EPA-600/9-76-011 '
This document has not been
submitted to NTIS, therefore it
should be retained.
PROCEEDINGS OF THE1
FIRST WORKSHOP ON
SAMPLING
GEOTHERMAL
EFFLUENTS
HELD ON
OCTOBER 20-21,1975
AT THE
ENVIRONMENTAL MONITORING
& SUPPORT LABORATORY
LAS VEGAS, NEVADA
PRO
-------
EPA-600/9-76-011
May 1976
PROCEEDINGS OF THE FIRST WORKSHOP ON
SAMPLING GEOTHERMAL EFFLUENTS
October 20-21, 1975
Las Vegas, Nevada
conducted by
Monitoring Systems Design and Analysis Staff
Monitoring Systems Research 6 Development Division
Environmental Monitoring and Support Laboratory
Las Vegas, Nevada
U.S. Environmental Protection Agency
Region 5 Library (PL-12J) V
77 West Jackson Blvd., 12th Floor
Chicago, IL 60604-3590
U.S. ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF RESEARCH AND DEVELOPMENT
ENVIRONMENTAL MONITORING AND SUPPORT LABORATORY
LAS VEGAS, NEVADA 89IIk
-------
DISCLAIMER
The responsibility for the technical accuracy and clarity of the
manuscripts included in these proceedings has been placed solely upon
their authors. Mention of trade names or commercial products does not
constitute endorsement or recommendation by the U.S. Environmental
Protection Agency.
ii
-------
FOREWORD
The U.S. Environmental Protection Agency contends that the
first step in any viable assessment program is to obtain samples
under standard conditions with quality-assured methods. Without
this first step, the balance of the examination for pollutants is
of questionable value.
In October 1975, the Agency's Environmental Monitoring and
Support Laboratory in Las Vegas held an initial workshop on geothermal
energy development. The purpose of this first workshop was to generate
that necessary exchange of ideas and knowledge needed in developing
a set of standard geothermal sampling methods with assurance of quality
in the methods. The goal of this effort was the formation of a recog-
nized Standard Sampling Method Handbook. The response to this first
workshop was encouraging, leading us to strongly believe that the
goal is attainable.
The Environmental Monitoring and Support Laboratory-Las Vegas
wishes to take this opportunity to express its appreciation and
gratitude to those organizations and persons who gave so freely
of their time and resources to make this first meeting a success.
George B. Morgan
Di rector
Monitoring Systems Research
and Development Division
Environmental Monitoring and
Support Laboratory - Las Vegas
i i i
-------
LIST OF ATTENDEES
Gordon W. Allen
R. J. AUmendinger
Jon Baldwin
J. Allen Bell
Richard Bell
J. E. Biesecker
H. K. Bishop
R. L. Booth
E. W. Bretthauer
J. Cardinal!!
J. B. Cotter
D. Christoffersen
E. Crecel ius
A. Crockett
Gene Culver
S. Dermengian
Y. Echstein
G. Edwards
T. D. English
D. Fach
J. Fleiner
R. C. Fowler
J. Fruchter
D. B. Gil more
P. H. Gudiksen
W. R. Hail
R. P. Hartley
F. B. Henderson, I (
M, H, Hyman
L. S. Ischlnger
Pacific Gas & Electric Company
Chevron Oil Company
Allied Chemical Corporation
Las Vegas Valley Water District
VTN Corporation
U.S. Geological Survey
San Diego Gas & Electric
U.S. Environmental Protection Agency
U.S. Environmental Protection Agency
State of Nevada Division of Water Resources
U.S. Energy Research & Development Adminis.
Union Oil Company
Battelle Northwest Laboratory
U.S. Environmental Protection Agency
Oregon Institute of Technology
Geothermal Energy Magazine
Hydro-Search, Inc.
Halliburton Services
Jet Propulsion Laboratory
U.S. Navy
Geonomics, Inc.
Mechanics Research, Inc.
Battelle Northwest Laboratory
U.S. Environmental Protection Agency
Lawrence Livermore Laboratory
W. A, Wahler & Associates
U.S. Environmental Protection Agency
Lawrence Berkeley Laboratory
Frederiksen Engineering Company
U.S. Fish and Wildlife Service
iv
-------
D. James
E. A. Jenne
Joel Jobst
R. Kaufmann
R. C. Kent
J. J. Koranda
P. Kruger
J. T. Kuwanda
J. La Fluer
V. W. Lambou
S. LeGore
J. D. Ludwick
A. MacTavish
J. R. McBride
L. B. McMillion
N. Melvin
D. E. Michels
M. J. Miles
J. A. Neil son
M. F. O'Connell
L. B. Owen
S. Porter
G. D. Potter
M. G. Reed
A. J. Regis
W. D. Riley
D. Robertson
J. A. Rogers
L. Schieler
E. A. Schuck
R. C. Scott
D. W. Shannon
D. Shirley
F. E. Smith
A. J. Soinski
U.S. Energy Research 6 Development Adminis.
U.S. Geological Survey
Edgerton, Germeshausen & Greer, Inc.
U.S. Environmental Protection Agency
U.S. Geological Survey
Lawrence Livermore Laboratory
Stanford University
Rogers Engraving Company, Inc.
Burmah Oil & Gas Company
U.S. Environmental Protection Agency
Parametrix, Inc.
Battelle Northwest Laboratory
Mechanics Research, Inc.
U.S. Environmental Protection Agency
U.S. Environmental Protection Agency
U.S. Bureau of Land Management
Aerojet Nuclear Company
Desert Research Institute
Ecoview Environmental Consultants
U.S. Environmental Protection Agency
Lawrence Livermore Laboratory
U.S. Bureau of Land Management
U.S. Environmental Protection Agency
Chevron Research Company
U.S. Bureau of Land Management
U.S. Bureau of Mines
Battelle Northwest Laboratory
U.S. Environmental Protection Agency
Woodward-Clyde Consultants
U.S. Environmental Protection Agency
U.S. Environmental Protection Agency
Battelle Northwest Laboratory
Jet Propulsion Laboratory
U.S. Fish £ Wildlife Service
LFE Corporation
-------
R. E, Stanley
R. T. Stearns
R. W. Steele
A. Stoker
R. Tin]in
A. C. Trakowski, Jr.
E. Wahl
E. F. Wehlage
E. A, We11 man
R. N. Wheatley
D. W. Wheeler
G. B. Wiersma
R. Williams
H. Wollenburg
U.S. Environmental Protection Agency
U.S. Energy Research 6 Development Adminis.
Parametrix, Inc.
Los Alamos Scientific Laboratory
General Electric-TEMPO
U.S. Environmental Protection Agency
Occidental Research & Development Company
Consultant
BWR Associates
Union Oil Company
Burmah Oil & Gas Company
U.S. Environmental Protection Agency
University of Denver Research Institute
Lawrence Livermore Laboratory
VI
-------
CONTENTS
Papers marked with * were not received in time to be included
in this publication.
Foreword i i i
List of Attendees iv
Session I
Geothermal Energy Development 1
Dr. Paul Kruger
Civil Engineering Department
Stanford University
Stanford, California
Geothermal Effluents, Their Toxicity and
Prioritization 36
Dr. Leroy Schieler
Senior Project Scientist
Woodward-Clyde Consultants
San Francisco, California
The Need for Standard Sampling Methods and
Sampling Preservations
Mr. Robert L. Booth
Technical Advisor
EMSL-Cincinnati, Ohio
Quality Assurance
Mr. ErichBretthauer
EMSL-Las Vegas
Las Vegas, Nevada
The Need for Defensible Data
Mr. James A. Rogers
Office of General Counsel
EPA, Washington, D. C.
VI I
-------
Session I I
Some Problems Involved with Sampling Geothermal 67
Sources
Alan K. Stoker and William D. Purtymun
Los Alamos Scientific Laboratory
Los Alamos, New Mexico
Drill Stem Testing and Sampling of Geo-Pressured 97
Brines
A. G. Edwards and J. M. Montgomery
Halliburton Services
Duncan, Oklahoma
Application of Oil Well Sampling Technology to
Geothermal Fluid Sampling
Dr. Marion G. Reed
Chevron Research
San Francisco, California
The Salinity Profile of the East Mesa Field as
Determined from Dual Induction Resistivity and
SP Logs
R. T. Littleton and E.E. Burnett
U.S. Bureau of Reclamation
Boulder City, Nevada
Field Sampling of Radioactive Geothermal Effluents 126
Arthur J. Soinski, David E. Claridge
and Rodney Melgard
LFE Corporation
Richmond, California
Sampling Hot Springs for Radioactive and Trace
Elements
Harold A. Wollenberg
Lawrence Berkeley Laboratory
Berkeley, California
VIII
-------
Page
Session I I I
Pacific Gas and Electric's Geothermal Sampling
Techniques
Gordon W. Allen
Pacific Gas and Electric Company
San Ramon, California
Union Oil Company of California's Geothermal 165
Sampling Techniques
D. J. Christoffersen, R.N. Wheatley, and
J. A. Baur
Union Oil Company of California
Brea, California
Brine Sampling and Analysis Techniques in Support
of Corrosion and Scaling Studies in the Imperial
Va11ey
P. Needham, W. Riley and A. Murphy
U.S. Bureau of Mines
College Park, Maryland
Geothermal Sampling Procedures at the Raft River
Project
Jon Baldwin
Allied Chemical Corporation
Twin Fal1s, Idaho
Atmospheric Discharge Sampling While Drilling
Geothermal Steam Wells
M. H. Hyman and G. R. Fox
Frederiksen Engineering Company
Oakland, California
Sampling a Two-Phase Geothermal Brine Flow 181
for Chemical Analysis
J. H. Hill and C. J. Morris
Lawrence Livermore Laboratory
Livermore, California
IX
-------
Page
Session IV
Approaches to Interpreting Environmental Data 204
Donald E. Michels
Aerojet Nuclear Company
Idaho FalIs, Idaho
Sampling and Preservation Techniques for Waters 218
in Geysers and Hot Springs
James W. Ball, Everett A. Jenne, and
J. M. Burchard
U.S. Geological Survey
Menlo Park, California
-------
SGP-TR 9
GEOTHERMAL ENERGY DEVELOPMENT
Paul Kruger
Civil Engineering Department
Stanford University
Stanford, California
*
based on reports prepared during Leave of Absence 1974^75 with the
National Science Foundation and the Energy Research and Development
Administration.
-------
INTRODUCTION
The Nation has embarked on an aggressive program to develop its
indigenous resources of geothermal energy. For more than a decade,
geothermal energy has been heralded as one of the more promising forms
of energy alternate to oil and gas for electric power generation, but
during the last fifteen years, the total capacity in the U.S. has
reached 502 MWe, about half the size of a single modern nuclear power
plant. And yet, the United States, especially its western and Gulf
coast states, is believed to possess a vast resource base of geothermal
heat at depths up to 3 to 10 km. Many estimates of these potential re-
sources suitable for the production of electric power have been pub-
lished and they range over a spectrum of more than a factor of 100.
This variation suggests that the potential is essentially unknown.
Table 1 gives a range of published forecasts for the year 1985 and
the equivalent potential in number of 1000 MWe power plants and in oil
consumption in millions of barrels per day. In view of the estimated
construction of about 200 to 250 nuclear power reactors by 1985-90, the
pessimistic forecasts clearly show that the contribution of geothermal
energy to the Nation's energy supply may indeed be small. The optimis-
tic forecasts represent more than 157o of the total electric power re-
quirements estimated for the year 1985. The Task Force for Geothermal
Energy, in the Federal Energy Aministration Project Independence Blue-
print report of November 1974, established a national goal for 1985 of
20,000 to 30,000 MWe, the latter value representing an equivalent en-
ergy supply of one million barrels of oil per day. This goal was
-------
GO
UJ
CD
GO
a.
CD
a:
o
CD
CD
CD
>
CD
CD
I
CD
CD
CD
\
CSI
00
CD
CD
CSI
oo
I
CO
r 1
UJ
_J
rt
<
CD
LU
CD
OO
ZD
U_
CD
1
!5^
LU
s:
Q_
CD
1
LU
>
UJ
a
Q£
CD
u_
1
~
^ ^~ x
j^
Q£ ^
UJ I-
mi "^
1 0
Q 0.
Z <
u >-
- tD
or a:
1 LU
o -z.
LU LU
J
LU
CO
cc.
o
1-
o
z <
« LU
rv**
h-
z a:
LU <
_J LU
> O
>- ID
a ~~
LU
CD
1
CD
CSI
^j-
1
CSI
CO
1-
1 1
z
LU
"^g
^r~
CD
CD
CD
i-H
.
O
s_x
o
l»4
1-
a.
2:
CO
2
OV
1-
LU
1
^
>
i_i
^t
a»
LU
^^j
t~i
i
f>»^
_
CD
^^
i I
CD
1
f^
CD
CD
^
^
Q
_l
Cfl
OQ
£
CD
t |
1
» «
CD
LU
O
-------
clearly a compromise between what is worth a national effort and what
might be realistically achieved. The potential for adding or replacing
the equivalent of some 25 nuclear power plants or for conserving one
million barrels of oil per day should be an adequate incentive for the
Nation to accelerate the development of a viable geothermal industry.
A puzzling enigma appears. If the potential resource base of geo-
thermal energy is so vast, why has significant utilization not occurred?
The entire U.S. production of electric power from geothermal resources
occurs at one location, the Geysers in California, where over a 15-year
period starting in 1960, generating capacity has grown from 12 MW sup-
plied by Unit No. 1 to the total of 502 MW attained with the startup
of the 106-MW Unit No. 11 in May 1975. The Geysers is the largest geo-
thermal electricity generating station in the world. The entire world-
wide capacity of electric power generation by geothermal resources is
slightly more than 1000 MW, the equivalent of the capacity of a single
modern nuclear power plant.
Utilization of geothermal fluids for thermal energy in the U.S. is
almost negligible. And yet throughout the country, fossil fuels are
consumed in large quantities to boil water for heating and electric
power generation, both at very low thermal efficiency. Some countries
already use geothermal fluids for its thermal energy, notably Iceland,
where municipal heating is an important utilization. Several countries,
responding to increased public awareness that future supply of fossil
fuel may be very limited, are examining the potential use of indigenous
thermal waters for industrial and municipal heating.
How is this enigma to be solved; how is the United States (and
other countries) endowed with potentially-bountiful geothermal resources
-------
going to develop these natural resources as a significant contribution
to its energy supply? The attainment of a national goal to contribute
an equivalent of one million barrels of oil per day from geothermal re-
sources clearly requires accelerated development of a geothermal industry
capable of providing 20,000 to 30,000 MW of electric power and thermal
energy in the next ten to fifteen years. And this objective will require
a national effort to accelerate and coordinate development in three
parallel tasks: (1) the discovery, proving, and extraction of geothermal
resources to provide a significant supply of hydrothermal fluids for
12
direct utilization and to produce more than 5 x 10 kWh of electricity
over the amortization period of the investment in resource development
and power plant construction, (2) the technology to convert the resources
as found in its various natural forms and qualities into electricity,
and (3) the removal of unnecessary institutional constraints to the
rapid development of a cost-effective and environmentally-acceptable
industry.
