing the chem-
ical analysis retain a custodian to maintain a record indicating:
Time and date of sample receipt
The person receiving the sample
t The sample number
The number assigned to each sample by the laboratory.
This custodian should provide for proper handling and storage of the samples
prior to analysis. In addition, the custodian should be responsible for dis-
tribution of the samples to the individual performing the analysis, recording
the individual's identity, and assuring that immediate analysis is conducted
to avoid water chemistry changes due to prolonged sample storage.
SELECTION OF CONSTITUENTS FOR MONITORING
The proper location of monitoring points is largely determined by the lo-
cale and character of the potential sources of groundwater quality impact and
the local source hydrogeology. The constituents for monitoring are selected
so as to provide a cost-effective indication of the nature and extent of im-
pact on groundwater quality. Assessment of enrichment factors (or concentra-
tion change above ambient), specific indicator constituents, and stable
isotopes are possible approaches for selection of constituents for chemical
analysis.
Enrichment Factors
In this subsection, enrichment factors, EF, will be calculated for major
possible sources of groundwater impact according to the expression:
EF - concentration from potential pollution source
concentration in aquifer*
For this assessment, representative baseline water quality levels were se-
lected (Table 21). Concentrations from the more saline sections of the Lower
Aquifer are included principally in Table 21 in order to most clearly demar-
cate the differences in enrichment factors. Representative concentrations of
constituents in retort water and in spent shale leachate were used in this
preliminary analysis.
129
-------
TABLE 21. REPRESENTATIVE CONCENTRATIONS IN GROUNDWATERS
ADAPTED FOR THIS STUDY
Gross Parameters (mg/1)
Conductance (unho/cm)
pH
IDS
Ammonia
Bicarbonate
Calcium
Carbonate
Chloride
Cyan i de
Magnesium
Nitrate
Potassium
Silica
Sodium
Strontium
Sulfate
Phosphate
Kjeldahl nitrogen
Nitrite
Sulfide
Minor and Trace Elements
Aluminum
Arsenic
Barium
Beryllium
Boron
Bromine
Springs,
seeps and
alluvial
aquifer
1,300
5 - 8
900
0.4
500
70
3
10
0.01
70
2
2
20
150
2
350
< 0.1
2
0.2
0.2
(vg/D
300
5
50
<100
500
20
Upper
Aquifer
1,500
7 - 8.5
1,000
0.5
500
50
^
10
0.01
70
1.0
2
20
200
2
350
< 0.1
0.6
200
10
100
<10
1,000
50
Saline
Lower
Aquifer
7,000
8
6,000
10
4,000
200
20
20
0.01
20
0.5
20
10
2,500
60
< 0.1
0.6
250
10
800
40,000
500
Lower
working limit
of detection,
Denver
Laboratory3
2
1
100
5
0.05
5
0.2 - 1
0.002
50
0.02
0.1
1
0.1
0.01
3 - 10
0.1
0.1
0.02
0.1
100
2-50
50
5
50
2,000
(continued)
130
-------
TABLE 21 (continued)
Springs,
seeps and
alluvial
aquifer
Upper
Aquifer
Saline
Lower
Aquifer
Lower
working limit
of detection,
Denver
Laboratory3
Minor and Trace Elements (ug/1) (continued)
Cadmium
Chromium
Cobalt
Copper
Fluoride
Iron
Lead
Manganese
Mercury
Molybdenum
Nickel
Radiation, alpha (pCi/1)
Radiation, beta (pCi/1)
Rubidium
Scandium
Selenium
Silver
Thallium
Titanium
Uranium
Vanadium
Zinc
Lithium
Gross Organic Parameters
TOC (mg/1)
Phenol (pg/1)
DOC (mg/1)
COD (mg/1)
Note:
aUsing standard methods.
17
11
8
30
400
500
50
30 - 500
0.4 - 3
40
30
5
4
10
4
<10
<1
200
5
200
<100
5
3
5
16
10
2 - 300
3
70
7,000
500 - 5,000
10 - 100
100
0.4 - 3
50
20
5
4
20
<10
10
100
<30
2
200
3
3
8
18
5
10
5
70
20,000
800
100
100
0.4 - 2
50
10
20
20
70
...
<10
10
100
<20
16
200
10
1 - 10
20
13
2
5 - 10
10
10
100
10
1 - 10
5
0.02
5
10
_._
5
0.05 - 10
5 - 50
300
5
5
5
1
1
...
10
131
-------
Also shown in Table 21 are the lowest concentrations typically reported
by a Denver water quality laboratory employing standard methods. As can be
seen, the average concentrations of P, V, Ti, As, Se, Ni, Co, Cu, Cd, Br, Be,
Ba, and As are close to or below these lower limits. It is therefore likely
that many of these trace element species were determined by spark source mass
spectroscopy which resulted in the improvement in detectability. This, how-
ever, resulted in a degradation in precision in comparison to standard
methods.
Table 22, which lists enrichment factors for the Lower Aquifer, is perti-
nent to the contamination of the Upper Aquifer, springs, and seeps by the
Lower Aquifer. As can be seen, Nfy, K, Na, B, and Br are enriched at least 10
times in the Lower Aquifer compared to either the Upper Aquifer or spring wa-
ters. In addition, Ba and F are enriched in the Lower Aquifer compared to
spring waters. It is likely, therefore, that these species would be indica-
tors of intrusion of waters from the Lower Aquifer.
TABLE 22. SPECIES ENRICHED IN THE LOWER AQUIFER
Enrichment factors
Lower Aquifer
Lower Aquifer
Upper Aquifer Springs and seeps
Conductance
TDS
Ammonia
Bicarbonate
Calcium
Potassium
Sodium
Barium
Boron
Bromi ne
Fluoride
Phenolics
4.6
6
20
8
4
10
13
8
40
10
2.9
0.3 - 10
5.4
6.6
25
8
4
10
17
16
80
25
50
0.3 -40
Table 23 lists enrichment factors for leachates and shows that the param-
eters pH, TDS, Cl, Na, $04, Mo, Se, and TOC are likely indicators (i.e.,
tracers) of contamination in the Lower Aquifer. Although carbonate appears to
be enriched in leachate, this reflects an increase in pH rather than an in-
crease in total HC03 + 0)3.
132
-------
TABLE 23. ENRICHMENT FACTORS ESTIMATED FOR SPENT MIS
OIL SHALE LEACHATE
Enrichment factors
Leachate
Upper Aquifer
Leachate
Lower Aquifer
Gross Parameters ;'mg/1)
Conductance (umhos/cm)
PH
TDS
Bicarbonate
Calcium
Carbonate
Chlor-ida
Cyanide
Magnesium
Nitrate
Potassium
Silica
Sod i urn
Strontium
Sulfate
Kjeldahl nitrogen
Sulfide
Minor and Trace Elements
Aluminum
Arsenic
Barium
Beryllium
(ug/i)
2.5 - 52
0.6 - 2
6 - 140
0.2 - 0.4
0.2 - 60
330 - 1,000
5.4 - 310
0.01 - 67
2.5 - 70
0.5 - 1.0
0.37 - 180
0.51 - 260
1.6 - 3.3
0.20 - 20
0.6 - 1.0
0.54 - 11
0.6 - 1.5
1-23
0.05 - 16
50 - 150
3.2 - 160
0.05 - 235
0.25 - 7
1.0 - 2.0
0.03 - 14
3 - 1,500
1.6 - 3.3
0.20 - 20
0.08 - 0.13
(continued)
133
-------
TABLE 23 (continued)
Enrichment factors
Leachate
Rubidium
Scandium
Selenium
Silver
Thallium
Titanium
Uranium
Vanadium
Zinc
Gross Organic Parameters
TOC
Phenolics
0.5 - 200
1.5 - 50
0.1 - 15
10 - 550
Leachate
Minor and Trace Elements
Boron
Bromine
Cadmium
Chromium
Cobalt
Copper
Fluoride
Iron
Lead
Manganese
Mercury
Molybdenum
Nickel
Upper Aquifer
(ug/1) (continued)
0.4 - 12
0.3 - 0.6
0.01 - 4
0.14 - 2.9
0.001 - 11
0.12 - 6
0.6 - 5
0.10 - 0.8
1.5 - 4,000
2.5 - 30
Lower
0.01
-
0.6
0.4
-
0.14
0
0.8
-
0.6
0.15
4
5
Aquifer
- 0.3
- 1.2
- 130
- 2.9
- 4
- 3.8
- 5
- 0.8
- 1,500
- 60
0.5 - 200
0.19 - 6.3
0.1 - 15
3 - 170
134
-------
The uncertainty in the enrichment factors reflects variations in the
original oil shale, methods of retorting, methods of analysis, and emphasizes
the necessity of preliminary controlled experiments prior to finalizing moni-
toring programs. Enrichment factors for those elements which are present in
concentrations near the detection limit, such as Se, would also be expected to
give variable enrichment factors.
Table 24 presents enrichment factors for retort waters and is relevant to
the extent that an in situ retort is not completely burned and retains a frac-
tion of the retort water. Most notably enriched in the retort water are Nfy,
COf, As, Br, Co, Hg, Se, V, U, and TOC, and possibly N0§, P0|-, and Ni.
The sulfur species shown at the bottom of the table will be discussed in the
next subsection.
Although COf is enriched in retort waters, it is unlikely that this
species would successfully pass through a spent retort because of the
reaction:
Ca2+ + 0)3 » CaC03
In fact, Parker et al. (1977) have shown that spent shale does, in fact, re-
move carbonate from surface waters. NH4, on the other hand, is likely a
highly mobile species, possibly after conversion to nitrate. In addition, the
more hydrophilic portions of the TOC may also travel with leachate and prove
indicative of groundwater contamination.
As in leachates from spent shale, retort waters appear enriched in those
species forming soluble an ions, such as As, Br, Se, and U. The origin of Co,
Hg, and V in the retort waters is less clear, although V is known to form or-
ganic complexes with organic compounds found in crude petroleum oils, and Hg
is known to vaporize from a simulated in situ retort and to recondense later
(Fox et al., 1978).
In summary, the water quality parameters pH, TDS, Cl, Na, S04, Mo, Se,
NH4, Br, Se, V, U, and TOC should be considered as potentially valuable in-
dicators of groundwater contamination, both for their elevated enrichment fac-
tors and for chemical reasons.
The utility of enrichment factor estimates is the identification of chem-
ical species likely to be detected in groundwater which indicate the impact of
a known source. To evaluate this possible monitoring approach, the enrichment
factors calculated above were categorized (arbitrarily) as follows:
Relative
likelihood
Table 25 Enrichment of detection
category factor range of impact
1 >500 High
2 50-500 Moderate
3 10-50 Low
135
-------
TABLE 24. ENRICHMENT FACTORS FOR RETORT WATERS
Enrichment factors
leachate
Gross Parameters (mg/1)
Conductance
Alkalinity
PH
TDS
Ammonium
Bicarbonate
Calcium
Carbonate
Chloride
Cyanide
Magnesium
Nitrate
Potassium
Silica
Sodium
Sulfate
Phosphate
Kjeldahl nitrogen
Sulfide
Minor and Trace Elements
Aluminum
Arsenic
Barium
Beryllium
Boron
Bromi ne
Cadmium
Chromium
Cobalt
Upper
10
19
1.0
i.a
3,400
34
0.01
170
0.002
40
0.001
0.17
1.5
0.02
0.001
0.06
0.8
1
0
(ug/i)
2.4
0.2
0.26
0.4
0.1
0.07
0.4
Aquifer
- 130
- 130
- 1.5
- 25
- 26,000
- 62
- 1.2
- 10,000
- 80
- 90
- 5
- 120
- 35
- 8
- 22
- 5.4
- 1,000
,700
.17
- 600
- 7
- 9
- 50
- 1.6
- 60
- 130
Leachate
Lower
2.1
2.4
1.0
0.3
170
4
0.002
25
0.001
40
0.01
0.34
0.15
0.04
j-0
0.3
0.8
-
0.
-
2.4
0.003
-
0.01
0.04
0.20
2.0
0.7
Aquifer
- 27
- 17
- 1.2
- 4
- 1,300
- 8
- 0.31
- 1,600
- 40
- 90
- 20
- 240
- 3.5
- 15
- 1.7
- 30
- 1,000
17
- 600
- 0.9
- 0.22
- 5
- 3.2
- 12
- 220
(continued)
136
-------
TABLE 24 (continued)
Enrichment factors
Minor and Trace Elements
Copper
Fluorine
Iron
Lead
Manganese
Mercury
Molybdenum
Nickel
Radiation, beta
Scandium
Selenium
Silver
Thallium
Titanium
Uranium
Vanadium
Zinc
Lithium
Organic Parameters
TOC
Unusual Sulfur Species
Total sulfur
Thiosulfate
Tetrathionate
Thiocyanate
Leachate
Upper Aquifer
(ug/1) (continued)
0.04 - 1.3
0.05 - 9
0.00001 - 15
0.05 - 10
0.23 - 1.4
3.3 - 1,000
2 - 11
3 - 50
9 - 35
___
>0.5 - >170
...
2-21
>0.33 - >150
2 - 5,500
0.20 - 25
10,000
6 - 20a
1,200 - 6,400b
400b
65 - 2,000°
Leachate
Lower Aquifer
0.04 - 1.3
0.02 - 3
0.001 - 100
0.4 - 10
0.23 - 1.4
5 - 1,000
2 - 11
6 - 100
1.8 - 7
>0.5 - >170
_
.
2-21
>0.50 - 230
0.25 - 700
0.20 - 25
3,000
35 - 120a
1,200 - 6,400b
400b
65 - 2,000b
Notes:
Calculated by assuming that all S in groundwaters is present
as sulfate.
Calculated by assuming background concentrations equal to a
detection limit of 0.5 mg/1.
137
-------
TABLE 25. RELATIVE LIKELIHOOD OF DETECTION OF MOBILITY FROM VARIOUS SOURCES TO UPPER AND
LOUER AQUIFERS AND SPRINGS BASED ON ESTIflATED ENRICHMENT FACTORS
00
Lower to Lower Aquifer
Constituent Upper Aquifer to springs
General water quality
measures
Conductivity
Total dissolved solids
Alkalinity
Major inorganic ions
Calciiw
MagnesiuM
Potassiu* 3 3
Sodium 3 3
Chloride
Sulfate
Fluoride 3
Bicarbonate
Carbonate
Amonia 3 3
Nitrate
Phosphate
Silica
Organics
Total organic carbon
Phenohcs 3 3
Kjeldah) nitrogen
Cyanide
In situ leacltjte
to Upper Aquifer
2
2
2
2
2
?