A major factor which helps create the enigma of vast resource base
and little utilization is the variability of geothermal resources. The
geothermal energy cycle, although simple compared to other alternate
energy sources, is actually complex in that geothermal resources occur
in many types of geologic, thermodynamic, hydrodynamic, and chemical
quality. As a result, the major problems in the energy cycle vary by
type of resource. Table 2 lists the key aspects of the cycle from
exploration to utilization that must be evaluated for each type of
resource.
Several general reviews of the state of the art of geothermal en-
ergy resources and technology are listed in the Bibliography. One is
-------
TABLE 2
PROBLEM AREAS IN THE DEVELOPMENT OF THE
GEOTHERMAL ENERGY CYCLE
VARIABILITY OF GEOTHERMAL RESOURCES
LOCATION OF SUBSURFACE RESERVOIRS
RESERVOIR EVALUATION
EXTRACTION TECHNOLOGY
CONVERSION TECHNOLOGY
POTENTIAL FOR MULTIPLE UTILIZATION
ENVIRONMENTAL IMPACT CONTROL
LEGAL AND INSTITUTIONAL CONSTRAINTS
-------
the proceedings of the 1970 United Nations symposium on the development
and utilization of geothermal resources. Another is the 1973 compila-
tion of an Ad Hoc Working Group convened by UNESCO. A general introduc-
tion to geothermal energy is the proceedings of the American Nuclear
Society conference on geothermal resources, production, and stimulation
held in 1972. Among other compilations of papers on geothermal energy
are the proceedings of the second and third Ail-Union conferences on
geothermal energy organized by the Scientific Council for Geothermal
Investigations of the USSR Academy of Sciences. Translations of these
proceedings are not generally available, but much of-the technical con-
tent is given in the Soviet papers of volume 2 of the United Nations
Symposium and in the ARPA reviews of Soviet literature in geothermal
energy. The proceedings of the Second United Nations Symposium held
in San Francisco in May 1975, adds another major contribution to the
literature of geothermal energy.
GEOTHERMAL RESOURCES
f\f
The upper 10 Ion of the earth's crust may contain more than 3 x 10
cal of heat, a resource base readily classified as vast. However, much
of this energy is too diffuse to be exploitable as an energy source.
Geothermal resources may be defined as localized deposits of geothermal
heat concentrated at attainable depths, in adequate volumes, and at tem-
peratures sufficient for commercial exploitation.
The only geothermal resources presently used for electric power
generation are high-quality hydrothermal convective systems which con-
tain high-enthalpy geofluids suitable for transferring the geothermal
heat to the surface for direct use in low-efficiency steam turbines.
-------
Unfortunately such resources have been discovered at only a few places
on earth. More than 75% of the world's geothermal electric power capac-
ity results from vapor-dominated hydrothermal systems which produce dry
or super-heated steam for direct conversion. The remaining capacity re-
sults from high-temperature, low-salinity, water-dominated hydrothermal
systems in which the geofluids are flashed on production, and only the
separated steam is used for electric power generation. The liquid frac-
tion is either wasted or reinjected into the ground. These systems are
commercially less desirable because only a small fraction of the water
flashes to steam, thermal efficiencies are low, and plant operational
problems are more severe.
Liquid dominated hydrothermal systems are expected to be many times
more abundant than vapor-dominated hydrothermal systems. Moderate-to-
high salinity hydrothermal resources may be more abundant than low
salinity resources. And other types of geothermal resources, such as
hypersaline brines, geopressured fluids, volcanic and magmatic deposits,
and impermeable hot-rock massives, which are not yet commercially ex-
ploitable, may be even more abundant than the currently exploited hydro-
thermal resources. Thus the answer to the utilization enigma may lie not
so much with the magnitude of the resource base, but more with the abil-
ity to locate suitable concentrated deposits of geothermal heat and the
technology to extract the energy in quantities which are economically and
environmentally feasible.
Although estimates of the geothermal resource base are available,
the magnitude of the potential reserves is not yet well defined. The
location of underground deposits of geothermal heat, especially where
8
-------
thermal manifestations are not visible at the surface, is a difficult
task. Over one million acres of "hot spots," areas of known geothermal
energy, were identified as early as 1967 by the U.S. Department of the
Interior in designated Federal lands in five western states as having
current potential value as geothermal resources. An additional 86 mil-
lion acres of land in thirteen states were designated as prospectively
valuable for geothermal resources. Since then several other inventories
of known geothermal resource areas (KGRA) have been compiled. A current
assessment of U.S. geothermal resources has been completed by the U.S,
Geological Survey and a summary of the resource base, by resource type,
is given in Table 3.
Exploration for geothermal resources has been undertaken by industry
on private lands, and through the Federal Leasing Act of 1970, on Fed-
eral lands by competitive and non-competitive leasing under supervision
of the Bureau of Land Management. Although total values are difficult to
ascertain, it is estimated that about 100,000 acres on Federal public
lands and about 200,000 acres on Federal Indian lands were under lease
for geothermal exploration in mid-1975.
Resource exploration and assessment of potential reservoirs of
geothermal energy are made by the variety of earth science methods
listed in Table 4. Details of these methods are available in the general
references listed in the Introduction. The final phase of geothermal
exploration is the drilling of exploratory wells. It is from these
wells that data for evaluating the suitability of the resource as a pro-
duction reservoir are obtained. Major factors in the economics of ex-
ploration and production of geothermal fields are the success of
-------
TABLE 3
SUMMARY OF GEOTHERMAL RESOURCE BASE OF THE UNITED STATES*
-1 Q
ESTIMATED HEAT CONTENT (10 CAL)
IDENTIFIED POTENTIAL
HYDROTHERMAL CONVECTION SYSTEMS
VAPOR-DOMINATED (STEAM) 26 50
HIGH T - HOT WATER (T > 150°C) 370 1,600
MOD T - HOT WATER (90° - 150°C) 345 1,400
TOTAL 740 3,000
HOT IGNEOUS SYSTEMS
MAGMA AND HOT ROCK 25,000 100,000
GEOPRESSURED BASIN PART OF
REGIONAL CONDUCTIVE SYSTEMS 10,920 44,000
TOTAL RESOURCE BASE 36,660 147,000
*
FROM D, F, WHITE AND D, L, WILLIAMS, EDS,, ASSESSMENT OF
GEOTHERMAL RESOURCES OF THE UNITED STATES - 1975, U, S,
GEOLOGICAL SURVEY CIRCULAR 726, 1975,
10
-------
TABLE 4
GEOTHERMAL EXPLORATION METHODS
EXPLORATION SURVEYS
AIRBORNE
AEROMAGNETIC SURVEY
THERMAL INFRARED SURVEY
GEOLOGICAL
TECTONICS AND STRATIGRAPHY
RECENT FAULTING
DISTRIBUTION AND AGE OF VOLCANIC ROCKS
THERMAL MANIFESTATIONS
HYDROLOGIC
SURFACE DISCHARGE OF GEOFLUIDS
TEMPERATURE OF FLUIDS
CHEMICAL COMPOSITION OF FLUIDS
GROUNDWATER HYDROLOGY
METEOROLOGY
GEOCHEMICAL
CHLORIDE CONCENTRATION
Si02 CONTENT
NA-K-CA RATIOS
ISOTOPIC COMPOSITION OF HYDROGEN AND OXYGEN
GEOPHYSICAL
GEOTHERMAL GRADIENT
HEAT FLOW
ELECTRICAL CONDUCTIVITY
SEISMIC ACTIVITY
EXPLORATION HOLE DRILLING
RESERVOIR CHARACTERISTICS
TEMPERATURE-DEPTH PROFILE
PRESSURE-DEPTH PROFILE
LlTHOLOGY AND STRATIGRAPHY
PERMEABILITY LOG
POROSITY LOG
FLUID COMPOSITION
11
-------
techniques available for surface exploration of potential resources and
exploratory drilling of potential reservoirs. Improvements and novel
methods for reducing costs in these two initial phases of the geothermal
energy cycle are thus of great importance. In order to achieve the goal
of providing an equivalent of one million barrels of oil per day by geo-
thermal resources, it is evident that exploration for geothermal re-
sources, especially hydrothermal, must receive a very high priority by
the U.S. energy resource industry.
The magnitude of hydrothermal resources required can be estimated
from the following calculation for a 100 MWe generating plant operating
with flashed steam of 555 kcal/kg (1000 Btu/lb) heat content. The re-
quired geofluid production rate for a hot-water system yielding 107o
steam on flashing with a thermal efficiency of 20 percent, would be
7.75 x 106 kg/h (1.7 x 107 Ib/hr). The amortization of the 100 MWe
plant over a period of thirty years would require a total production of
12
2.1 x 10 kg hot water, and a mean reservoir porosity of 10 percent
3
would require a geothermal reservoir volume of about 2 km . At a 50
percent condensation efficiency, the plant would discharge a hot water
3 7
supply of about 100,000 m /d (2.5 x 10 gpd).
For a national capacity of 20,000 MWe, these values are multiplied
by a factor of 200. Thus reservoirs supporting 200 units of 100 MWe
generating plants must be located. These reservoirs will produce about
Q
1.5 x 10 kg/h of hot water. For a mean well production flow rate of
250,000 kg/h a total of 6,000 production wells will be needed, and for
2
a mean spacing of 100,000 m /well (25 acres/well), a total reservoir
82 5
area of 6 x 10 m (1.5 x 10 acres) of geothermal resources must be
12
-------
found. It is evident that if hydrothermal systems are to provide the
nation with 20,000 MWe, very high priority for resource exploration and
assessment is indeed required.
UTILIZATION TECHNOLOGY
Utilization of geothermal energy varies with the quality of avail-
able resources. It has been noted that the present geothermal industry
has focused on high quality hydrothermal resources. Extraction and con-
version technologies for dry-steam reservoirs are sufficiently advanced
to be commercially attractive. Conversion technologies for hot-water
resources are more complex, and for hot brines, geopressured basins, and
hot dry rock formations, they are even more complex; commercial utiliza-
tion is still further away. Since these latter types of resource hold
great promise, technology to exploit them must be developed.
Stimulation of geothermal energy production can be achieved by re-
search and development to (1) increase the modes of resource utilization,
(2) improve energy conversion technology, and (3) provide advanced
methods of energy extraction. Increased efficiency in each of these
three aspects of the geothermal energy cycle is attainable.
Development of a geothermal field generally involves the geofluid
characteristics, steam separation and gathering facilities, turbine and
generator equipment, cooling systems, and condensate disposal methods.
Such development presupposes that electric power generation is the sole
purpose of the field development. It may turn out, however, that for
many geothermal reservoirs, non-electric utilization of the resource may
make the reservoir economically feasible, with significant conservation
of fossil and nuclear fuels. Several modes of utilization of geothermal
13
-------
resources are listed in Table 5. Hydrothermal fluids with temperature
or enthalpy too low for economic electric power production may be use-
ful for water or mineral sources and for industrial, agricultural, and
municipal heating. However, since major interest in geothermal energy
is for the production of electric power, combined or total utilization
may help make many geothermal reservoirs submarginal in power production
alone become economically feasible. The possibility of building a com-
munity around a geothermal resource, with municipal heating, an industrial
park of process firms requiring hot water and concomitant electric power
production appears feasible. Thus, research for methods stimulating
geothermal resource utilization in all forms is well warranted.
General methods for producing electricity from geothermal fluids are
summarized in Table 6 and are described adequately in the several cited
references. The choice of a conversion cycle is generally dependent on
the thermodynamic and chemical properties of the geofluid. Present com-
mercial plants utilize low-salinity hydrothermal systems with steam or
water at temperatures above about 200°C in the single-stage direct steam
turbine conversion system. To utilize lower temperature fluids, inves-
tigations are underway to develop other conversion systems; among these
are multiple-flash low-pressure steam turbines, single and multiple
stage binary cycle systems, and hybrid systems combining these two.
The binary system appears to be the most promising for utilization of
geofluids with temperatures between 100°C and 200°C. However only one
experimental facility, the Pauzhetka station in the Kamchatka peninsula
of the USSR, has been constructed to date. The binary system most
likely to be successful in the U.S. will require a downhole pump to
prevent flashing, a heat exchanger which can operate without excessive
-------
TABLE 5
UTILIZATION OF GEOTHERMAL ENERGY
ELECTRIC POWER PRODUCTION
DIRECT USE OF DRY STEAM
FLASHING OF HOT WATER TO STEAM
SURFACE FLASHING
IN-SITU FLASHING
BINARY AND HYBRID CYCLES
INNOVATIVE SINGLE-WELL CONVERTERS
DIRECT USE OF THERMAL WATERS
AGRICULTURE
AQUICULTURE
SPACE HEATING
INDUSTRIAL PROCESSING
MEDICAL THERAPY
BYPRODUCTS
MINERAL EXTRACTION
WATER RESOURCES
15
-------
TABLE 6
TYPES OF GEOTHERMAL POWER PLANTS
HEAT SOURCE
DRY STEAM
HOT WATER (T >180°C)
HOT WATER (T <150°C)
HOT WATER
(MODERATE SALINITY)
HOT BRINE (PRESSURIZED)
HOT BRINE (FLASHED)
GENERATION MODE
STEAM TURBINE
STEAM TURBINE
BINARY CYCLE
HYBRID CYCLE
BINARY CYCLE
IMPACT TURBINE
HELICAL SCREW EXPANDER
BLADELESS TURBINE
16
-------
corrosion and deposition, and & circulation system which allows for re-
injection of the geofluids for environmental control purposes.
Several types of downhole pumps are under development, involving
design concepts which use (a) in-situ heat to operate a closed steam-
generator-turbine to drive the pump, (b) a high-speed, high-temperature
high length-to-diameter electric-motor driven pump, or (c) a hydraul-
ically driven unit with hydraulic power from the surface. Heat exchanger
concepts include fluidized sand beds to enhance heat transfer rate and
maintain clean surface, and liquid-liquid systems with direct contact
of immiscible fluids, tray-tower contactors, or subcritical or super-
critical power cycles.
Flash and binary systems are useful in large power plants having
capacity in excess of 50 MWe. They require complexes of multiple-well
field development and extensive networks of gathering lines. Innovative
conversion systems are under development in which small power plants, in
sizes of 1 to 15 MWe, may be installed at individual wells. These sys-
tems may involve a total flow concept in which both the thermal and
kinetic energy of the geofluid is used for production of electricity.
One of these is the impulse turbine, in which the thermal energy is
converted to kinetic energy through a converging-diverging nozzle, and
the high-velocity output drives a hydraulic impulse turbine operated at
low pack pressure. Calculations indicate that a large unit (e.g.,
220 MWe) might be feasible for the Salton Sea geothermal brines, which
contain as much as 230,000 ppm total dissolved solids. The material
handling problems of such brines are indeed enormous, but the dissolved
solids may also represent a source of valuable minerals, such as lead,
manganese, and copper, if they can be processed economically.
17
-------
Another total-flow concept is the helical rotary screw expander
which expands the vapor from hot saturated liquids by continuous pres-
sure reduction in the expanding screw, in essence creating an infinite
series of flashing stages. A small 62.5 kV prototype model was tested
successfully with moderate salinity geofluids with indications that it
can accept the total flow of untreated brines. Still another concept is
the bladeless turbine, in which a series of closely-spaced disks are
rotated by viscous drag exerted by geofluids introduced by a nozzle.