2
2
3
_
1
-.-
...
1
...
.--
...
In situ leachate
to Lower Aquifer
3
3
---
3
2
3
2
1
_
-_.
2
_
_._
___
2
Retort water to
Upper Aquifer
2
3
2
-
--.
3
3
2
2
1
1
2
1
---
1
1
2
Retort water to
Lower Aquifer
3
---
3
-
3
..-
3
3
1
1
2
1
3
1
2
(continued)
-------
TABLE 25 (continued)
Constituent
Sulfur species
Total sulfur
Thiosulfate
TetratMonate
Thiocyanate
Trace elements
Arsenic
Barium
Boron
Bromide
Chromium
,_. Cobalt
!o Iron
Lead
Mercury
HolybdentM
Nickel
Selenium
Titanium
Uranium
Vanadium
Zinc
Radiological
Gross beta
Lower to Lower Aquifer In situ leachate In situ leachate Retort water to
Upper Aquifer to springs to Upper Aquifer to Lower Aquifer Upper Aquifer
3
1
2
1
3 31
3
32 3
3 3 --- 3
2 2
2
3
3
1
... 1 13
3 23
2 22
3
2
3 ... i
3 33
3
Retort water ID
Lower Aquitur
2
1
2
1
1
.--
---
_.-
3
2
2
3
1
3
2
2
3
2
1
3
---
Note:
Enrichment factor (EF) categories: 1 = high likelihood of detection (II"
relatively low likelihood (EF = 10 to SO).
moderate likelihood (liF - bO to 500);
-------
The results of this categorization are shown in Table 25.
For monitoring in the Upper Aquifer for the impact from two major in situ
sources, consider the following listing:
Water quality constituent
Potential source
of impact
Retort water
Enrichment factor >500 Enrichment factor 50 - 500
Conductivity
Alkalinity
Chloride
Bicarbonate
Nitrate
Cyanide
Tetrathionate
Chromium
Cobalt
Selenium
Uranium
Conductivity
TDS
Calcium
Magnesium
Potassium
Chloride
Sulfate
Selenium
Examination of this listing indicates that the following constituents may be
unique indicators of the impact of retort water or spent shale leachate on the
Upper Aquifer. A unique indicator is one which is in the above listing for
one source, but not for the other:
In situ spent
shale leachate
Carbonate
Ammonia
Phosphate
TOC (or DOC)
Kjeldahl N
Thiosulfate
Thiocyanate
Arsenic
Mercury
Vanadium
Carbonate
TOC (or DOC)
Molybdenum
140
-------
Possible unique Indicators
Retort water In situ spent shale leachate
Alkalinity TDS
Bicarbonate Calcium
Ammonia Magnesium
Phosphate Potassium
Nitrate Sodium
Kjeldahl N Sulfate
Thiosulfate Molybdenum
Thiocyanate
Tetrathionate
Cyanide
Arsenic
Chromium
Cobalt
Mercury
Uranium
Vanadium
Following the same procedure for consideration of monitoring in the
Lower Aquifer, the following listing was extracted from Table 25:
Water quality constituent
Potential source
of impact Enrichment factor >500 Enrichment factor 50 - 500
Retort water Carbonate Nitrate
Ammonia Cyanide
Phosphate Total sulfur
TOC Tetrathionate
Thiosulfate Cobalt
Thiocyanate Iron
Arsenic Nickel
Mercury Selenium
Vanadium Uranium
(continued)
141
-------
Water quality constituent
Potential source
of Impact Enrichment factor >500 Enrichment factor 50 - 500
In situ spent Molybdenum Chloride
shale leachate Carbonate
TOC
Chromium
Nickel
Selenium
Possible unique indicators were then identified from this listing:
Possible unique indicators
Retort water In situ spent shale leachate
Ammonia Sulfate
Phosphate Magnesium
Nitrate Chloride
Tetrathionate Chromium
Thiosulfate Molybdenum
Thiocyanate
Arsenic
Cobalt
Iron
Mercury
Uranium
Vanadium
Indicator Constituents
In addition to those water quality parameters for which baseline values
have been established, additional species have been measured on a random basis
in oil shale effluents. These species will be discussed in this subsection.
Inorganic Species
Data presented earlier suggest that those trace elements forming stable,
soluble anions under basic, oxidizing conditions are most likely to be en-
riched in leachates from a spent in situ retort. It is thus interesting to
speculate whether additional elements not discussed above may behave simi-
larly. Other trace elements which form anions under basic, oxidizing
142
-------
conditions include Te, Sb, Bi, Po, W, Re, and I, and their monitoring may
prove valuable. However, a more complete investigation of the geochemistry of
these species is beyond the scope of this book and their potential mobility
remains speculative.
Species such as SCN~, S20§, and $405 are normally not detecta-
ble in groundwater and should, therefore, form excellent indicators of ground-
water contamination. Since background concentrations of these species have
not been measured, enrichment factors (Table 21) were calculated using esti-
mated detection limits as background concentrations, based on the assumption
that their concentrations were less than the detectable limit. The enrichment
factors shown in Table 21 for these species recommend them as possible tracers
of groundwater contamination, especially if even lower detection limits can be
achieved.
Organic Species
The enrichment factors for TOC (or DOC) for both leachates and retort wa-
ter suggest organic matter as a valuable indicator. However, the baseline or-
ganic content of groundwater actually varies widely; Leenheer and Huffman
(1976), for example, indicate levels of DOC of 30,700 mg/1 for trona water
collected near Eden, Wyoming. Few measurements in the Piceance Basin have
been greater than about 10 mg/1. Leachates from raw shale may contain more
organic acids than leachates from spent shale.
For these reasons, individual organic compounds (or compound classes)
which are absent in natural groundwater, but which are produced by the retort-
ing process, should prove to be more sensitive probes of groundwater movement.
For this reason, organic (DOC) fractionation methods, such as those described
by Leenheer and Huffman (1976), may provide a set of useful indicators for
monitoring.
One such type of organic compound could be aromatic acids, which are en-
riched in leachate from spent shale compared to raw shale. In addition, the
smaller (lower molecular weight) aromatic acids should be highly soluble in
the basic conditions expected and should, therefore, follow water movement
closely. The larger acids, although ionized, could be more readily sorbed
and, therefore, migrate less slowly. Polynuclear aromatic hydrocarbons, which
are products of combustion, may also increase during combustion.
Another likely organic tracer would be in hydrophilic bases. Much inter-
est has focused on such compounds,lately because of their biological activity
and unusually large occurrence in oil shale products. Fruchter et al. (1977),
for example, have found that indoles, substituted pyridines, quinolines, and
acridines are highly enriched in shale oil as compared to coal-derived syn-
crude. Sievers and Denny (1978) have also detected numerous organic bases,
many of which could not be readily identified, in retort waters. To the ex-
tent that such organic bases are retained by groundwater, they should provide
sensitive and unusual indicators of groundwater contamination.
143
-------
Stable Isotopes
It is well established that variations in isotopic abundancesespecially
for the light elementsoccur naturally through such processes as diffusion,
evaporation, dissolution, and chemical reaction. For example, "C, is about
3 percent more abundant in ocean bicarbonate than in terrestrial petroleum
(Roboz, 1968).
Similar variations in the isotopic ratios of other light elements, such
as H, N, 0 and S, suggest this measurement as a o robe for studying the migra-
tion of groundwater. As an example, suppose the 2H/*H ratio is slightly higher
in kerogen than in natural groundwater. Water produced by combusting kerogen
will thus be labeled with a higher 2H/*H ratio and could be distinguished
from natural groundwater. Similar considerations should be given to natural
and combustion-produced Nlfy, C0§, and SOf.
The variation in stable isotope abundances is normally reported as parts
per thousand variation from a standard:
(WI.) - (VI,)
s sample standard
5 3 Tyn
standard
where 13 and 1^ refer to the minor and major isotope, respectively.
Variations in isotope ratios are measured almost exclusively by mass
spectrometry. Although any mass spectrometer is capable of measuring isotope
ratios, the measurement of naturally occurring variations requires highly spe-
cialized instruments. Indeed, many isotope ratio mass spectrometers are dedi-
cated to a single element. Consequently, such instruments are found almost
exclusively in research laboratories and are numerically absent from commer-
cial laboratories.
Isotope ratio mass spectrometers are characterized by dual detector sys-
tems which are designed to collect both isotopes simultaneously, thereby mini-
mizing errors due to ion current instability. Detector electronics are
specifically designed to yield the isotope ratio directly, and ion sources
typically include a means of switching rapidly between the sample and a stan-
dard of known Isotopic composition. The precision with which 6 may be mea-
sured in a routine matter .1s about 1 mil for H and 0.1 mil for C, 0, and N.
The precision of 5 is typically limited by isotope fractlonatlon which oc-
curs during sample preparation and introduction into the mass spectrometer.
Although studies of Isotope ratios in the Green River Formation have not
been found in the literature, other relevant Investigations deserve mention.
Friedman et al. (1964), for example, discuss the natural variations of deute-
rium in the hydrologic cycle, including the theory of the fractionatlon pro-
cesses which occur during evaporation, transport, and deposition. They also
report the results of over 1,000 determinations of 'H in waters of North Amer-
ica. Oansgaard (1964) also discusses both the theory and the measurements of
2H and 180 in precipitation.
144
-------
Holt et al. (1972) and Jensen and Nakai (1961) both discuss natural vari-
ations of 34S in environmental samples. Holt et al. (1972) observed perturba-
tions of 5^4$ in surface waters due to rainfall, earth-surface distur-
bances, and effluents from sewage treatment plants.
N isotopic ratios have been studied widely, principally as a means of
identifying pollutant sources and characterizing the atmospheric N cycle
(Moore, 1977; Moore, 1974; Hgering and Moore, 1958; Wada et al., 1975). Natu-
rally occurring values of 515N ranging from -15 to +25 have been observed.
Possible problems which may be encountered in the application of the sta-
ble isotope technique to the Green River Formation include lack of background
data, insufficient difference in 5 for natural and contaminated groundwater,
and exchange reactions such as the following:
1H2HO + ^CO" - ^0 + 2HC03
H2180 + HC1603 + H2160 + HC180160; .
Thus, to the extent that carbonates and bicarbonates exchange with, or precio-
itate as solid materials, the isotopic composition of certain elements ^av be
altered.
SAMPLE ANALYSIS AND COSTS
This discussion is meant to aid the reader in the efficient selection of
analytical techniques suitable for monitoring groundwater movement. Both sur-
vey and element-specific techniques are discussed.
Trace Elements
The most common techniques which are used for trace element analysis are
instrumental neutron activation analysis (INAA), inductively coupled plasma
emission spectroscopy (TCP), spark source mass spectroscopy (SSMS), and atomic
spectroscopy with its various modifications (AA). Each technique has
strengths and weaknesses which should be recognized.
Table 26 compares these techniques on the basis of their abilities to de-
tect trace levels of 44 elements. Although not shown on the table, the limit
for SSMS is typically 1 ug/1 for most elements. The detection limits for ICP
were obtained from a recent review of an ICP spectrometer in use at a DOE syn-
fuels laboratory, and were determined with artificial, multielement standards.
The detection limits shown for a flameless (carbon rod) and flame AA were
taken from the manufacturer's literature. The limits for INAA were for a rou-
tine survey available on a commercial basis. The working limits shown in the
table are the lowest concentrations typically reported by a routine analytical
services laboratory located in Denver. In this case, the working limits are
typically several times the detection limit, since the method of choice in an
analytical services laboratory is determined by regulatory requirement,
145
-------
TABLE 26. COMPARISON OF ANALYTICAL TECHNIQUES FOR TRACE ELEMENT DETERMINATIONS3
fk
a\
Detection limits
Element
Ag
Al
As
B
Ba
Be
B1
Ca
Cd
Co
Cl
Cr
Cu
Ga
Ge
Fe
F
ICP
(ug/l)
3
3
16
15
1
1
80
2
15
5
3
4
30
20
5
Flameless
AA
(ug/1)
0.03
2
15
2
0.2
1.4
0.06
0.02
0.8
0.5
0.4
2
0.5
Flame
AA
(M9/D
2
20
100, 2d
2,000
20
0.7
46
2
0.7
7
5
2
40
100
6
Instrumental
Neutron
Activation
Analysis5
(ug/1)
0.5
100
0.5
NA
100
NA
NA
1,000
0.10
0.5
500
2
300
a-70
NA
200
200
Colorado water
Working limit, quality standards
Denver Laboratory cleanest classification
(M9/I)
O.Ob
100
2
50
50
5
bOO
50
2
10
2QQ
10
10
10
100
Methodc
B
A
B
C
A
A
A
A
A
A
C
A
A
A
0
(ug/1)
0.1
100
50
750
1,000
10
0.4
50
10
300
(continued)
-------
TABLE 26 (continued)
Detection limits
Element
Hg
K
Li
Hg
Hn
Ho
Na
Nb
Ni
P
Pb
Sb
Se
Si
Sn
Sr
S
ICP
(M9/1)
600
50
50
1
5
7
90
30
9
30
20
60
20
30
12
10
Flaroeless
AA
(ug/1)
12
0.2
0.4
0.006
0.04
0.6
0.02
1
0.3
3
6
7
1
0.8
Flame
AA
(M9/D
0.4b
2
2
0.2
2
30
0.3
3,000
8
15
40
250, 2d
200
30
2
Instrumental
Neutron
Activation
Analysis'*
(ug/1)
0.5
300
NA
5,000
20
3
70
^25,000
NA
NA
NA
0.5
1
NA
80
2,000
NA
Colorado water
Working limit, quality standards
Denver Laboratory cleanest classification
(ng/D
0.02
100
5
50
5
5
100
10
100
1
50
5
1,000
500
10
Hethodc (ug/1)
d 0.05
A
A
A 125,000
A 50
B
A
A 50
C
B 4
A
B 10
C
A
A
(continued)
-------
TABLE 26 (continued)
-p.
oo
Detection limits
Workiixj limit,
Denver I dboratory
Colorado water
quality standards
cleanest classification
Element
Te
Ti
Th
Tl
U
V
Zn
U
Br
I
ICP
(M9/0
5
1
200
500
2
10
Flame less
AA
(yg/D
0.6
1,000
10
0.02
---
Flame
AA
(M9/D
40
50
13
60,000
50
1
Instrumental
Neutron
Activation
Analysis'1
(M9/D
2
200
0.2
NA
1
1
10
30
1
30
Methodc
(M9/1)
300
2
b
5
B
E
A
A
15
30
50
Notes:
Detection limits correspond to approximately 20 times the background noise level. Working limits
typically correspond to several times the background noise level and are based on a wide variety of
groundwater and surface water using equipment in a routine fashion.
bNote INAA not approved EPA method.