The device seems simple and self-cleaning, but the overall efficiency
may be small.
Increased extraction efficiency represents a major means to stimu-
late geothermal energy production, especially for non-hydrothermal reser-
voir systems. Calculations show that hot-water reservoirs contain a
larger amount of available energy than steam-filled reservoirs under the
same reservoir conditions because of the much larger mass of water; but
in either system, the heat contained in the rock formation is much
larger than the heat in the fluids. Thus recovery of the formation
heat would be of major economic significance. Extraction of formation
heat must be a non-isothermal process, which can be achieved either by
flashing geothermal liquids to steam within the formation or by recycl-
ing colder fluids back into the formation. Laboratory investigations
and theoretical calculations of reservoir models are underway to deter-
mine the extent of heat extraction from fractured reservoir formations.
The natural extraction efficiency of energy from impermeable hot
dry rock formations is extremely small. And yet hot dry rock in the up-
per 10 km of the earth's crust represents a major potential resource of
18
-------
geothermal energy. The volumetric energy extractable from hot dry rock,
calculated for average expected properties and possible technical extrac-
9 3
tion efficients, is of the order of 1.2 x 10 kWh/km of fractured rock,
3
equivalent to a volumetric power extraction of 1.4 MW/km for one century.
The technical challenge is the ability to fracture such volumes of hot
rock massives and achieve an extraction efficiency of the order of
10 percent.
Fracture stimulation methods are useful for many types of geothermal
reservoirs. In vapor-dominated systems, stimulation may restore declin-
ing pressure or connect dry holes in commercial steam fields to producing
sections. In liquid-dominated systems lacking sufficient productivity
for economic power generation, fracture stimulation may provide larger
wellbore diameter for increased flow rate, greater surface area for heat
transfer, or restore porosity or permeability around wells having depos-
ited silica, calcite, or other precipitated minerals. In dry geothermal
systems, stimulation is needed to provide large fracture volumes for
heat transfer to an artificial convective extraction system.
Several fracturing methods are under study; these include hydraulic
fracturing, thermal stressing, and chemical and nuclear explosive frac-
turing. Hydraulic and explosive fracturing methods have already proven
successful in stimulation of natural gas reservoirs.
Experiments to evaluate the potential for hydraulic and thermal
stress fracturing for recovery of geothermal energy from hot dry rock
formations are underway. In this concept a large diameter vertical
crack is created hydraulically at the bottom of a boreholde in the geo-
thermal formation. A second hole is drilled to intersect the upper part
-------
of the fracture, and a pump is used to initiate artificial heat-
extraction circulation. It is hoped that pumping can be discontinued
if a natural convective circulation is achieved. The major technical
problems are the attainment of. a vertical crack of about 2 km diameter
with sufficient fracture area, the creation of additional fracture
area by thermal stress of cold water injection, and the ability to
achieve a natural convective circulation without undue losses of
water, especially in arid regions. Calculations indicate that under
favorable conditions, the system might provide an average power of
about 100 MW (thermal) for twenty years.
INSTITUTIONAL ASPECTS
Although much remains to be done in locating adequate reserves
and developing adequate technology to meet the goals for exploiting
the Nation's geothermal resources, there is great confidence that these
will be achieved. These problems involve advances in physical research
and technology. Institutional problems however, also exist. Such
problems are complex; they involve public acceptance, vested interests,
historical precedents, existing regulations from other resources, over-
lapping jurisdictions, and economic and financial factors. These prob-
lems are often more difficult to resolve than are engineering problems,
and they may in the long run be the major constraints to an accelerated,
but orderly development of geothermal resources. The solutions to many
institutional problems may require broad public interaction, changes in
regulations and legislation, and perhaps changes in traditional invest-
ment and marketing procedures.
20
-------
Economic factors affect all forms of energy supply; they involve
total capital costs per installed power unit and operational costs per
unit of energy production. For geothermal energy, both of these cost
factors are strongly dependent on the specific characteristics of indi-
vidual reservoirs and the size of the installed power plant units.
Important capital costs include the investments for exploration, drill-
ing and completion of wells, gathering lines and waste handling systems
for all utilizations. For thermal energy applications, they also in-
clude the distribution system, and for electric power production, they
include the power plants and the transmission network. The production
costs are influenced by the cost of capital, operations and maintenance,
and plant utilization factor. In the United States, additional costs
must also be added for environmental pollution control.
Factual cost data for geothermal electric power production in the
United States are available only for the Geysers field. The electric
utility purchases steam from only one supplier, but has negotiated to
purchase steam for future plants from additional suppliers. In the
development of future geothermal power stations, an option exists for
an integrated operation from exploration to power production in contrast
to the traditional roles of an electric utility purchasing steam or hot
water from an independent supplier. The general effect would be an in-
creased investment cost per kilowatt hour of energy.
Data for costs of recently-construeted power plant units at the
Geysers are sparse, but estimates for the original plants range from
about $100 to $150 per kW. Production costs were estimated at about
7 mill/kWh of which about 3.5 mill/kWh was the price of the purchased
steam. These estimates included a cost of 0.5 mill/kWh for injection
21
-------
of water condensate as a disposal method. With escalation of drilling,
construction, and environmental reporting costs over the past few years,
these cost values are not useful for estimating costs of new facilities,
especially for reservoirs which do not produce dry steam. Recent esti-
mates indicate installation costs may range from $500 to $700 per kW
and operating costs for binary conversion systems of the order of 20 to
40 mill/kWh. These costs, of course, are hypothetical, and more precise
costs will be generated as other major reservoirs and plants are devel-
oped and operated. A large uncertainty in the total cost is the fixed
exploration cost for the resource, which is independent of plant capac-
ity, and the average drilling costs of the production,' dry, and injection
wells. Computer models to evaluate the relative importance of these re-
source and utilization costs are under development.
Because of large uncertainties in the technical costs of explora-
tion and drilling, conversion efficients, and stimulation techniques,
and because of the rapid escalation rate of these costs, it is difficult
not only to estimate costs on an absolute basis, but even to compare
costs of other forms of electric power generation. Besides the costs
affected by these technical factors, other factors more social in
nature must be considered. Among these are public acceptance and govern-
ment stimulus for accelerating the development of geothermal energy in
relation to other energy sources, the interpretation of compliance with
the National Environmental Protection Act of 1969, and the availability
of investment capital for development of geothermal resources and elec-
tric and thermal power plants. These socio-economic factors may require
much public and government deliberation before general philosophies are
widely accepted in practice.
22
-------
Although geothermal energy is considered to be one of the least pol-
luting of the many forms of energy available, it should be assumed that
the public will insist that the environmental impact of producing geo-
thermal energy in all of its natural and stimulated forms, be thoroughly
investigated in accordance with NEPA and any additional requirements
under state and local legislation. Furthermore, in addition to environ-
mental impact, it is also evident that assessment will be required of
the operational aspects of the various types of resources which affect
personnel safety and plant maintenance.
In the evaluation of a benefit-risk analysis, geothermal energy is
expected to compare favorably with respect to other energy resources,
especially when viewed over the entire fuel cycle. Since geothermal
energy must be utilized or converted in the vicinity of the resource,
the entire "fuel cycle" from reservoir to transmission is located at one
site. This is in contrast with material fuels in which the cycle in-
volves mining, storage, refining, transportation, reprocessing, and
waste disposal, many or all of these at different locations. Further-
more, increased utilization of geothermal energy may result in a cor-
respondingly reduced demand for material fuels in short supply, such as
natural gas, oil, coal, and uranium. And still further, geothermal
fluids may provide byproduct sources of water with reduced demand for
cooling water.
Geothermal energy, nevertheless, has its array of potentially
deleterious environmental impacts. A list of potential environmental
impacts is given in Table 7. A review of the more important ones has
recently been completed in a workshop sponsored by the National Science
23
-------
b
CO CD
« CO i
LU < _l
J- CJ3 _J O
00 o: 2 2
° * o £
LU O Z I
Q 5 Q
LU
CO
a
3
CO
LU
oo
o =~r
LU a:
co LU
LU
cr:
CO
LU
CO
co e>
LU O
C3 _J
o:
o
LJ
o:
o
LU
fe
PO
UJ
co
Q
CO ^J
25
CO Q
3 e
Q- H
OO
LU
CO
o
o z
Q_ (_3
? H
S [if ^
8 IS
O LU p|
CO
£
co
(I
en
LO
00
O
CO
O
- O
is: o_
CO
2 co
LU LU
fsl
Q- CO
° d
O LU
I
cS
o: a.
CO
LU
i
OJ
UJ CO
~) CL
CO
LU
o
S £
gat
U_ CO eC
CO '
LU h-
O
LU
00
-------
Foundation (see [7] in Bibliography) as the basis for a program to sup-
port research for baseline data and technology for monitoring potential
impacts and controlling actual hazards. The major impacts include
gaseous emissions, liquid waste disposal, and geophysical effects such
as seismicity and subsidence. Other concerns involve thermal releases,
surface water contamination, land use planning, cooling water consump-
tion, and visual and noise pollution.
An array of legal problems associated with geothermal resource
development also exists. These have been reviewed in another workshop
sponsored by the National Science Foundation (see [8] in Bibliography).
The legal problems of geothermal resources begin with resource
definition, which varies from state to state. For example, in Cali-
fornia geothermal resources are defined as "the natural heat of the
earth, the energy--which may be extracted from naturally heated fluids--
but excluding oil, hydrocarbon gas or other hydrocarbon substances."
This definition leaves open the question whether geothermal resources
are legally defined as water, mineral, or gas resources, and results in
large uncertainty with respect to Federal, state, and local jurisdictions.
On the other hand, the State of Hawaii considers geothermal resources as
minerals, whereas the State of Wyoming has declared them water resources.
As water resources, they would be subject to the very complicated set of
state laws concerning water rights and regulation. As minerals, they
would be subject to mining laws and such problems as ownership, depletion
allowances, and write-off of intangible drilling costs. Geothermal re-
sources have already been classified in court decisions in different
ways. In one case a U.S. District Court in San Francisco treated the
25
-------
Geysers geothermal resource as "nothing more than superheated water" and
therefore not a mineral, but in another case, the resource was held to be
a gas within the meaning of the Internal Revenue Code provisions for de-
pletion allowance and intangible drilling costs.
Ownership rights is also a serious institutional problem. The Fed-
eral government has given some 35 million acres of land to the home-
steaders, States, and railroads, but generally reserved the mineral
rights to the Federal government. However some State grants included
mineral rights and thus many problems exist in the ownership aspects of
Federal and State lands under the leasing of these lands for geothermal
energy development. Land utilization for geothermal resources also
comes under the jurisdiction of local governments, except for resources
on State or Federal lands.
Other institutional questions at the State level include the acreage
level for commercial development, the need for long-range financial and
land use planning, and the overlapping of State regulatory agencies with
each other and with jurisdictions of local governments for permits, li-
cences, taxation, and especially environmental control. The latter may
be affected at the Federal, State, regional, county, or city government
levels. For example, in some areas, authority may be divided between
such agencies as a Regional Land Development Commission and a County Air
Pollution Control Board.
The institutional aspects of licensing and regulation of power
plants is very complicated; they cover the spectrum from Federal to local
jurisdictions. Regulations already exist with respect to the exploration,
drilling and operation of water and mineral wells in all states. The
26
-------
extension to geothermal wells should be relatively simple. Yet the need
to satisfy the provisions of NEPA and any specific State environmental
requirements may make geothermal resource development a slow process.
For example, in California, the State Lands Commission, before it can
lease any lands under its jurisdictions, must make a finding at a public
meeting that the lease will not have significant detrimental environ-
mental effect and must prepare an environmental impact report available
to the legislature and the public. The corresponding problems of en-
vironmental impact from geothermal resources in private lands are not
yet fully resolved.
Once the field is developed to the point where a utility contracts
to purchase the resource and construct a power plant, other regulatory
agencies come into the picture, such as the Federal Power Commission and
corresponding state and local agencies. Site selection and environ-
mental analysis criteria are becoming of major importance in power plant
licensing for all types of energy resources and their effect on geo-
thermal energy development will probably be determined by solution of
these problems on a generic basis, rather than specifically for geo-
thermal energy alone.
Institutional problems thus involve many social, legal, environmen-
tal, and economic questions. The problems become more complex for land
use planning when geothermal resources span Federal, state, and private
lands. They involve capital investment problems for geothermal develop-
ment which may be considered to be high-risk and involve long-delay times
until they become income producing. They involve inter-industry arrange-
ments when multi or total utilization is needed to support economic
27
-------
development of electric power generation, thermal power heating, desal-
ination and mineral recovery. And they involve multi-government arrange-
ments in the realms of regulation, licensing, and environmental control.
NATIONAL.GEOTHERMAL PROGRAM
Although significant growth of the one natural steam field in the
United States has occurred since 1960, it has become apparent that a
major national effort of industrial development supported by Federal
stimulation is needed to develop the potential of geothermal resources
in its several forms as an alternate energy source. Early efforts to
achieve a coordinated Federal program for the support of research and
development were undertaken by an informal Interagency Panel for Geo-
thermal Energy Research. From these efforts evolved a 5-year program
whose objective was the rapid development of a viable geothermal in-
dustry for the utilization of geothermal resources for electric power
production and other products. The goals and plans for this program
were prepared by the Interagency Task Force on Geothermal Energy under
direction of the National Science Foundation in the Federal Energy Ad-
ministration "Project Independence Blueprint" (see [9] in Bibliography).
The task force evaluated two alternate strategies. The first was
"business-as-usual" which assumed continuation of current policies af-
fecting levels of geothermal production. The second was "accelerated
demand" which assumed specific changes that would result in a more rapid
expansion of potential production.
The task force estimated that under the "business-as-usual" assump-
tions, electric power capacity could reach 4000 MWe by 1985 and perhaps
59,000 MWe by 1990. The corresponding numbers for the "accelerated
28
-------
demand" assumptions were 20,000 to 30,000 MWe by 1985 and 100,000 MWe by
1990. These latter values were adopted as the primary goal of a pro-
posed National Geothermal Energy Research Program, which was directed
towards (1) providing the necessary technological advances to improve
the economics of geothermal power production, (2) expanding the knowledge
of recoverable resources of geothermal energy, and (3) providing care-
fully researched policy options to assist in resolving environmental,
legal, and institutional problems.
The major research funding agencies which contributed to the task
force program were the Atomic Energy Commission, the Department of the
Interior, and the National Science Foundation which served as lead
Federal Agency. The status of the research carried out under support
from these agencies is described in the proceedings of a conference on
research for the development of geothermal energy resources (see [10]
in Bibliography).
During 1974, two acts of Congress resulted in a marked change in
direction for the national development of geothermal energy. The first
was PL 93-410, the Geothermal Energy Research, Development, and Demon-
stration Act of 1974, which established a Geothermal Energy Coordination
and Management Project. The Project was given responsibility for the
management and coordination of a national geothermal development program
which included efforts to: (1) determine and evaluate the geothermal
resources of the United States; (2) support the necessary research and
development for exploration, extraction, and utilization technologies;
(3) provide demonstration of appropriate technologies; and (4) organize
and implement the loan guarantee program authorized in Title II of the
Act.