CA - flame atomic absorption; B - carbon rod atomic absorption; C - colorimetric; D - electrode;
E - fluorometric.
Vapor generation.
NA - not available under normal circumstances or very insensitive.
-------
economics, and ease of operation. It should be recognized that data in Table
26 represent a common basis for discussion; however, detection limits are of-
ten degraded in complex samples or improved by special pretreatment processes.
In addition to the detection limits, the precision and importance inter-
ferences should be considered. ICP is relatively free of matrix interfer-
ences, but is subject to spectral interferences. For example, the DOE
operators have reported poor accuracy for U, Co, As, and Cd on complex sam-
ples, presumably because of spectral interferences. AA has fewer spectral in-
terferences, but special corrections may be needed for background or matrix
interferences. The precision of AA or ICP spectroscopy is typically ±10
percent when used by trained personnel. INAA is often considered a reference
method for trace elements because of its relatively high precision at trace
levels and freedom from matrix interferences. SSMS is typically subject to
fewer interferences than either ICP or AA, but the routine precision for this
technique is about ±40 percent, although precisions of ±3 percent have
been reported in the literature using electrical detection under tightly con-
trolled conditions. Since samples for SSMS must be dried onto a graphite sub-
strate and placed in a vacuum, volatile elements such as Hg, S, and Se may be
lost, especially under acidic conditions.
It is obvious that no single method is a panacea. INAA is attractive be-
cause of its detectability for the potential low-level indicators As, Sb, Se,
Te, U, and V. SSMS is favored as a survey technique because it provides uni-
formly low detection levels and broad elemental coverage. The other methods
listed in Table 26 are attractive as monitoring tools because of their ade-
quate precision and detectability for many elements.
Organic Methods
Common techniques which are available for the determination of trace or-
ganic species in complex mixtures include gas chromatography (GC), combined
gas chromatography/mass spectroscopy (GC/MS), high-pressure liquid chromatog-
raphy (HPLC), and thin-layer chromatography (TLC). Recent advances in con-
trolling the variables in TLC are also giving rise to high-performance,
thin-layer chromatography (HPTLC).
Standardized methods are not normally available for specific organic com-
pounds since operating parameters are optimized for each substrate and
analyte.
For more tractable species, literature references may be found for simi-
lar substrates, -although as a general rule a significant effort will be re-
quired for implementing, adapting, and "debugging" methods for groundwater in
the oil shale area. Organic bases are a particular problem since they readily
decompose and since analytical methods are poorly developed.
Instrumentation should include a GC, GC/MS, and HPLC as a minimum, along
with other standard analytical equipment. The GC/MS should be capable of op-
erating with capillary columns and be capable of peak switching and single ion
monitoring. A specific nitrogen detector on the GC should be considered es-
sential for the determination of organic bases (Sievers and Denny, 1978).
149
-------
Nonspecific separation schemes are also available for classifying the
types of organic compounds in water (Hamersma et al., 1976; leenheer and Huff-
man, 1976). Such schemes can provide a first warning of the groundwater
changes and can indicate otherwise unsuspected changes. The procedures by
Leenheer and Huffman may be of special interest since it was originally con-
ceived as an aid in understanding the movement of organic materials in ground-
water. The procedure operates by separating hydrophilic and hydrophobic
acidic, basic, and neutral compounds based on their adsorptive characteristics
on artificial resins. In this scheme, the hydrophilic fractions should be
most mobile in groundwater, while the hydrophobic fractions should most read-
ily be retained by sorptive clays and minerals.
Other Inorganic Species
For a wide variety of commonly occurring inorganic species, standard
methods have been developed and tested which are reliable when applied to typ-
ical surface water or groundwater and which can be performed with a minimum of
equipment (U.S. EPA, 1974; American Public Health Association, 1976; U.S. Geo-
logical Survey, 1970). Although standard methods must not be applied blindly
to oil shale waste water (or to other waste water), it is believed that many
standard methods can be modified slightly in order to produce more reliable
results. In any case, a carefully designed quality assurance program is
highly recommended.
This subsection first discusses several representative standard analyti-
cal procedures, analytical problems which occur, and possible solutions. A
discussion of possible additional procedures which could be used to better or
more efficiently analyze oil shale waste waters then follows.
Total Suspended and Dissolved Solids
Normally, these are determined by drying an aliquot of water at 103° to
105°C. In retort waters, this may cause the loss of ammonium carbonate and
result in an artificially low result. A possible solution is evaporation at a
different pressure and temperature to more selectively remove the water, or
complete evaporation of ammonium carbonate, which is then determined
separately.
Alkalinity-
Normally, alkalinity is measured by titrating with dilute acid. Results
are typically interpreted as total bicarbonate and carbonate. In retort wa-
ters, dissolved ammonia and organic adds are also titrated so that the re-
sults should be interpreted as "total tltratable base." Another method is to
determine carbonate and bicarbonate by measuring total inorganic carbon in a
TOC analyzer and adjusting the pH and ionic strength. Other options include
acidification of the sample and determination of the evolved C02 titremetrl-
cally, colorimetrically, or by hydrogenatlon and the detection of methane.
150
-------
Chloride-
Chloride is often determined by the subsequent reactions in a continuous
flow system:
2C1" + Hg(SCN)2 * HgCl2 + 2SCN"
SOT + Fe3* - Fe(SCN)x .
The colored ferric thiocyanate complex is then detected colorimetrically.
In retort water, thiocyanate is thus detected as chloride. This problem
should be removed by chemically oxidizing the thiocyanate prior to analysis.
Alternatively, analyzing subsequent samples with and without the addition of
Hg(SCN)2 may provide a determination for both chloride and thiocyanate.
pH-
pH electrodes are subject to fouling by oils. This common problem can
be overcome by frequent standardization or a cross check with a series of pH
indicators, which are certainly as accurate, if not as convenient.
Nitrate-
Often nitrate is determined by the automated Cd reduction method. A com-
mon problem is the fouling of the Cd reduction column by organic materials. A
possible solution is extraction of the organic material prior to analysis, or
the use of an alternate reducing agent, such as hydrazine.
BOD
In our experience, the normal SOD determination is not reproducible un-
less acclimated seed is used.
Ammonia
Often ammonia 1s determined with a selective ion electrode (which is sub-
ject to fouling by organic materials). A likely solution is removal of the
organic materials by extraction, by filtration with a hydrophilic filter, or
by the use of macroreticular resins.
Other Constituents-
It is likely that similar problems and relatively straightforward solu-
tions may exist for other assays, such as fluoride and sulfate. Such minor
modifications may be simple and, indeed, are often practiced by the alert ana-
lytical chemist. There are, of course, requirements for entirely new or
greatly improved analytical methods. Possible analytical schemes are dis-
cussed below as examples.
151
-------
Determination of the complex mixture of sulfur and nitrogen species found
in retort waters is an unresolved problem. In addition, S, $203, $40!,
5305, SOj, SCN~, and CN~ can interract and thereby change their chemical form.
SCN~ can further react with oxidizing agents, which might be used in water
treatment, to form the highly toxic cyanogen chloride.
One approach which has been used (Stuber et al., in press) for this prob-
lem is the cyanolysis of the various sulfur oxides with selective catalysts
(Kelly et al., 1969). The resulting SCN" was detected colorimetrically as the
ferric thiocyanate complex. However, it has not yet been shown that the cata-
lysts are sufficiently selective or that they do not occur naturally in suffi-
cient quantities in waste waters.
There are several possible approaches to this problem which would be
considered:
t Ion chromatography
The development of coloring agents specific for thiosulfate,
thiocyanate, tetrathionate, etc.
t Polarographic techniques which distinguish between the various S
and N species on differing oxidation potentials
Surrogate tests.
The latter tests assume that speciation of the various forms of S and N is not
essential. As an example, $205, SsOf* and S40§ could be determined as a group
using the cyanolysis procedure of Kelly et al. (1969).
An especially attractive technique for such complex waters is ion chroma-
tography. Because it is a separatory technique, complex and selective reac-
tions are not required. Ion chromatography holds the possibility of
chromatograpnically determining cyanide, thiocyanate, sulfate, thiosulfate,
trithionate, tetrathionate, sulfide, as well as phosphate, fluoride and ni-
trate, minutes after sample collection. Because ion chromatography detects
ions nonselectively, the presence of unexpected peaks alerts the analysts to
unknown ions. Thus, the analyst can often detect previously unexpected
compounds.
At the other extreme are tests which would measure, for example, total
sulfur in all forms. Such a technique could be used to alert the analysts to
the need for a more detailed analysis of sulfur species.
INTERPRETATION OF WATER QUALITY DATA
Data Analysis
Data analysis procedures Include (1) checks on data validity, and (2)
methods for presenting data for interpretation for environmental description
or control purposes. Data checking procedures include:
152
-------
Cation-anion balance
t TDS-conductivity comparison
Conductivity-ion comparison (meq/1)
Diluted-conductance method.
The cation-anion balance check involves considering the theoretical
equivalence of the sum of the cations (expressed in milliequivalents per liter
(meg/1)) and the sum of the anions (in meq/1). Because of variations in anal-
ysis which may be unavoidable, exact equivalence is seldom achieved. In gen-
eral, the inequality observed can be expected to increase as the total ionic
concentration increases. When using this method, it is assumed that analyses
of all significant ions have been included and that the nature of the ionic
species is known. In addition, it should be noted that compensating analyti-
cal errors can fortuitously produce a close ion balance. Hence, a combination
of quality control (e.g., replicate analyses, use of standard references,
spiked samples, etc.) and data checking procedures should be employed.
For other analysis checks, samples can be evaporated to dryness at
130°C and the weight compared to the total solids determined by calculation.
""his check is approximate because losses may occur during drying by volatili-
zation and other factors may cause interference (Brown, Skougstad, and Fish-
man, 1970). Another recommended check on analyses involves multiplying
specific conductance (ymhos/cm) by a factor ranging from 0.55 to 0.75. The
product should approximately equal total dissolved solids, in mg/1, for water
samples with TDS below 2,000 to 3,000 mg/1. Also, the specific conductance
divided by 100 should approximately equal the meq/1 of anions or cations.
This relationship is useful in deciding on which sum, cations or anions, is in
error. A more refined method for checking TDS by the electrical conductivity
relationships, called the diluted-conductance method, may also be employed.
Proper design of the monitoring program with regard to selection of moni-
toring sites, sampling frequency and analytical methods, and implementation of
quality control measures will alleviate such data interpretation problems.
Good monitoring design can deal effectively with sources of data variability,
such as operational variability of field instrumentations and errors in calcu-
lations or analysis.
Other significant sources of data variability are events such as in-plant
spills, poor in-plant housekeeping practices, temporary process or control
equipment failure or modification, and other in-plant events. These events
may be entirely random (e.g., spills) or somewhat cyclic (e.g., equipment
maintenance) in nature. Effectively dealing with these sources of data varia-
bility requires liaison with facility operators. Ideally, this communication
should be of two types, namely to assure that (1) monitoring personnel have
adequate knowledge of facility operations (and deviations), and (2) that plant
developers have access to monitoring data and the evaluations made of that
data. Such intercommunication can enhance data interpretation efforts.
153
-------
Data Presentation
Data presentation and interpretation are key aspects of monitoring for
environmental detection and control. Several methods are available for orga-
nization and presentation of water quality data. These include tabulation and
graphical tabulation of appropriate water quality criteria or standards, pro-
viding a format for screening data and identifying important sites or pollu-
tant constituents. Ionic concentrations can be expressed as milligrams per
liter or milliequivalents per liter. Other water quality measures may be seg-
mented into contributing components, such'as total and noncarbonate hardness
or phenolphthalein and methyl orange alkalinity.
Graphic representations of analyses of the chemical quality of water are
useful for display purposes, for comparing analyses, and for emphasizing simi-
larities and differences. Graphs can also aid in detecting the mixing of wa-
ters of different composition and in identifying chemical processes occurring
as water moves through the hydrologic regime of the monitoring area. A vari-
ety of graphic techniques is available; some of the more useful ones are de-
scribed in the following paragraphs.
A widely used method of data presentation is the bar graph. On a bar
graph, each sample analysis appears as a vertical bar whose total height is
proportional to the total concentration of anions and cations, expressed in
mi Hiequivalents per liter. One-half of the bar represents cations and the
other half anions. These segments are divided horizontally to show the con-
centrations of major ions or groups of closely related ions, which are shown
by distinctive patterns. Variations include the addition of individual bar
graphs to express levels of other water quality measures, such as hardness or
un-ionized solutes such as silica.
Water quality data can also be plotted as a set of radiating vectors
(Figure 36). Related methods of showing concentrations as linear vectors re-
sult in constructions of polygons. These approaches are useful in displaying
changes in water quality as changes 1n, for example, the shape of these poly-
gons. Trilinear diagrams are another useful method for representing and com-
paring water quality analyses (Figure 37).
Here, cations, expressed in percentage of total cations (as milliequiva-
lents per liter), plot as a single pofnt on the left triangle. Anions, simi-
larly expressed as a percentage of total anions, appear as a point in the
right triangle. These points are then projected into the central, diamond-
shaped area parallel to the upper edges of the central area. This single
point is thus uniquely related to the total 1on1c quality, and at this point a
circle can be drawn with an area proportional to the total dissolved solids
concentration. The trillnear diagram is a convenient way to distinguish simi-
larities and differences among various water samples as waters with similar
qualities will tend to plot together as groups. Also, simple mixtures of wa-
ters can be identified as the mixture data will plot at locations intermediate
between the mixture component waters.
154
-------
Ml + K 10
12-6
18-1
C«
S04
HCQ3
N«+K 17-3
C*
CI
HC03
0 S 10 18
MIUJEOUIVALENTS PER UTER
Figure 36. Water quality data display using vectors,
i I
I
5 SCALE OF OIAMCTEHS
^
I
PBMCENT OF TOTAL
MILLIEOUIVALENTS PER UTER
Figure 37. Trillnear diagram for displaying water quality data.
155
-------
Other graphic methods include time series plots, plots of variation in
water quality constituents with distance or depth, area or cross-section plots
of equal water quality lines, and plane maps. The choice of data presentation
is determined by the goals of the monitoring program and the type of audience
to which the data are to be presented. The goal of data presentation is to
provide a clear portrayal of the data for evaluation of environmental quality.