29
-------
The second law was PL 93-438, the Energy Reorganization Act of 1974,
which established the Energy Research and Development Administration,
ERDA, with responsibility as lead Federal agency for activities related
to R&D of all energy sources. The Act abolished the AEC and trans-
ferred the geothermal development function of the AEC and NSF to ERDA.
On January 19, 1975, ERDA assumed responsibility for the national pro-
gram of geothermal energy development. It has also assumed direction of
the Geothermal Energy Coordination and Management Project which has com-
pleted the Final Report required by PL 93-410 (see [11]). In addition,
ERDA, in response to Congressional requirements and internal needs, pre-
pared a comprehensive R,D&D plan (see [12]) for developing energy tech-
nology options. The geothermal section of the plan built upon the
predecessor plans of the Task Force for Geothermal Energy and the Geo-
thermal Project and has based the goal for the national program on the
rational given in Volume 2 of the Plan.
The objectives being considered in the ERDA program for geothermal
energy include methods to stimulate the industrial development of indige-
nous hydrothermal resources to provide the Nation with 10,000 to 15,000
MW of electric power and thermal energy during the 1985 to 1990 period
and to develop new and improved technologies for cost-effective and
environmentally-acceptable utilization of all types of geothermal re-
sources as a long-term alternate source of energy.
The strategy of the program which might accomplish such objectives
would be to accelerate industrial development of the nation's geothermal
resources by (1) coordinating efforts for exploration and assessment
of geothermal resources necessary to establish reserves by 1978-1980
which can support production of 20,000 to 30,000 MW of power,
30
-------
(2) demonstrating near-term and advanced technologies needed to utilize
many types of geothermal resources in a cost-effective and environmentally-
acceptable manner, and (3) fostering rapid development of a viable geo-
thermal industry by appropriate incentives, timely reduction of institu-
tional impediments, and direct participation of the private sector in
development and demonstration of geothermal energy technology.
Although EKDA assumes overall responsibility for effective manage-
ment and coordination of Federal geothermal activities, the scope of the
Federal program includes the efforts of many Federal agencies. The
Geothermal Steam Act of 1970 authorized the Department of the Interior to
lease Federal lands for geothermal resource exploration, development, and
production of energy and useful byproducts (such as methane, desalinated
water, and valuable minerals). The leasing program is conducted by the
Bureau of Land Management which is responsible for selecting lands for
lease and holding lease sales and the U.S. Geological Survey which
classifies the lands by appraised value.
The U.S. Geological Survey's geothermal research program is focused
on the characterization and description of the nature and extent of the
geothermal resources of the United States. The output of the U.S. Geo-
logical Survey's program is the determination of the magnitude of the
geothermal resource base on a national and regional basis. The Survey's
program includes development of exploration technology, methodology for
estimating energy potential of geothermal systems, environmental effects
of geofluid withdrawal, and geochemical aspects of reservoir permeability.
Some of the problems that have retarded the delineation of the
Nation's geothermal resources through the leasing of public lands include
31
-------
the lack of reliable information regarding suitable resources, even on
lands classified as KGRAs, insufficient requirements for early explora-
tion and development of leased lands, and legal problems involving owner-
ship and control of use of geothermal resources. Under a national
program, coordinated effort by ERDA and the Department of the Interior
would help to accelerate the establishment of geothermal reserves by
the resource industries. Potential actions include (1) accelerated
estimation by the U.S. Geological Survey of the available resources by
geologic type, (2) improved technology for resource exploration and
assessment and for reservoir evaluation, (3) easing of leasing impedi-
ments by better methods for designating KGRAs and establishing minimum
acceptable bids, (4) incentives for early development of leased lands,
and (5) recommendations for legislation to resolve legal uncertainties
pertaining to geothermal resources.
The second part of the strategy for the Federal program would cen-
ter on ERDA efforts for demonstration of near-term and advanced systems
for resource utilization, development of supporting research and tech-
nology, and execution of the Federal loan guarantee program. Demonstra-
tions of utilization technology could occur as (1) commercial-scale
demonstration plants to provide the public sector with operational ex-
perience with full-scale electric power plants capable of generating
energy at design production cost under pertinent environmental and
institutional conditions, (2) pilot-plant facilities to prove technical
feasibility, provide preliminary economic data, and provide capability
for testing new and improved extraction and conversion systems for
electric power production, and (3) field test facilities to improve
32
-------
reservoir assessment technology, evaluate reservoir characteristics and
performance, test and evaluate energy extraction and conversion components
and processes, evaluate material compatibility with geothermal fluids,
and test environmental control technologies. A supporting research and
development program could provide for development of hardware systems,
components, processes, and control techniques for installation in the
demonstration facilities and field testing of reservoir evaluation tech-
nologies for the range of resource types. Supporting research and devel-
opment program could also provide advanced research and technology to the
geothermal industry and its supplier and support industries for improved
productivity and utilization.
Implementation of the Loan Guarantee program should be coordinated
with the Bureau of Land Management's geothermal leasing program and
ERDA's research, development, and demonstration program. The program
could involve venture capital companies, reservoir developers, and lease
holders to maximize the impact of the loan program in stimulating early
development of commercial electric and thermal power facilities. The
Loan Guarantee program might be used primarily for income-producing
projects, such as field development and power-plant construction.
Smaller industrial firms could benefit from guaranteed loans by gaining
access to necessary private capital. Regulations and procedures govern-
ing the implementation of the loan guarantee program are currently being
drafted in coordination with other Federal agencies, such as the Small
Business Administration and the Economic Development Administration.
Approved regulations and operating procedures setting forth specific
information requirements to be met by the applicant and criteria govern-
ing the approval process should be widely publicized as early as possible.
33
-------
The third part of the strategy of the Federal program would involve
several Federal agencies, notably the National Science Foundation, in-
volved in assessing environmental, legal, and institutional problems of
advanced energy technology under its RANN program, the Environmental
Protection Agency, involved in .environmental emission standards, moni-
toring, and control technologies, and the Federal Energy Administration,
involved in institutional aspects of the national energy situation.
With the mutual efforts of the Federal program and the geothermal
industry, the attainment of the National electric and thermal energy
goals for geothermal resources could add a significant alternate energy
source to the national economy before the end of the present century.
-------
BIBLIOGRAPHY
1. Proceedings, United Nations Symposium on the Development and Utiliza-
tion of Geothermal Resources, Geothermics, Special Issue No. 2,
2 volumes, 1970.
2. H. C. H. Armstead, ed., "Geothermal Energy: Review of Research and
Development," Earth Science Series No. 12 (UNESCO, Paris, 1973).
3. P. Kruger and C. Otte, eds., "Geothermal Energy: Resources, Pro-
duction, Stimulation" (Stanford University Press, Stanford, 1973).
4. Scientific Council for Geothermal Energy, USSR Academy of Sciences,
"Geothermal Investigation and Utilization of the Heat of the Earth"
(Nauka, Moscow, 1966).
5. Scientific Council for Geothermal Energy, USSR Academy of Sciences,
"Study and Utilization of the Deep Heat of the Earth" (Nauka,
Moscow, 1973).
6. Informatics, Inc., Recent Soviet Investigations in Geothermy
ARPA-1622-3, May, 1972, and Soviet Geothermal Electric Power Engi-
neering, ARPA-1622-3, December, 1972.
7. Proceedings, Workshop on Environmental Aspects of Geothermal Re-
sources Development, National Science Foundation Report No. AER
75-06872, 1974.
8. Proceedings, Conference on Geothermal Energy and the Law, National
Science Foundation Report No. NSF-RA-S-75-003, 1975.
9. Federal Energy Administration, Project Independence Blueprint,
Final Task Force Report, Geothermal Energy, November, 1974.
10. Proceedings, Conference on Research for the Development of Geo-
thermal Energy Resources, National Science Foundation Report No.
NSF-RA-N-74-159, 1974.
11. Energy Research and Development Administration, Definition Report:
Geothermal Energy Research, Development and Demonstration Program,
ERDA-86, October, 1975.
12. Energy Research and Development Administration, A National Plan for
Energy Research, Development and Demonstration: Creating Energy
Choices for the Future, ERDA-48, 2 volumes, June, 1975.
35
-------
GEOTHERMAL EFFLUENTS, THEIR TOXICITY AND PRIORITIZATION
by
Leroy Schieler
Woodward-Clyde Consultants
San Francisco, California
Exploitation of geothermal resources is hampered by a lack of under-
standing of the basic chemical interactions and toxicity hazards associ-
ated with the gaseous and aqueous effluents characteristic of various
geothermal areas. The objective of this paper is the evaluation of the
relative hazards of the various effluents in both liquid- and vapor-
dominated fields in terms of concentration and chemical toxicity effects.
Many chemical species must be considered in evaluating liquid-dominated
systems, but only a few volatile species are significant in vapor-dominated
.systems. This is a direct consequence of the thermodynamic equilibria
established among the chemical elements over the wide range of temperatures
>
encountered in various geothermal areas. Vapor-dominated systems are
lacking the aqueous phase required for dissolving water soluble salts.
Since water is present as steam rather than as liquid, vapor-dominated
systems tend to be lower in total dissolved solids than liquid-dominated
systems. Volatility is the primary transport mechanism in vapor-dominated
systems.
As might be anticipated from the preceding discussion, gas phase
effluents are present in higher concentrations in vapor-dominated systems
than in liquid-dominated systems. In the predominately vapor-dominated
36
-------
systems, such as The Geysers and Lardello, over 98 percent of the gas is
steam. The remaining 2 percent is composed of gases including primarily
hydrogen sulfide, carbon dioxide, methane, ethane, nitrogen, and hydrogen.
Since water is a condensible gas at ambient temperatures, it is not
normally listed although it is initially .present in the gas phase.
Typically only the noncondensible gases are listed as in Table 1.
Data for The Geysers are typical of the vapor-dominated systems
which account for approximately 5 percent of the known geothermal
resources. Wairakei is typical of the majority of the remaining 95
percent of the geothermal areas. Although the absolute concentrations
of the noncondensible gases are quite different for the vapor- and
liquid-dominated systems, the ratios are generally similar for those
major species which are involved in the same chemical equilibrium
systems. Carbon dioxide and water react to form carbonic acid which, in
turn, can liberate hydrogen sulfide from metal sulfides. The complex
chemical equilibria involving these and other species are temperature
dependent and can be expected to be linear only over a narrow tempera-
ture range.
Ammonia, nitrogen, and hydrogen are involved in another chemical
equilibrium reaction system by virtue of the following thermal dissociation:
2NH ^=^ N_ + 3H
O £ £
Hydrogen is also contributed by the dissociation of water.
37
-------
^2
OH
Ci
UJ
^c
(X
jjj
^^
5?
2
s
o
s
§
P-,
co
UJ
^?
<:
o
w
1-3
CO
l_l
CO
z
p
^4
8
z
o
^
i~4
fn
0)
X
0)
e?
0)
<
X
rt
us
c
g
«£«
o
1
w
rt
13
O O LO rH 00 IO O
^t O rH
VD
CM O t ** CM VO
CM VO Ot O IO IO
CM CM rH 1 rH
to
O O t-» O> O 00 to
O O "* rH VO IO rH
vO VO Tj- O VO CM
\ «s * *\
rH O rH' rH
to
t
10 O to to CT> VO rH
C7l rH t4
CM
f~\
/ \
f^l
O 01 rH IO
rH CJ rH 1 1' 1 CM 1 1
tO rH
o o
o o
V^* IM*
vO
co CM n- ac to
rH
*
0
CM
1
10
rH
0)
S
5
~
M
G
O
P
W
0
CQ
^
3
(X,
<
tW
O
bo 10
C £
H CT>
J..1 . 1
+* ri
5 3
rH CM
-------
Thus, ammonia and nitrogen tend to occur at a ratio of about 10 to 1.
Hydrogen ratios appear to be variable since hydrogen is contributed by
two major sources, water and ammonia.
These constant ratio characteristics provide a valuable generalization
for estimating the relative toxicity hazards of specific geothermal projects
or areas on the basis of limited data.
The two noncondensible gases listed in Table 1 which present the
greatest potential hazard are hydrogen sulfide and ammonia. Both exceed
the maximum allowable concentrations by factors which cannot be ignored.
The relative hazard of radon 222 and associated daughter products will be
considered later. The global greenhouse effect associated with the
evolution of carbon dioxide will not be considered because the magnitude
is small compared with combustion sources.
Ammonia is primarily an upper respiratory poison. Inhalation of 1000
ppm causes irritation of the eyes and upper respiratory tract with coughing,
vomiting, and redness of the mucous membranes of the mouth, nose, lips,
and pharynx. Higher concentrations cause swelling of the lips and eyes,
temporary blindness, restlessness, tightness in the chest, frothy sputum
indicating pulmonary edema, and weak, rapid pulse. In geothermal operations
ammonia is not likely to present a direct toxicity hazard except possibly
in the immediate vicinity of the power plant. Ammonia toxicity risk is
always small compared to that of hydrogen sulfide since it occurs at a
much lower concentration.. Atmospheric dilution would reduce the ammonia
effluent levels to acceptable values very rapidly.
39
-------
On the other hand, ammonia docs present indirect hazards which have
not received much attention to date in the evaluation of geothermal
hazards. Ammonia is a base which reacts with acids such as hydrogen
fluoride, sulfur dioxide, and hydrogen sulfide to form the corresponding
salt. The resultant particulates are lower respiratory poisons which
are generally more toxic than the original reactants and which, in
addition, may settle out in given geographical areas depending upon
local meteorological conditions. For example, ammonium sulfate par-
ticulates are approximately 5 times as toxic as sulfur dioxide. This
is the basis for current concern and the basis for consideration of
possible standards for sulfates. As is the case with many air pollutants,
there is insufficient available evidence to prove that air pollution per
se produces disease, but there are many indications that air pollution can
aggrevate symptoms of pre-existing disease which may then prove fatal.
Human beings with cardiovascular or respiratory disease appear to be
particularly vulnerable. It is believed, however, that a particulate such
as ammonium sulfate, not only has an adverse effect by itself, but is
even more toxic when inhaled along with other common air pollutants. This
synergistic effect of particulates formed from ammonia effluents may be a
valid factor to consider in complying with requirements of the Significant
Deterioration Act.
In addition to the human health hazards, the environmental impact
on plant and animal species sensitive to changes in ammonium ion con-
centration and pH are likely to be significant. It has been postulated,
-------
for example, that the surface waters in some gcotherraal areas are neutral
whereas equivalent waters in adjacent areas are slightly acid because of
ammonia liberated from gcothermal operations. Continuous operation could
result in a shift in species distribution. Intermittent operation could
result in a periodic change in pH which might have more significant
environmental impacts.
The most noticeable geothermal effluent is hydrogen sulfide. The gas,
which has a characteristic rotten egg odor even at very low concentrations,
is known to cause irritation to the eyes and respiratory tract as well as
deleterious effects to the nervous system of humans exposed to it (Ref.
1,2). In the immediate vicinity of the geothermal area, the hydrogen
sulfide concentration far exceeds the maximum allowable concentration
levels set by OSHA', federal, and state standards. The ambient air levels
arc high enough to fall within the concentration range where serious
health effects and death have been well documented. Atmospheric dilution
and possibly efficient scrubbing systems must be depended upon to reduce
the ambient air levels to acceptable levels. Since geothermal operations
are conducted in remote areas, adequate atmospheric dilution is easily
achieved. On the other hand, atmospheric hydrogen sulfide emissions do
pose a significant industrial hygiene hazard to workers in the immediate
area. The fact that the OSHA limit of 20 ppm is exceeded by orders of
magnitude cannot realistically be ignored. Hydrogen sulfide also has a
major environmental impact on water quality as will be discussed later.
-------
Hydrogen sulfido is much more toxic than is commonly realized.
More concern is directed toward the odor nuisance than the health hazard.
The human health and physiological effects of exposure to varying concentra-
tions of hydrogen sulfide are presented in Table 2. As can be seen, the
maximum concentration of 1600 ppm reported for The Geysers could result
in death after only a brief exposure. The average value of 222 ppm
reported for The Geysers is clearly within the range of serious health
effects. Even the relatively lower hydrogen sulfide levels characteristic
of liquid-dominated geothermal areas are well within the .range of potential
health hazard effects. Compliance with the 0.03 ppm limit set by
California and other states will be difficult except at large distances
from the geothermal site (Table 2).
As with most toxic chemicals, there is a wide variation in the
individual response to hydrogen sulfide. It is well documented that
persons who have consumed alcohol within the past 24 hours as well as
psychotic or neurotic personalities are at high risk with respect to
hydrogen sulfide. This is a majority rather than a minority of the
American population (Table 3).
Alcoholics or individuals who have consumed alcohol within 24 hours
of exposure have been overcome by unusually small concentrations of
hydrogen sulfide (Ref. 3). Alcoholics may constitute a hypersusceptible
population. Persons having psychiatric problems are a poor risk at any
hydrogen sulfide level. Individuals with schizoid or paranoid tendencies
-------
0"
U4
Q
P-,
{/)
pj
u
o
Q
5*
C/i
rH
0
X
£2
o
X
O
CJ
(/)
^
O
rH
rt
M
«
Q}
10
S
p
MH
03
rt
H
_f\
O
«> X
rH 3 PL,
rH & O
(D -H P
§ S ri
>
C C
0) -H -H
f^ £ P P
oj Q) nj cd
tO (O P P
H !-( -H
»< tl J J, t J
fj T*I M rn
pi O fn fn
H -H
H W
O 10 (U 0)
T3 0 X X
O J W PJ
0
tO rH IT)
6 1 O rH
*v. rH r-l rH 1
bO 1 O
S O O rH rH
V
f^
VO
0 0
1 vO O
fx. f^» 1 rH
VO vO l^ I
B| O O VD t-^
p^ J
pj O O O vO
V
rH
f>
H
to
to
0
t
Jj
^
o
X
rH
C
H
r"t
P
rH
G
O
H
^j
rt
p
H
fH
fH
H 10
fH
X 3
fH 0
O X
p
rt oo
H
PL, PI
(/) -H
Q) f~*
fH P
H
*X3 ^
C
rt X
4-*
0 Oj
X ^
P-3 *T3
O
O
to
i
o
tn
rH
O
0
(N
1
O
O
rH
crj
(D
P: o
PH-H
w B
X
X (/>
P fH
> O
P! ^
O 00
r-l rt
-P
rt X
P P
rl -H W
fH S f-l
fH 3
H X O
P X
X rt
fH (U 00
O 'O I
p ^
Crj rrj
fH C P!
rH rt -H
O^ f*j
to P
C) fl) 'H
fH bO S
ri
C fH P
rt fn o
O O
O 6 M-i
X
X «
bO
H -a
rH C
to rt
o
o
f^
1
o
o
LO
t^
vO
^f
1
^>
to
to
pj
H
X
P
H
^
X
p
rt
(L)
rrj
*
fH
3
O
X
iH
C
H
X
4-)
H
S
to
g
o
p
o.
1
O
H
£3
O
P
(O
X to
10 fH
P O
X X
bO
H 00
rH 1
CO ^"
o
o
o>
1
o
o
f^.
o
o
VO
1
(^
vO
«^
*
fH
3
o
X
rH
C
H
X
p
H
V
X
P
rt
Q
O
o
«*
rH
1
0
o
o>
^
to
o>
1
o
o
vO
-------
CO
2
O
H
_3
:D
OH
O
a,
CO
r-i
CrS
3C
CJ
( 1
X
w
Q
r 1
ft ,
,-J
ITD
co
2
1
oi
Q
KJ
O
43
tf
H
^*
H
<£
Q
CJ
2
r- 4
H
OS
o
O-,
OH
co
co
"D
[-H
<£
C t
CO
:c
1
C
i 1
t-H
1 1
^^
^
O
rt
H
ft
OJ
OS
rt
O
f-l
ft
+j
^J
o
13
X
w
(D
^
H
(rt
O
f*\
4-)
43
13
(D
4->
C
0
|
O
0
4_)
o
z
rt
H
1
.
13
0)
4-1
C
0
O
8
rH
r-l
0
to
?H
3
O
43
«^-
(M
4_)
to
rt
t
r-l
0
O
0
iH
.
13
0
^j
c
0
E
3
O
O
13
rt
i-H
0
W
o
H
4_>
O
fH
3
0
2
v
to
o
H
O
43
O
X
OH
r-H
0)
H
trt
!H
0
O
.p
c
o
u
0
3
O
r^t
0
T3
0
cti
ft
0
OS
-------
become markedly worse following exposure. Neurotic individuals have
developed innumerable bizarre symptoms, many of which remain for a long
time as after-effects of the exposure (Ref. 3).
Other persons at high risk are those with respiratory illness, eye
infections, and anemia. Although it is not well documented, the magnitude
of additional risk could relate directly to the exposure of persons with
respiratory illness to sulfur dioxide. Persons with anemia are presumed
to be at high risk because hydrogen sulfide is known to r^eact rapidly with
oxyhemoglobin of the blood and thereby interfere with the body oxygen
transport system.
Hydrogen sulfide is extremely toxic and is as rapidly fatal as
hydrogen cyanide. Although it is a caustic irritant which reacts with
the mucous membranes of the respiratory tract, the maior and more serious
toxic effect is paralysis of the respiratory center. The potential victim
is deprived of a warning since even low concentrations paralyze the
olfactory nerve. The principle route of absorption of hydrogen sulfide
into the blood stream is through the lungs. High hydrogen sulfide con-
centrations cause almost instant paralysis of the entire central nervous
system. Dissolved hydrogen sulfide exists only momentarily in the blood
stream. It reacts with the oxygen of oxyhemoglobin to form thiosulfate
and sulfate almost instantaneously. Hydrogen sulfide is prevented from
accumulating in the body or acting as a cumulative poison by the very
rapid detoxification process. Slow intravenous injection of several
-------
times the lethal close of sodium sulfide has no apparent effect whereas
a rapid injection of a much lower dose is fatal. (Ref. 4,5).
Systemic toxic effects result from absorption into the blood stream
at a rate faster than it can be detoxified. This results in exposure
of the central nervous system to the toxic effects of unoxidized hydrogen
sulfide. The precise mechanism by which hydrogen sulfide exerts its
toxic effects has not been firmly established, but it is generally agreed
that enzyme inhibition results from the formation of sulfides of numerous
cations. This reactivity removes enzyme metal cofactors required for
optimal activity. This effect lias been well established in laboratory
investigations on numerous respiratory enzymes (Ref. 6,7,8) but has not
been conclusively demonstrated in the body (Table 4).
Radon 222 is a radioactive gas which occurs in trace amounts in the
noncondensible gases of geothermal effluents. The question of the
(
relative radioactivity hazards associated with geothermal power plants is
currently a topic of lively discussion. The discussions center around
three major points:
What is the concentration of radon 222 in the geothermal gas
phase effluents?
What are the concentrations of the radon 222 daughter products
in the groundwaters?
What is the radioactivity hazard of radon 222 and its daughter
products relative to those of uranium 238 from nuclear power
plants?
46
-------
Table 4. TOXICOLOGY OF HYDROGEN SULFIDE
Reacts with oxyhemoglobin to form thiosulfate or sulfate
Tolerance depends on oxygen content of blood
Physical condition, hyperventilation, alcohol, asthma
Detoxification is rapid
Dose vs. rate
Permanent vs. temporary effects
Subjective symptoms
Lower respiratory poison
Humidity
Not a cumulative poison
-------
The reasons for the uncertainty regarding the adequacy of the gco-
thermal experimental data and the subsequent interpretation of the signif-
icance of the data are illustrated in Table 5.
The emissions that accompany the disintegration of a radioactive
element depend upon the rate at which the disintegration proceeds. This
is commonly expressed in terms of the half-life of the radioactive
clement. For example, radium atoms show some 3 million times the activity
of uranium atoms. This factor will have a different value for each radio-
active element and will be large for those elements whose radioactivity is
slight. A given amount of radium 226 will disintegrate into radon 222
at a rate that will remain practically constant over a period of several
months because of the.comparatively long half-life of radium 226. When
formed, the radon 222 atoms will start to disintegrate into polonium 218
at a rate that is proportional to the number of radon atoms present and
their tendency to disintegrate. At first this rate is slow since only a
few radon 222 atoms have been formed. As radium 226 atoms continue to
decompose, the number of radon 222 atoms disintegrating per unit time
increases. Eventually, the rate at which radon 222 atoms are disintegrating
will become equal to the rate at which they are being formed from radium
226. Tlien the amount of radon 222 will remain constant.
In such a series of radioactive elements in equilibrium, the amounts
of each will remain constant.
-------
Table 5. PRINCIPAL DECAY SCHEMO OF THE URANIUM ELEMENTS
ELEMENT
Uranium
Thorium
Protactinium
Uranium
Thorium
Radium
Radon
Polonium
Lead
Bismuth
Polonium
Lead
Bismuth
Polonium
Lead
MASS NUMBER HALF-LIFE
238
234
234
234
230
226
222
218
214
214
214
210
210
210
206
4.5xl09
24.1
69
2.5xl05
8.3xl04
1.62x10
3.82
3.05
26.8
19.7
1.64x10
25
4.85
138
Stable
years
days
sec.
years
years
3
years
days
min.
min.
min.
-4
sec.
years
days
days
ENERGY
(MEV)
4.2
0.19,0.10
1.18,2.31
4.76,4.71
4.68,4.62
4.78
5.49
6.00
0.65
1.65,3.17
7.68
0.017
1.16
5.30
-
RELATIVE HAZARD*
(a) (B)
1
(380)
(l.lxlO7)
1.8xl04
5.4xl04
2.7xl06
4.3xlOU
7.7xl014
(4.9xl05)
(6.6xl05)
9xl020
(1)
(1880)
1.2X1010
-
*assuming equal radiation energy levels
-------
Human illness duo to industrial exposure to radon 222 and its
daughter products is well documented. Radiation from radon 222 and its
daughter products in metal mines in Joachimsthai, Czechoslovakia was
found to be the cause of a sharp rise in lung cancer in 1949. Similar
increases in the incidence of lung cancer have been reported in the
United States uranium mining and milling industry (Ref. 9) and in
fluorospar mining (Ref. 10).
Both concentration and half-life must be considered in assessing the
relative radiation hazard of radon 222 and its daughter products in geo-
thermal operations. This is always true; it is not unique to the geothermal
situation.
For example, radon 222 is 4.3 x 10 times as hazardous as uranium
20
238. Similarly, polonium 214 is 9 x 10 times as hazardous as uranium
238. The following type of example illustrates the basis for current
concern and uncertainty regarding the potential hazards associated with
radioactive emissions from geothermal operations. Assume that equivalent
radioactivity hazard criteria are applicable to both nuclear power plants
and geothermal power plants. The limit for uranium concentration at the
3
site boundary is usually taken to be 0.05 mg./meter . If radon 222 is
4.3 x 10 times as hazardous as uranium 238, then the equivalent maximum
allowable radon 222 concentration would be 0.05 / 4.3 x 10 or approxi-
-13
mately 10 mg. Although the methodology in this example is correct,
the conclusion is open to question because radon gas is not accumulated in
the human body as is the case for uranium. However, if the same methodology
-------
is applied to polonium 218, a radioactive daughter product of radon 222,
the logic is totally correct and the comparison of relative hazard is
valid. In the case of polonium 214, the equivalent maximum allowable
20 -23
concentration would be 0.05 / 9 x 10 or 5 x 10 mg.
As can be seen from Table 5, the same situation is true in varying
degrees for all of the other daughter products of radon 222. All are
relatively more hazardous than uranium 238 in varying degrees.
In this radiochemical hazards analysis no distinction has been made
between the relative hazards of alpha, beta, and gamma radiation or the
energy levels of the specific radioisotopes. Relative hazard ratings
have been subdivided and treated separately in Table 5 but cross-comparisons
were not attempted.
Externally, alpha particles do not penetrate the skin. However, when
given off internally after ingestion or inhalation, they produce serious
damage because the energy content is completely absorbed.
High energy beta rays can penetrate the protective layers of the skin
but usually do not reach deep-seated organs when delivered externally.
When ingested or inhaled, they often produce more wide-spread damage than
alpha emitters.
This type of analysis suggests that a thorough examination of the
concentrations of radon 222 in the geothermal non-condensible gas phase
51
-------
as well ns the concentrations of the other non-volatile daughter products
in the aqueous phase may be warranted. This is not a simple problem.
Many of the daughter products of radon 222 are present at concentration
levels far below the levels detectable by conventional analytical methods.
Only the most high]y sophisticated chemical and radiochemical trace metal
methods arc capable of yielding significant results.
In all geothcrmal operations, but particularly in those involving
liquid-dominated fields where the volume and concentration of the geothermal
fluids is high, water quality problems related to total dissolved solids
and dissolved heavy metals is a major problem. Although the absolute
concentrations of the individual trace elements may vary within a given
area, the ratios of the constituents are generally similar. The concen-
tration of the solution is the primary variable. The composition of the
solution is relatively constant (Table 6).
These are valuable generalizations to keep in mind during the sampling
and analysis of geothermal waters and associated trace metals. For example,
assume an initial sample of the initial composition and dilution as given
in Table 7.
The extent of dilution determines the sample size and analytical
method required to determine the chemical composition. In the case of
very diluted geothermal fluids, valid chemical analysis of the trace
metal content can be achieved only by use of the most sensitive analytical
methods or by sufficient sample concentration prior to analysis. Much of
the uncertainty related to assessing the toxicity hazard of geothermal
52
-------
Table 6. GEOTHERMAL FLUIDS
Chemical Composition Varies With
Temperature, pressure
Geology
Vapor- vs. liquid-dominated
o Within a Given Area
Absolute concentrations vary
Ratios of constituents are similar
Table 7. HYPOTHETICAL EXAMPLE OF DILUTION EFFECT
COMPONENT
Na
Cl
Mn
6
Pb
F
Hg
Initial
300,000
300,000
2,000
800
400
20
0.01
CONCENTRATION, PPM
Dilute 1/lk Diluted 1/1M
300 0.3
300 0.3
2
0.8
0.4
0.02
_ _
53
-------
fluids stems from the practice of reporting "nil" as the concentration of
a trace metal without specifying the lower limit or accuracy of the
analytical method. As can be seen from Table 7, the conclusions regarding
toxicity are quite different depending on the particular set of data used.