Data Interpretation and Reporting
Water quality data from monitoring should be analyzed and interpreted so
as to define quality trends, identify new pollution problems or regions of im-
provement, and assess the effectiveness of pollution control activities. As-
sessments include such things as identifying segments of the groundwater
systems not meeting water quality standards and projections of impact on vari-
ous water uses. The monitoring program should incorporate pertinent data from
all agencies and organizations involved in the monitoring region.
The final result of a monitoring program organized in an area is informa-
tion on water quality. The final task of the monitoring program is to dissem-
inate the information gained in usable forms to the agencies and organizations
concerned with such information.
Monitoring should be summarized in appropriate forms for convenient study
before wide distribution outside of the monitoring agency. This may involve
preparation of tables showing averages and/or changes in water quality. Simi-
larly, graphs prepared to readily display long-term trends may be helpful, as
described previously. Maps showing, for example, locations of major known
sources of pollution, area! distribution of concentrations of key pollutants,
and regions having groundwater with qualities not meeting some water quality
criterion can also be shown to be both useful and effective.
Monitoring information should be distributed regularly to appropriate
public agencieslocal, State, and Federal. Major industries in the area
should also receive the material as well as cooperating agencies and organiza-
tions that contribute monitoring data.
Finally, the monitoring agency would have the responsibility to alert ac-
tion and enforcement agencies of critical problems or situations which are
discovered within the monitoring program. This may involve, for example, de-
tection of hazardous or toxic pollutants which could affect water users.
Prompt reporting of such Instances is essential, as is following up with spe-
cialized monitoring efforts for documenting and controlling emergency
situations.
156
-------
REFERENCES
Ahrens, T.P., and A.C. Barlow, Permeability Tests Using Drill Holes and Wells,
Including Comments Regarding Equipment, Etc., Bureau of Reclamation Geol-
ogy Report No. G-97, 1951.
American Public Health Association, Standard Methods, 19th Edition, Washing-
ton, O.C., 1976.
Brown, E., M.W. Skougstad, and M.J. Fishman, "Methods for Collection and Anal-
ysis of Water Samples for Dissolved Minerals and Gases," Techniques of
Water-Resources Investigat 1.on, Book 5, Chapter A-l, U.S. Geological Sur-
vey, 1970.
Bureau of Reclamation, Ground Water Manual, U.S. Department of Interior, 1977.
C-b Shale Oil Venture, Quarterly Data Releases, Oil Shale Tract C-b, submitted
to the Area 011 Shale Supervisor, U.S. Geological Survey, U.S. Department
of Interior, 1974 and 1975.
C-b Shale 011 Venture, Environmental Baseline Program, Final Report, Oil Shale
Tract C-b, submitted to the Area. 011 Shale Supervisor, U.S. Geological
Survey, U.S. Department of Interior, 1976.
C-b Shale Oil Venture, Application for a Permit to Inject Water Produced from
011 Shale Tract C-b, submitted to the State of Colorado, February 1979.
Chow, V.T., "On the Determination of Transmlssivity and Storage Coefficients
from Pumping Test Data," American Geophysical Union Transactions, Vol.
33, pp. 397-404, 1952.
Cooper, H.H., and C.E. Jacob, "A Generalized Graphical Method for Evaluating
Formation Constants and Summarizing Well Field History," American Geo-
physical Union Transactions, Vol. 27, pp. 526-534, 1946.
Dansgaard, W., '"Stable Isotopes in Precipitation," Tellus, Vol. 16, No. 33,
1964.
Everett, L.S., K.O. Schmidt, R.M. T1nl1nt and O.K. Todd, Monitoring Ground-
water Quality; Methods and Costs, U.S. Environmental Protection Agency
Report 600/4-76-023, May 1976.
157
-------
Everett, L.G. (editor), Groundwater Quality Monitoring of Western Coal Strip
Mining; Identification and Priority Ranking of Potential Pollution
Sources, EPA-6QO/7-79-OZ4, U.S. Environmental Protection Agency. January
T577I
Everett, L.G., Groundwater Monitoring, General Electric Co., Technology Mar-
keting Operation, Schenectady, NY, 1980.
Fox, J.P., R.D. McLaughlin, T.C. Bartke, J.J. Duvall, K.K. Mason, and R.E.
Poulson, "Mercury Emissions from a Simulated In Situ Oil Shale Retort,"
llth Annual 011 Sha1e Sympos i urn Proceed ings, Golden, Colorado, April
12-14, 1978.
Friedman, I., A.C. Redfield, B. Schoen, and J. Harris, "The Variation of the
Deuterium Content of Natural Waters in the Hydrologic Circuit," Reviews
of Geophysics, Vol. 2, No. 177, 1964.
Fruchter, J.S., J.C. Laul, M.R. Peterson, P.M. Ryan, and M.E. Turner, "High
Precision Trace Element and Organic Constituent Analysis of Oil Shale and
Sol vent-Refined Coal Materials," ACS New Orleans Meeting, Division of Pe-
troleum Chemistry, Symposium on Analytical Chemistry of Tar Sands and Oil
Shale. March 20-25, 1977.
Hamersma, J.W., S.L. Reynolds, and R.F. Maddalone, Level 1 Environmental As-
sessment, EPA-600/2-76-160a, U.S. Environmental Protection Agency, June
T376":
Hantush,, M.S., "Analysis of Data from Pumping Tests in Leaky Aquifers, Ameri-
can Geophysical Union Transactions. Vol. 37, pp. 702-714, 1956;
Hantush, M.S., "Analysis of Data-from Pumping Tests in Anisotropic Aquifers,"
Journal of Geophysical Research, Vol. 71, pp. 421-426, 1966.
Hantush, M.S., and C.E. Jacob, "Non-Steady Radial Flow in an Infinite Leaky
Aquifer," American Geophysical Union Transactions, Vol. 36, pp. 95-100,
1955.
Hantush, M.S., and R.fi. Thomas, "A Method for Analyzing a Drawdown Test in
Anisotropic Aquifers," Water Resources Research, Vol. 2, pp. 281-285,
1966.
Hendricks, O.W., and J.C. Ward, "Environmental Analysis of an Oil-Shale Indus-
try in the Upper Colorado Region," 011 Shale (T.F. Yen and G.V. Chiling-
arian, editors), 1976.
Hodgson, R.A., Photo Linear Study. MDP Area, Rio Blanco Oil Shale Project,
Tract C-a. Qulf R&O Report, March 1979.
Hoering, T.C., and H.E. Moore, "The Isotopic Composition of the Nitrogen in
Natural Gases and Associated Crude 011s," Geochemica et Cosmochemlca
Acta. Vol. 13, No. 225, 1958.
158
-------
Holt, B.D., A.G. Engelkeneir, and A. Benters, "Variation of Sulfur Isotope
Ratios in Samples of Water and Air Near Chicago," Env. Sci. & Tech.,
Vol. 6, pp. 338-341, 1972.
Homer, D.R., "Pressure Build-up in Wells," Proceedings, Third World Petroleum
Congress. Vol. 11, p. 503, The Hague, 1951.
Humenick, M.J., L.J. Turk, and M.P. Cochin, "Methodo "logy for Monitoring at
Uranium Solution Mines," Ground Water Journal. Vol. 18, No. 3, 1980.
Jacob, C.E., "On the Flow of Water in an Elastic Artesian Aquifer," American
Geophysical Union Transactions, Vol. 72, Part II, pp. 574-586.
Jensen, M.L., and N. Nakai, "Sources and Isotocnc Comoosition of Atmosoheric
Sulfur," Science. Vol. 134, pp. 2102-2104, 1961.