These data may be adequate for preliminary engineering design but are
inadequate for toxicity evaluation.
Engineering data of this type are given in Table 8. As can be seen,
it is difficult to evaluate the potential hazard for most of the heavy
metals. Taking lead as an example, numerical data are reported only for the
most concentrated brine from the Niland area. This presentation of data
leads to the conclusion that lead may be a toxic hazard at Niland but
not at The Geysers or Cerro Prieto. As will be shown later, this is not
necessarily a valid conclusion.
In most cases, evaluation of relative toxicity hazards requires
consideration of worst case conditions and an assessment of average
operating conditions and the probability of occurrence. A worst case
example for the Imperial Valley area is presented in Table 9. This
Table lists the maximum value reported for each chemical species in any
geothcrmal well in the area. This is an unrealistically severe worst
case example from the point of view of environmental assessment of the
Imperial Valley area. The probability of-any one geothermal well having
all of the listed high concentrations is very low. This type of presenta-
tion was chosen because it permits presentation of all of the potentially
toxic chemical species in a single table.
-------
Table 8. COMPOSITION OF CGOTI1CRMAL FLUIDS (Rcf. 11)
Component
Sodium
Potassium
Calcium
Lithium
Magnesium
Strontium
Barium
Rubidium
Cesium
Iron
Manganese
Lead
Zinc
Silver
Copper
Silicon dioxide
Chlorine
Boron
Fluorine
Sulfur
Total dissolved
solids
Ammonium
Bicarbonate
Parts
The Geysers,
California
.12
.10
.20
.002
.06
.10
--
--
--
--
--
.50
20.00
.10
.10
7.10 (sulfate)
28.38
236.0
775.0
per million by
Cerro Prieto,
Mexico
5,610
1,040
321
14
Negative
28
57
__
--
--
--
--
'Trace
Trace
--
9,694
12
Trace
17,000
weight
Niland,
California
53,000
16,500
27,800
210
10
440
250
70
20
2,000
1,370
80
500
--
400
155,000
390
--
259,000
-------
Table 9. TRACE METALS IN GEOTHERMAL FLUIDS - A WORST CASE EXAMPLE*
Water
Component
Silver
Arsenic
Barium
Cobalt
Chromium
Cesium
Copper
Iron
Mercury
Manganese
Lead
Rubidium
Antimony
Tin
Strontium
Thallium
Vanadium
Zinc
Bromide
Iodide
Ammonium
Nitrate
Fluoride
Boron
Max. Cone
(ppm)
3
15
570
0.4
1.8
22
10
4200
0.008
2000
400
168-
0.5
0.65
740
1.5
6
970
146
22
570
35
18
745
Public Supply
(mgm/1)
0.001
0.05
1.0
-
0.05
-
1.0 (taste)
0.3(taste)
0.002
0.05
0.05
-
-
-
-
-
-
S.O(taste)
-
-
0.5
10.0
1.4-2.4
-
(Irr. 0.75-2.0)
TLV 3
Rel.Haz. mgm/m
3,000 X
150 X 0.25
570 X 0.5
0.5
36 X 0.1
_
10 X
14,000 X
4 X 0.1
40,000 X 5.0
8,000 X 0.2
-
0.5
-
-
0.1
0.1
194 X
-
-
1,140 X
3.5 *
8-13 X 2.5
-
(373-1000X)
Air
Rel.Haz.
_
300 X
5,700 X
4 X
90 X
-
-
-
0.4 X
2,000 X
10,000 X
-
5 X
-
-
75 X
300 X
-
-
-
-
-
36 X
-
*Geothermal Wastes and the Water Resources of The Salton Sea Area, Dept.
of Water Resources Bulletin No. 143-7 (February 1970).
56
-------
Maximum concentrations of each species are evaluated in terms of
relative hazard with respect to public water supply and air quality
criteria. In a complete study, all applicable criteria would be
evaluated in a similar manner. The relative hazard was calculated by
dividing the maximum concentration observed by the appropriate limit.
This gives a number which indicates how much any given value exceeds the
maximum allowable concentration.
Manganese, lead, and silver appear to be potentially, the most
serious water toxicity hazards on the basis of this set of data. The
other components such as arsenic, barium, chromium, ammonium ion,
nitrate, and fluoride cannot be ignored but relatively speaking are a
lesser hazard than manganese, lead, and silver which occur at much
higher concentrations. Boron is essentially nontoxic to humans but is
toxic to plants. Although any mercury is bad, the relatively small
concentrations indicated by these data indicate that it may be a lesser
problem relative to some of the others. The question of mercury toxicity
limits is currently a subject of active concern; the ultimate assessment
in geothermal operations will be determined by the established regula-
tory limits as well as by more accurate analysis of geothermal fluids
and gases.
Chronic manganese poisoning has not been as extensively investigated
as lead and mercury, but the physiological effects are similar to those
of the other heavy metals. Manganese is primarily a central nervous
57
-------
system poison. It affects the large ganglion cells of the cortex and
the mid-brain. The symptoms are acute anxiety, compulsive behavior,
hallucinations, and physical disorientation.
Silver poisoning was common at the beginning of the century from
the inclusion of silver compounds in cosmetics. Although fatal poisoning
is not a likely possiblility, continued exposure to soluble silver salts
leads to a permanent blue-black discoloration of the skin and eyes.
Although soluble silver salts may cause local corrosive effects, they
are not likely to produce systemic effects because silver ion is precipi-
tated by protein and chloride.
The physiological effects of other potentially toxic water pollutants
will not be discussed in this paper. The reader is referred to "Handbook
of Poisoning" by Robert H. Driesbach, Lange Medical Publications, Los
Altos, Calif, for an excellent source of toxicological and physiological
information in summary form.
When evaporated to dryness as in waste disposal areas or from
spillage in work areas, the dust and particulate matter may produce an
air quality problem. Generally, the inhalation of dust and particulates
is a more serious human health hazard than direct ingestion. For this
reason, maximum allowable concentrations in air as well as in water
should be considered.
58
-------
It is significant to note that the relative hazards in air are not
the same as in water. The particulate components most hazardous with
respect to inhalation are lead, manganese, and barium. Several others,
including fluoride, vanadium, thallium, chromium, and arsenic, would be
considered to be at dangerous levels if they were encountered in urban
air samples. As was the case for noncondensible gases and radon radio-
active decay products, it is important to consider the relative toxicity
hazard both in terms of industrial hygiene where the primary consideration
is safety of workers and in terms of the safety of surrounding com-
munities after dilution or scavenging.
The sulfur cycle presented in Table 10 summarized the chemical fate
of hydrogen sulfide in the environment. It is converted to sulfur
dioxide, sulfur trioxide, sulfuric acid, or particulates within a matter
of hours or, at most, a few days. The particulates are metal sulfides
or metal sulfates. The mere fact that hydrogen sulfide values drop to
low levels within a short time and distance from the geothermal site
does not necessarily mean that the hazard is less. The nature of the
hazard has been transformed to that of the new chemical species.
In summary, it is appropriate to discuss the potential health
hazards associated with geothermal power plants in a broader perspective.
Hydrogen sulfide is a chemically reactive gas. It will not remain in
the form of hydrogen sulfide for long time periods as shown in Tables
11 and 12.
59
-------
Table 10. SULFUR CYCLE
AIR POLLUTION
0,
Sun
H0
_*. so.
so
H2so4
CONCENTRATION MECHANISM
H2S + Metals
Metal Sulfides
(NH4)2S,PbS,
, HgS etc.
WATER POLLUTION
Metal sulfide j OH- no reaction
Metal sulfide | H+ ^ metal ion | +
60
-------
*
to
J^
U)
fe
0,
w
13
^*
s
w
s
o
r n
(i,
0
S
<
ix
S
§
t-«l
H
S
i-H
i-H
<1>
r-l
JO
rt
H
/* N
E
P.
ft
*?.
d
i i
H
Ct
H
u3
^
8
I
oi
CQ
H- 1
hH
|~^
O*
w
J
hH
H
h- 1
Z
n
Q
H
CQ
-< >-
CtJ CQ
H
c3
C_)
0
ij HH
< H
r ^ .__.
J^* c/D
y o
w ^-_
i-J O
W U
to o oo to
1 t IO t/5 ^ LO
CM r-t CM CM
t*-- O) CM \O
O Tf M 1 1
CM rH -3-
to to
3C tC
Z Z
- C/5 "CM
CM CM CM O
a: E z u
,_^
r-l
s^^*
CM CM CM tO Tt
33 W Z 0 O 0
CM
1
pj
O
O T3
C C
o rt
C
10 CM
h Z
S) S §
X ft H
U/ (ij DO
0 0
CM f-(
(U m a.
W)
rt - j-i
f^ _to Z ^
Co ft
6 6
c ft o
fn "*>-.
0 0 f-i
(4-1 rH 4.)
CO
'O *^ *r^
O CMg
POD
rt U X
rH CJ
0 ft f-l
r-< ft-H
cd <^
U 0
C CM 4->
O >0 C
H rt
+J ^ 4-)
H in 1-1
t/) CM3
O K 10
B B o
o p^ tj
0 ft
0
10 CM T3
rt CM X
W)CM r-H
> ^ C 3
U 1
H C -U
6 0 M
C £ S
T3 10 O
Q O O
<1> 0 ,£2
rC O -P
!-> -H
§ S)Z
H CM
^ <" Z
,0 1 (
H ^Q S
t-l -H ft
H 10 ft
3 C
CT 0) VO
I-H *^ LO
*
61
-------
uj
UJ
OH
P
^r*
Q
r-H
X
2
I-H
H
z.
UJ
,J
tu
j
§
pyj
UJ
H
§
IX,
O
<
u«
H
s!
. .
H
-J
.
CM
i-H
CD
r-H
8
H
f_^
e
P.
P-
2:
o
it
't
00
o
a,
~J
u
I
QS
CQ
3
UJ
, ]
^
HH
H
I-H
I 1
Q
Qj
P
1^3
CQ
t-t >-.
Di CQ
H
c5
O
2
O
_J hH
H I-H
H g
afe
nJ O
UJ U
to o ^i-
1 1 IO LT) O
CM tO LO .
O O) CM \O
o o rr to i
o
CM
HP*
> o
tO IO CM
12 IS **
« CO > CM
CM CM CM O
X X Z U
i-H
^^
CM CM CM «*
X CO 2 0 0
CM
X
CD
f" ^
C V)
CO -H
*T^ £
c
W C
^% TO
O 4-*
CJ r-H
3
W> C
Ctf 0
(H O
CD
> 0
^^
O <~H
x; o
4-> 1
13
o rt
"U O
0) O
+J S:
cd
3 4->
O -H
1 1 5
cd
0 O
CM
>~« »J-«
O
H &,
o to
Ot rH
S CM
O
O (/>
W r-H
co r* i
&o
/ V
O i-H
H i-H
a 0
P! i-H
*o cd
1 ^ e
c^ c ro
X O fH
4-> -H 00
- tl P
E -H (H
3 W P,
,H s «
fH P< rH
XI E CD
H O 4J
I-H O 3
C7* Cd O
UJ W) U
*
62
-------
Even more important, a mechanism exists for the chemical concentra-
tion of toxic heavy metals as their water insoluble sulfides or sulfatcs.
The particular geographical sites of concentration depend on the meteorology
and hydrology of the region. In general, both the heavy metal sulfides
and sulfates will remain insoluble and consequently immobilized as long
as they are in alkaline soils or water. Thus, although hydrogen sulfide
is not a cumulative poison and is not concentrated significantly in the
biosphere, it indirectly functions as a concentration mechanism. However,
if the region in which they are accumulated becomes acidic, high con-
centrations of toxic heavy metals and hydrogen sulfide will be released
at some later date. Geothermal accumulation areas can become acidic in
a number of ways including, for example, diversion of water courses,
inadvertent industrial waste disposal, or intentional addition of chemical
soil additives, such as calcium sulfate. A significant number of toxic
episodes involving heavy metals and sulfides have occurred in this way.
63
-------
REFERENCES
(1) G. L. Waldbatt, Health Effects of Environmental Pollutants,
C. V. Mosby Co., St. Louis, Mo. (1973).
(2) Hydrogen Sulfidc Health Effects and Recommended Air Quality Standard,
Illinois Institute for Environmental Quality Report No. 74-24
(March 1974).
(3) G. A. Poda, Hydrogen Sulfide Can Be Handled Safely, Arch Envir.
Health .12, 795-800 (1966) .
(4) H. W. Haggard, The Fate of Sulfides in the Blood, J. Biol. Chem. 49,
519-529 (1921).
(5) C. W. Mitchell and W. P. Yant, Correlation of the Data Obtained'
from Refinery Accidents with a Laboratory Study of H S and Its
Treatment, U.S. Bureau of Mines Bulletin, 231, 59-79 (1925).
(6) C. L. Evans, The Toxicity of Hydrogen Sulfide and Other Sulfides, Quart.
J. Exp. Physiol. S2_, 231-48 (1967).
(7) S. Kosmider, E. Rogala, and A. Pacholek, Electrocardiographic and Histo-
chemical Studies of the Heart, Muscles in Acute Experimental Hydrogen
Sulfide Poisoning, Arch, Immun. and Therapeutic Exper. 15, 731-740
(1967). .
(8) S. Kosmider, Electrolyte Balance Disturbances in Serum and Tissues
in Subacute Experimental Poisoning with Hydrogen Sulfide, Int. Arch
Gewcrbepath 13, 392 (1967).
6/t
-------
(9) J. K. Wagoner, V. E. Archer, F. E. Lundin, et.al., Radiation as the
Cause of Lung Cancer Among Radium Miners, New England J. Med. 273,
181-88 (1965).
(10) C. W. Cooper, Uranium Mining and Lung Cancer, J. Occupational
Medicine 10_, 82 (1968) .
(11) EPA, "Control of Environmental Impacts from Advanced Energy Sources,"
1974.
-------
ADDITIONAL DISCUSSION OF THE NONEQUILIBRIUM REACTIONS OF I-I^S
PRESENTED AT THE LAKE COUNTY GEOTHERMAL SEMINAR
The rate of conversion of hydrogen sulfide to thermodynamically stable
end products can be calculated by means of a nonequilibrium air chemistry
computer program. This program is analogous to a photochemical smog model
except that it was written to consider up to 36 elements . Thus, it has the
capability of considering photochemical reactions of sulfur species as well as
carbon, hydrogen and oxygen.
This program indicates that hydrogen sulfide is converted to sulfur dioxide,
sulfur trioxide and sulfuric acid rapidly. The rate of conversion varies between
3 and approximately 24 hours depending on humidity, temperature and sunlight
intensity. Within the context of geothermal emissions, this means that hydrogen
sulfide with an air quality standard of 0.04 ppm in California is rapidly converted to
sulfuric acid which does not have an established legal limit. It is highly probable
that the rapid conversion of hydrogen sulfide to other products is part of the
reason why field measurements do not detect signifi cant quantities at distances
remote from the site boundary.
66
-------
LA-UR-75-2335
UNITED STATES
ENERGY RESEARCH AND
DEVELOPMENT ADMINISTRATION
CONTRACT W-7405-ENG. 36
SOME PROBLEMS INVOLVED WITH SAMPLING GEOTHERMAL SOURCES
by
Alan K. Stoker and William D. Purtymun
Los Alamos Scientific Laboratory
University of California
P. 0. Box 1663
Los Alamos, New Mexico 87545
ABSTRACT
Geothermal wells must be sampled for a variety of
purposes including geologic and geochemical interpretation,
engineering design of facilities, environmental release
evaluation, and documentation of baseline conditions. Basic
factors influencing the choice and application of sampling
methods are reviewed including the type of the geothermal
resource, the analyses of interest, well production parame-
ters, utilization processes, and possible sample contamination
or instability. Three basic methods of sampling are described
including condensation, phase separation, and use of evacuated
containers. Several practical problems experienced by various
workers are discussed. These include the natural variability
of fluid composition with time, effects of Veil-bore heat
losses, effects of well flow rate and production time, sam-
pling locations, laboratory simulation stuj'es, contamination
by corrosion reactions, and documentation of hydrologic systems
possibly connected to the geothermal resource.
67
-------
I INTRODUCTION
Discharges from geothermal sources must be sampled
and analyzed for a variety of purposes. Knowledge of the
physical and chemical nature of geothermal fluids is
necessary for understanding the geologic and geochemical
conditions of the natural resource, for designing equipment
and processes to utilize the resource, for anticipating
and evaluating potential environmental releases or required
controls, and for documenting baseline conditions which
may change during the period of resource extraction.
The unique aspects of geothermal sources, especially
high temperatures and pressures, impose constraints on the
methods of sampling. This presentation provides a review of some
factors important in selecting and applying sampling methods
to geothermal discharges. Some examples of practical problems
are included to suggest sampling difficulties encountered in
certain situations. The need to collect and use information
about related hydrologic systems is discussed in the context
of a case study.
II SAMPLING METHODS
A. General Considerations
A variety of sampling techniques have been applied to
discharges from thermal sources. Finlayson reviewed
literature on methods for collecting and analyzing volcanic
and hydrothermal discharges. The basic methods included air
68
-------
displacement, liquid displacement, vacuum tubes or flasks,
condensation, and adsorption. Most of the methods were
applicable to sampling vents, fumaroles, bubbling hot
springs, or other natural openings. The methods preferred
for collecting steam condensate from geothermal boreholes
involved condensation and the use of evacuated flasks.
Another technique is the separation of liquid and vapor.
The choice of a sampling technique requires considera-
tion of the type of fluid to be collected and the analyses
which will be performed. The fluid may be liquid or gas
or a two-phase mixture depending on the type of geothermal
resource and pressure-temperature conditions. A vapor
reservoir will yield either saturated or slightly super-
heated steam containing some fraction of non-condensable
gases. A hot liquid reservoir could produce either a
pressurized liquid with some dissolved gases, or a two-
phase flow of steam and entrained water if pressure-
temperature conditions permit flashing. No natural fluids
will be present in a hot dry rock resource prior to
injection of water, but it will probably contain pres-
surized liquid when operational.
The analytical methodology may impose constraints on
sampling technique. Sample size may be important when
analyzing for minor gaseous or dissolved constituents.
Analysis for gases such as C02 and H~S may require pre-
conditioning in the sample container to control solubility.
-------
Some solids such as silica may precipitate on cooling
requiring predilution of the sample. Constituents with
a propensity for adsorption such as mercury require careful
consideration of materials used in sampling apparatus.
Effluent streams from a facility using geothermal
fluids may or may not require sampling methods similar
to those used for wells. For example, the steam-gas mix-
ture from noncondensable gas ejectors could be sampled by
techniques applicable to steam wells. Howe.ver, the cooled
condensate from a power plant could be sampled by more
conventional water sampling techniques.
The possibility of samples becoming contaminated or
otherwise changing after collection requires special pre-
cautions. Samples may be contaminated during the collec-
tion process by such things as inadequate flushing of
connectors, lines, or containers. Corrosion reactions in
the well casing can contribute gases or dissolved mater-
ials. Formation fluids may be contaminated for some time
after completion by drilling fluids. Changes in sample
composition can result from precipitation, adsorption,
permeation of gases through containers, radioactive decay,
or chemical reactions between constituents in the sample.
B. Examples of Sampling Methods
Three basic techniques depending on separation of
phases, condensation, and the use of evacuated containers
70
-------
have been used to sample fluids from geothermal wells. An
indirect laboratory simulation technique has been used to
obtain predictive information for a. hot dry rock geothermal
resource.
1. Separation of Phases
Liquid-vapor separation has been employed at the
2 3
Wairakei geothermal field in New Zealand. » A simple
separator (Figure 1) operating at atmospheric pressure is
used for sampling low pressure (30-40 psig) two phase flow.
A calorimeter is used to measure the enthalpy of the dis-
charge so separate results from steam and water analyses
can be related to concentrations in the total flow. High
pressure (>100 psig) samples are taken using small Webre
separators (Figure 2) which have very small pressure drops.
Steam samples are condensed in evacuated glass flasks
cooled by water. The flasks may be partly prefilled with
alkaline solution to absorb CO- and H-S for laboratory
analysis by titrations. Water samples are collected after
passing the hot pressurized water through a cold-water
jacketed pipe.
A unique advantage of this method is the capability
to obtain separate samples of the liquid and vapor phases.
Disadvantages include the need for careful control of heat
losses in the equipment so as not to alter the steam/water
ratios.
71
-------
1 STEAM
^»
PIPE
8. V VALVE TO
CONTROL OUTLET
SIGHT GLASS
S. S. TUBE
^HEAVY LAGGING
ON SEPARATOR.
CALORIMETER &
PIPEWORK.
Ji VTO'lia 3 V
Jiflaa-jaafl -^suna aaf
CLIP TO CONTROL
WATER
PIPE INLET SET
TANGENTIALLY TO
3 16" DIA. SEPARATOR.
l" PIPE & T VALVE
STIRRING PADDLE
FLOATING
THERMOMETER
WATER OUTLET f ? r^^'
COPPER
TUBE SPIRAL
CONTAINER WITH WATER TO ACT AS CALORIMETER
FOR MEASURING ENTHALPY OF DISCHARGE.
LOW PRESSURE SEPARATOR SAMPLING APPARATUS
FIGURE 1
72
-------
WEBRE
GLASS FLASK
WITH CONDENSATE
TUBING
HIGH PRESSURE SEPARATOR SAMPLING APPARATUS
FIGURE 2
73
-------
The location at which samples are taken from the well-
head and associated delivery system piping can influence
sample composition. Differences in steam/water ratios,
pressure, and velocity can all be particularly important
when sampling two-phase flow. Mahon has discussed such
problems encountered at Wairakei. Table I (from Ref. 3)
presents data from various sampling points on a wellhead
and by-pass line incorporating an orifice constriction as
illustrated in Figure 3. The numbered points identify
the sampling locations. All sampling locations upstream
of the constriction give consistent results. Downstream
from the constriction, the pressure drops and some water
flashes to steam resulting in reduced CO^/steam ratios at
points 6 through 10. It is possible to relate the CO- to
total discharge in a consistent manner for all points except
number 7 by knowledge of the enthalpy. Mahon suggests that
after passing through the constriction a proportion of the
water in the discharge is thrown to the top of the pipe
resulting in poor sampling conditions for some distance
downstream as indicated by the irregularities at points 5,
6, and 7. About 30 feet downstream, at points 8, 9, and
10, representative steam samples could be obtained but most
of the water was evidently at the bottom of the pipe (Ref. 3).
Similar problems can occxir in steam well and steam line
sampling if there are points at which heat losses cause con-
densation. Each different configuration requires evaluation
-------
UJ
CO
a:
to
o
c
o
c_>
to
to
D_
I
CC
Q
LU|
I-
(J
»«4
OH
h-
(O
Z
o
cc
UJ
i!
u
CO
o
cv
< to
I UJ to
C _l 01
H O _J
s. o
«-i c:
< to
UJ UJ CO
I- _J UJ
10 O _l
cv o
CD s: i«
LT> LH CV rH
CC CC- CO CO
CV CV CV CV
CC CC CC CO
o
UJ
Q
o
o
oc
c
a.
to:
o
N^
c- oo cr. cr. oo cr.
cv cv cv cv cv cv
UJ
OQ
o.
oo
o
a:
UJ
Z 51
a.
a.
i^ r-^ o o o c" cc
oo oc =:- cr oc oc cc-
cx^ c* cc cr c>j
cxi r-^. CD .=r LT\
cr a: ^
T-H tn oc
CSICVJCXCMCSlCNJCVJCNJCMCSlt-Hi-H
I iI CM CM CXJ
Q
UJ
cu
to
to
UJ
u.
o
U
LU
_l
UJ
OO
UJ
a:
^> s~*
to o
LU (O
0£ Q_
(-1 >»x
0
a. z
2: -
< o
oo o_
cocsccrcrcriccrc
o. cr OC' esc. oo co co oc
X.
crcrcrocoocooccccc o
3" i" -3" C^i CV C>J CV CV CV CQ
o:
UJ
75
-------
WELL HEAD .
1
'3~
i
;
i
s
.4'.
L-i-f
i
- fl
2
i r
;
i
1 R' J
!* VALVE
f
t
_,., ... «?/| ,. , , ,-
t f
4_
1
JXL o ^ ; os 90
\
<»
I
\
~ff
j » i
4 ORIFICE
\GATE VALVES
J U
7 10
-
SILENCER
II
*LwEiR BOX
SAMPLING LOCATIONS
FIGURE 3
76
-------
of factors that could degrade the representativeness of the
sample.
2. Condensation
Sample condensation is accomplished hy directing a
portion of the discharge through a condenser to completely
convert steam to liquid or reduce the temperature of water
so no flashing will occur. The condenser may be a coil of
metal tubing, such as aluminum or stainless steel, immersed
in a container of water or ice or a finned air heat exchanger,
At the outlet of the condenser the liquid and noncondensable
gas portions can be collected and handled in different ways.
4
Barnes, et al., collected the liquid in plastic bottles.
Some portions were filtered and then acidified for analyses
of cations or left untreated for analyses of other components,
Noncondensable gases were collected by displacing condensed
water from plastic syringes, and the sealed syringes were
transported to the laboratory submersed in the condensate to
avoid gas diffusion through the plastic. The gases were
analyzed by gas chromatography.
Alternatively, the output of the condenser can be
directed to an interconnected sampling train. The train
could include containers for accumulating the condensate,
others partly prefilled with alkaline solution for removing
C02 and HLS or lead acetate solution for reacting H2S»
a flask for collecting remaining noncondensable gases.
77
-------
The basic method is applicable to single-phase or two-
phase flows where total discharge analysis is of interest.
An advantage of the method is that almost any desired size
of sample can be obtained. If gases are separated in the
field the possibility of reactions prior to analysis is
reduced. Cooled liquids are immediately available for
field determinations of critical parameters such as pH.
Disadvantages include the need for considerable equipment
and availability of cooling water or ice in the field.
3. Evacuated Containers
Evacuated containers can be used for sampling total
discharge. An evacuated pressure cylinder can be connected
directly to a tap on the wellhead or delivery line (Fig-
ure 4). A "T" fitting with a valve and bleed line permits
purging of air from the connecting line and fittings. The
sample size can be controlled by the time allowed for flow
into the container, or by inducing additional condensation
in the container through cooling with water. Certain pre-
cautions are necessary. Tf pressure equilibrium is reached
between the container and delivery line, continued condensa-
tion of steam can displace non-condensable gases out of the
container. If the inlet to the container is at the bottom,
liquid may be able to drain back into the connecting line.
As the sample cools, the contents of the cylinder will be
at less than atmospheric pressure requiring tightly sealed
fittings to avoid atmospheric contamination.
-------
DELIVERY LINE
TAP ON DELIVERY LINE
1/4 PRESSURE TUBING
WELLHEAD
,T- COUPLING WITH VALVE
TO BLEED SAMPLING
LINE
STEEL SAMPLE
CYLINDER
EVACUATED CYLINDER SAMPLING APPARATUS
FIGURE
79
-------
Non-condensable gases can be removed from the con-
tainer for analysis by direct pumping or by introducing
inert carrier gases to strip dissolved gases from the
liquid. The liquid portion of the sample can be drained
after pressure equilibration.
The technique has been applied to single-phase liquid
or vapor flow with reproducible results. ' For two-phase
flow reproducibility was poor, possibly because of the
constantly decreasing flow rate through the connecting line
as pressure builds up in the container.
The technique has the advantage of being simple to use
in the field. Disadvantages include possible reactions
before laboratory analysis. The use of metal cylinders may
be problematic when corrosion reactions are possible or
when minor elements are of interest. Without induced con-
densation, the sample size may be small, making some
anslyses difficult.
4. Laboratory Simulation
The Hot Dry Rock Geothermal Source Demonstration
experiment being conducted by the Los Alamos Scientific
Laboratory presents a unique problem to source sampling.
The geologic system from which energy is to be extracted
contains no naturally present fluids. Two holes are to
be completed at a depth of about 3000 m (10,000 feet) in
granitic basement rock and connected by hydraulic fractur-
ing. Then water will be circulated in a pressurized loop
80
-------
to extract heat from the 200°C. formation.
Thus it is necessary to resort to simulation studies
in the laboratory to obtain an indication of dissolved
materials that will be dissolved by the circulating fluid.
Cores from the zone to be fractured were taken during
drilling. These rock samples have been used in both flowing
and non-flowing laboratory experiments at the pressure and
temperature conditions expected to occur during the in-situ
circulation experiments. Preliminary results indicate that
there will not be any evolution of gases. Dissolved con-
stituents are apparently subject to change with time because
of differential dissolution rates for various mineral types
and because of reprecipitation of certain compounds con-
7 8
trolled by complex geochemical equilibria. * The labora-
tory simulation results will provide an indication of what
materials may be of importance in planning waste disposal
operations and environmental monitoring.
Ill Factors Affecting Fluid Composition During Sampling
A. Natural Variability
An indication of the possible importance of natural
variability in fluid composition is implicit in data for
sequential samples taken from steam wells during different
periods. Figure 5 shows the time dependence of three
constituent ratios measured in five samples collected
during a 24-hour period. The flowrate of the well was
81
-------
.SHORT TERM VflRIflBILITY
2-2
2-
t'8-
e
a:
a:
oc
a:
a:
or
QJ
O
Q_
O
O
i -
0-8-
LEGEND
RflDON/CONDENSflTE
METHflNE/CONDENSFITE
RfiDON/METHflNE
0-4-
0-2-
I -1
5 10 15 20
RELATIVE TIME (HOURS)
FIGURE 5
82
-------
constant within about 1.5% of the average during the period.
The total variation of the Rn/Condensate ratio was about
10%, or roughly twice the standard deviation of about 4%
expected for individual measurements. The ratios of
CH./Condensate and Rn/CH. each cover a range with a factor
of about 2 between the lowest and highest values. Indivi-
dual measurements had standard deviations of about 15%.