Johnson Division, UOP Co., Ground Water and Wells, Saint Paul, Minnesota,
1966.
Johnson, E.E., Ground Water and Wells, Johnson Division, UOP Inc., Saint Paul,
Mi nnesota, l97Sj.
Kelly, O.P., L.A. Chambers, and P. A. Trudinger, "Cyanolysis and Soectrophoto-
metric Estimation of Tr^-.hionate in lixtjr? -Hth ""'osulfats and Tetra-
thionate," Anal. Chem. .. Vol. 41, p. 398, 1969.
Knutson, C.F., Project Rio Blanco: Evaluation of Possible Radioactivity
Transport in Groundwater. CER Seonuclear Corporation, Las Vegas, Nevada,
1973.
Kruseman, G.P., and N.A. Ridder, Analysis and Evaluation of Pumping Test Data,
International Institute of Land Reclamation and Improvement, Bulletin 11,
Wageningen, The Netherlands* 1976.
leenheer, J.A., and E.W.D. Huffman, Jr., "Classification of Organic Solutes in
Water by Using Macroreticular Resins," Journal Research, U.S. Geological
Survey. Vol. 4, No. 6, p. 737, 1976.
Marsh, J.M., and J.W. Lloyd, "Details of Hydrochemical Variations in Flowing
Wells," Ground Water Journal. Vol. 18, No. 4, 1980.
Miller, L.M. (editor), Ground Water. American Water Well Association, New
York, New York, 19731
Moore, H., "Isotoplc Measurement of Atmospheric Nitrogen Compounds," Tell us.
Vol. 26, No. 167, 1974.
Moore, H., "The I so topic Composition of Ammonia, Nitrogen Dioxide and Nitrate
1n the Atmosphere," Atmos. Environ. II. No. 1239, 1977.
Neuman, S.P., and P. A. WUherspoon, "Theory of Flow in a Confined Two-aquifer
System," Water Resources Research. Vol. 5, pp. 803-816, 1969.
159
-------
Odeh, A.A., and L.G., Jones, "Pressed ur-awdown Analysis, Variable Rate Case,"
Journal of Petroleum Technology, pp. 960-964, August 1965.
Parker, H.W., R.M. Bethea, N. Gunen, M.N. Gazdar, and J.C. Watts, "Interac-
tions Between Ground Water and In Situ Retorted Shale," Proceedings Sec-
ond Engineering Congress, 1977.
Rio Blanco Oil Shale Company, Quarterly Report, 1974.
Rio Blanco Oil Shale Company, Final Environmental Baseline Report for Tract
C-2 and Vicinity, May 1977.
Roboz, J., "Introduction to Mass Spectrometry, Interscience, New York, 1968.
Schmidt, K.D., "Water Quality Variations for Pumping Wells," Ground Water
Journal, Vol. 15, No. 2, 1977.
Sievers, R.E., and D. Denny, Investigation of Trace Organic Compounds in Oil
Shale Wastes. Progress Report, 1978.
Slawson, G.C. (editor), Groundwater Quality Monitoring of Western Oil Shale
Development: Identification and'Priority Ranking of Potential Pollution
Sources. EPA-600/7-79-023, U.S. Environmental Protection Agency, January
WFT.
Slawson, G.C., Jr., Monitoring Groundwater Qu*Ilty:_J?on^t:or'*nq Program Devel-
opment. GE78TMP-90, General ElectricTEMPO, 1980a.
Slawson, G.C., Jr., Monitoring Groundwater Quality: The Impact of In-Situ
Shale Retorting. GE78TMP-J03. General ElectricTEMPO, 1980b.
Sokal, R.R., and F.J. Robly, Biometry. W.H. Freeman Co., San Francisco, 1969.
Stuber, H.A., J.A. Leenheer, and D.S. Farrier, Inorganic Sulfur Species in
Waste Waters from In Situ Oil Shale Processing, in press.
Summers, W.K., "Factors Affecting the Validity of Chemical Analyses of Natural
Water," Ground Water. Vol. 10, No. 1, p. 12-17, 1972.
Summers, W.K., and L.A. Brandvold, "Physical and Chemical Variations in the
Discharge of a Flowing Well," Ground Water Journal. Vol. 5, No. 1, 1967.
Theis, C.W., "The Relation between the Lowering of the Piezometric Surface and
the Rate and Duration of the Discharge of a Well Using Groundwater Stor-
age," American Geophysical Union Transactions. Vol. 16, pp. 519-524,
1935.
Tiratsou, E.N., Qilf 1elds of the World (2nd Edition), Scientific Press, Ltd.,
Beaconsfield, England, 1976.
160
-------
Todd, O.K., R.M. Tinlin, K.D. Schmidt, and L.G. Everett, Monitoring Groundwa-
ter Quality: Monitoring Methodology . EPA-600/4-76-076, U.S. Environmen-
tal Protection Agency, 197&.
U.S. Atomic Energy Commission, Environmental Statement; Rip Blanco Gas Stimu-
lation Project, Rio Blanco County, Colorado, April 19/2.
U.S. Environmental Protection Agency, Methods for Chemical Analysis of Water
and Wastes. Report 625/6-74-003, IfHT.
U.S. Environmental Protection Agency, Procedures Manual for Ground Water Moni-
toring at Solid Waste Disposal Facilities. Report No. 530/SW-611. pre-
pared for Office of Solid Waste, 1977.
U.S. Environmental Protection Agency, Methods for Chemical Analysis of Water
and Wastes. Report 600/4-79-020,
U.S. Geological Survey, "Methods for Collection and Analysis of Water Samples
for Dissolved Minerals and Gases," Techniques of Water Resources Investi-
gations of the U.S. Geological Survey, U.S. Government Printing Office.
Washington, D.C., 19/u.
U.S. Geological Survey, Inflatable Straddle Packers and Associated Equipment
for Hydraulic Fracturing and Hydroloqic Testing. Water Resources Investi-
gations 78-55, PB 288-403, 1978.
Wada, E., T. Kadonga, and S. Matsuo, "15N Abundance in Nitrogen of Naturally
Occurring Substances and Global Assessment of Oenitrification from Iso-
topic Viewpoint," Geochemical Journal. Vol. 9, No. 139, 1975.
Walton, W.C., Selected Analytical Methods for Well and Aquifer Evaluation,
Illinois State Water Survey Bulletin No. 49, 1962.
Weeks,. J.B., G.H. Leavesley, F.A. Welder, and G.J. Saulnier, Jr., Simulated
Effects of Oil Shale Development on the Hydrology of Piceance Basin, Col-
orado. U.S. Geological Survey Professional Paper 908. 1974.
Welch, P.S., "Lirnno logical Methods." (Cited in Everett 1976).
Welder, F.A., and G.J. Saulnier, Jr., Geohydraulic Data from Twenty-four Test
Holes Drilled in the Piceance Bas_'P* 310 B1Jl"c°__Coun*y* Colorado.
19/&-/6. U.S. Geological Survey Open File Report 78-734. 1978.
Wood, W., "Guidelines for Collection and Field Analysis of Ground-Water Sam-
ples for Selected Unstable Constituents," Techniques of Water -Resource
Investigations. Book 1, Chapter 02, U.S. Geological survey 1975.
161
*US GOVERNMENT PRINTING OFFICE 1983 - 659-095/0763
------- |