4
Barnes, et al., reported a difference in C02 and CH.
contents of samples from steam wells taken 7 days apart.
The ratio of CC>2/CH4 varied by a factor of almost 5 for one
of the wells.
A longer test reported by Kruger and Umana showed
variation in the Rn/condensate ratio measured in 18 samples
collected during a 20-day period. The left-hand portion of
Figure 6 shows the dependence of the ratio with time while
the steam well was flowing at a constant rate of about
100,000 kg/hr (^200,000 Ib/hr) . The extreme A/-alues are
about 28% lower and 32% higher than the average.
These few examples are sufficient to indicate that
variability of fluid composition may be a significant
consideration when establishing a sampling program for
many constituents in geothermal fluids. Variability may
be due to inhomogeneities in the reservoirs, geochemical
changes, or geologic and hydrologic changes.
B. Well Production Conditions
83
-------
I, Well Flowrate
The radon content of geothermal steam discharges can
change significantly at different flowrates. Referring
again to Figure 6, it can be seen that the Rn/Condensate
ratio was higher during the 20-day period when the flow-
rate was about 100,000 kg/hr 0200,000 Ib/hr) than during
the 28-day period after the flowrate of the well was
dropped to about 50,000 kg/hr 0100,000 ib/hr). Following
a transient, the average Rn/condensate ratio measured in
the last 5 samples was about 50% of the average measured
in samples collected at the higher flowrate. Kruger and
Umana suggest that this flowrate dependence of the Rn/Con-
densate ratio may be explained by different flowtimes which
permit different decay periods for the radon, or by differ-
ent emanating power in the different flow volumes swept out
around the well bore at the two rates.
It is possible that similar effects may occur for other
non-condensable gases or for dissolved constituents. One of
the authors has observed an increase in the concentration of
some trace elements in water wells when they are pumped at
lower rates. A possible but as yet unverified explanation
is that certain zones in the aquifer with different chemical
and hydrologic characteristics contribute varying proportions
of the total flow at different drawdown conditions. Similar
variations may be important for geothermal wells.
-------
LONG TERM VflRIflBILITY flT TWO'RflTES
32
30
en
28-
24-
OQ
or
o
o
O
a
or
16-
12-
10-
8-
-5 0
LEGEND
RflOON/CQNQENSflTE
PLO/JRRTE: aooo LB/HRI
i r
i i i i
L 200
CJ
f-
cn
ce:
o
- JOO
10 15 20 25 30 35 40
RELATIVE TIME (DflYS)
45 50 55
FIGURE 6
85
-------
2. Wei1bo re Heat Losses
Heat losses from the wellbore can cause condensation
of steam and thereby increase the proportion of non-condensable
gases. This effect is of most consequence at low flow rates
where the velocity in the wellbore will permit the condensed
fluid to drain under the influence of gravity,, These con-
ditions have been observed in steam wells flowing at so-called
"bleeding" rates. For example, measurements of radon content
in the discharge of one steam well at various times yielded
Rn/Condensate ratios ranging from one to two orders of
magnitude higher at bleeding rates (estimated at about 2,500
to 5,000 kg/hr O5,000 to %10,000 Ib/hr) than were observed
after several hours of performance tests (at rates of about
30,000 to 60,000 kg/hr (^60,000 to ^120,000 Ib/hr)5 In the
same sets of samples CCK/Condensate ratios were as much as
three orders of magnitude higher and CH./Condensate ratios
were as much as two orders of magnitude higher at bleeding
rates compared to those observed after several hours at the
performance test rates.
Thus it is clear that, at least in the case of non-
condensable gases, measurements made at flow rates where
heat losses are proportionally large may not be representa-
tive of conditions at typical production rates
In some situations it may be possible to make theoret-
ical corrections for condensation due to heat Iosses0
Relations developed for predicting and evaluating effects
86
-------
of heat losses during injection of hot fluids for oil
Q
recovery can be modified to predict heat losses from
geothermal wells.
3o Well Production History
The length of time a well has been produced can be
an important parameter affecting the concentrations of
some constituents observed in the discharge. Changes in
some non-condensable gases to condensate ratios have been
observed on both short and long time scales. During six
pressure drawdown tests of steam wells measurements of
Rn/Condensate ratios and CH,/Condensate ratios in sequen-
tial samples fell into a repeated pattern. An example of
the data obtained from these tests is presented in Figure 7,
Within about 1 hour after starting this test the values of
the Rn/Condensate and CH./Condensate ratios dropped by
about 1 order of magnitude from the values observed in
bleeding rate samples taken just prior to test initiation.
(This phenomenon has already been noted as due to the
buildup of noncondensables in the wellbore at the low
bleeding rate.) In the time interval between 1 hour and
about 4 to 6 hours after starting the test, both ratios
decreased by about 50%. After that, generally constant
value of the Rn/Condensate ratio was observed through the
end of the tests which lasted about 6 to 10 hours and in
one test which lasted 16 hours. The data for CH./Conden-
sate and Rn/CH. ratios showed more variability.
87
-------
PERFQRMflNCE TEST COMPONENT RflTIOS
100
10-
or
or
CD
o:
or
o:
or
i-
o
o
0-1
I \
t \
t \
/ \
I \
LEGEND
RflDON/GONDENSPITE
METHflNE/CONDENSflTE
-2
0 2 4 6 8 10 12
TIME PROM STflRT OP TEST (HOURS)
FIGURE 7
88
-------
In addition to being fairly constant after about 4
hours, the Rn/Condensate ratios did not appear to depend
significantly on the flowrate during the short-term draw-
down testsc In one series of three drawdown tests where
the maximum flowrate was 56% higher than the minimum, the
average Rn/Condensate ratio at 4 or more hours varied by
about 4%. In the second series, the maximum flowrate was
about 10% higher than the minimum, but the average Rn/Con-
densate ratio at 4 or more hours varied by about 13%.
This contrasts with data on Rn/Condensate ratios observed
over a period of many days, as shown in Figure 6, where
there is an apparent dependence on flowrate. Figure 6
suggests that there may be a period of several days fol-
lowing a marked change in well flow rate before an approxi-
mate steady-state condition is achieved. In the case of
radon, this may be partly due to the time required to
establish radioactive equilibrium between the transport-
ing fluid and the effective emanating power of the forma-
tion,, For other non-condensable gases or dissolved con-
stituents similar patterns may occur and it would seem
necessary to investigate such possibilities carefully when
planning sampling schedules.
On a longer time scale, Ellis notes that the con-
centration of gas in steam from steam fields tends to
decrease with time. At Wairakei, salinity as measured by
chloride did not change more than 2% in 15 years. In
89
-------
systems of lower permeability concentrations are more
- u-, 10
variable«,
4. Corrosion Reactions
4
Barnes, et al., note that hydrogen released from
water during the oxidation of well casings may constitute
a large fraction of the total noncondensable gas in steam
well discharges. They report hydrogen volume percents
ranging from 50 to 73%. Such reactions occur at approxi-
mately constant rates and thus would contribute a much
smaller proportion of H~ to the noncondensables at higher
flow rates.
Dissolved constituents could also be introduced by
corrosion reactions and their significance would generally
be expected to be related to flowrate. The importance of
any contaminants is, in part, dependent on the types of
analyses performed. It may be possible to analyze for the
contaminant and make suitable corrections, as in the case
of hydrogen0 However, some other substances, such as iron,
may react chemically and alter the sample irreversibly.,
IV DOCUMENTATION OF HYDROLOGIC SYSTEMS RELATED TO GEO-
THERMAL RESOURCES
Problems arise in identifying the presence and chemical
characteristics of water overlying geothermal resources as
most exploratory holes are drilled with fluids (water and/or
mud) for cooling and cutting removal from the test hole0
90
-------
Most, if not all, states require well or test hole construc-
tion to prevent the mixing of potable or fresh water by non-
potable or highly mineralized water penetrated by the hole.
Thus it becomes necessary to initiate a testing program de-
signed to evaluate the water resources at the site or use
indirect methods of investigation to evaluate possible con-
tamination of fresh water aquifers with highly mineralized
fluids.
Indirect methods are generally employed due to lower
costs. Careful collection of data can yield good results.
A regional reconnaissance of the ground water is made by
inventorying existing wells and springs that furnish water
for domestic, municipal, industrial or agricultural use.
In this way the depth, thickness, and chemical quality of
the water bearing rocks are determined. This data can be
extrapolated to the exploration site. Lithologic and geo-
physical logs of the test hole are necessary to confirm
the extrapolated data.
Monitoring the chemical quality of the drilling fluids
can aid in evaluating the penetration of an aquifer con-
taining highly mineralized water. Preliminary field deter-
minations are made with conductance cells and can be supple-
mented by laboratory analyses of selected constituents
(e.g. SO,, Cl, or TDS) The monitoring of increases or
losses of the circulation fluids are also indicative of
potential aquifers during drilling operations0 Collection
91
-------
of water samples, though they may in part be contaminated
by drilling fluids, can be obtained by packer tests, bailing
during periods of lost circulation, or from the discharge
line when air is employed as a cutting carrier.
At the LASL Dry Rock experiment, a regional reconnais-
sance was made of the surface and ground water and its
chemical characteristics. During drilling of GT-2, a
fresh water aquifer occurred at depths of 125 to 137 m in
volcanic rocks. In the underlying sediments, ten potential
aquifers occurred between depths of 137 and 560 m, with a
main zone of saturation occurring at depths of 560 to 730 m
above the granite.
Interpretations were based on the results of the re-
connaissance, lithologic and geophysical logs, and water
samples collected and analyzed when circulation losses
occurred or when air was substituted for water-mud as a
circulation fluid» Monitoring of the quality of circula-
tion fluids indicated a general increase in IDS and
chlorides in the sediment section which was con-
firmed by the water samples collected. In the granite
section, total uranium in the fluids increased from <1 to
60 yg/lo The sample of water from a fracture zone in the
granitic rocks (identified from geophysical logs) collected
during a packer test, contain a total uranium concentration
of 125 ug/1,12
92
-------
Thus, indirect methods can he used to estimate quality
of water as well as hydrologic characteristics during dril-
ling operation. This data is of value to determine well
construction (casing schedules), to meet state criteria, and
to establish a monitoring network for evaluating environmen-
tal effects of geothermal development.
V CONCLUSION
Many factors must be considered in the design and
execution of a geothermal source sampling program and in
the interpretation of the results. The composition of
samples taken from a well may vary with factors such as
heat losses in the well, the flowrate, the production
history, and as a result of contamination introduced by
corrosion reactions. All of these problems suggest that
the most representative samples will be obtained when the
well is operated at conditions close to those of actual
production and after a long enough time to ensure that
steady-state conditions have been reached for the con-
stituents of interest.
The content of some materials encountered in geo-
thermal fluids may fluctuate on both short and long time
scales The nature of such fluctuations must be understood
for each situation in order to plan a sampling program with
statistical validity,,
93
-------
Additional possibilities for variability and uncer-
tainty related to sampling methods and sampling locations
must be examined to preclude adverse effects on results.
Where several alternative sampling techniques and analyti-
cal methods may be applied to measuring a given constituent,
it would be desirable to have some intercomparison studies
performed before adopting a standard or preferred method.
Ceothermal source sampling must include techniques
such as laboratory simulation in order to obtain predic-
tive information for systems which do not contain natural
fluids. leothermal sampling programs must consider the
need to document related hydrologic systems with some
potential for connection to the geothermal resource.
-------
REFERENCES;
1 J. B. Finlayson, "The Collection and Analysis of Volcanic
and Hydrothermal Gases," in Geothermics, Special Issue No. 2_,
Proc. U.N. Symposium on the Development and Utilization""!^ ~~
Geothermal Resources, Pisa, 1970, Vol. 2, Part 2, nr> 1344-1354
2 A. .T. Fills, W. A. J. Mahon, and T. A. Ritchie, "Methods of
Collection and Analysis of Geothermal Fluids," Second Edition,
C^emistrv Division, Department o^ Scientific and Industrial
Research, New Zealand, Report No. C.D. 2103 (.Tulv, 1975).
3 W. A. J. Mahon, "Sampling of Geothermal Drillhole Discharges,"
in Mew Sources of Energy, Proc. of the Conference in Rome,
AugusT Zl-31, 1*F5~1, Vol. II Geothermal Energy T, pp 269-
277 (1964).
4 I. Barnes, M. F.. Hinkle, J. B. Papp, C. Heropoulos, and
W. W. Vaugh, "Chemical Composition of Naturally Occurring
Fluids in Relation to Mercury Deposits in part of North
Central California," U. S. Geological Survey Bulletin 1382-A
(1973).
5 A. K. Stoker and P. Kruger, "Radon Measurements in Geothermal
Systems," Stanford University, Stanford Geothermal Program
renort SGP-TR-4 (January, 1975).
6 P. Kruger and A. Ilmana, "Radon in Geothermal Reservoir
Engineering," paper presented at I.A.E.A. meeting on
Application of Nuclear Techniques to Geothermal Studies,
Pisa, Italy, September 8-12, 1975.
7 J. P. Balagna, Jr., R. W. Charles, and C. C. Herrick, "Lab-
oratory and Numerical Studies of Water-Granite Interactions
at Elevated Temperature and Pressure," Bulletin of the New
Mexico Academy of Science 15:2, 21-2 (1975)
8 J. P. Balagna, Jr., Los Alamos Scientific Laboratory, personal
communication, July 1975.
9 H. J. Ramey, Jr., "Wellbore Heat Transmission," Trans, of
the AIME, Journal of Petroleum Technology 205, 427-435
(April, 1962).
10 A. J. Ellis, "Geothermal Systems and Power Development,"
American Scientist 63:5, 510-521 (1975).
11 W. D. Purtymun, F. G. West, and W. H. Adams, "Preliminary
Study of the Quality of Water in the Drainage Area of the
Jemez River and Rio Guadelupe," LASL Report LA-5595-MS
(April, 1974)
95
-------
12 W. n. Purtymun, F. H. v.rest, and R. A. Pettit, "Oology i
Ceothermal Test Hole GT-2 Fenton 'Hll Site, Tuly, 1974,'
LASL Renort ].\- 5780-M.S (November, 1^74).
96
-------
DRILL STEM TESTING AND SAMPLING OF GEO-PRESSURED BRINES
By
A. G. Edwards and J. M. Montgomery,
Halliburton Services, Duncan, Oklahoma
ABSTRACT
Technology and equipment are available today for obtaining
samples and reservoir data on geo-pressured and most geo-thermal
wells. The ultra-deep search for hydrocarbons has fostered develop-
ment of subsurface equipment capable of withstanding pressure dif-
ferentials of 10,000 psi at 500° F. This equipment has been
successfully used in a limited number of geo-thermal wells but has
seen wide uses in the oil field. In addition to obtaining samples
of the formation effluents, the following formation characteristics
can be calculated: Static Reservoir Pressure, Indicated Flow Capa-
city, Transmissibility, Average Effective Permeability, Damage Ratio
Theoretical Potential with Damage Removed, and the Approximate Radius
of Investigation.
INTRODUCTION
Drill Stem Testing is a temporary completion of a well to
gather data. As early as 1963, Drill Stem Testing (DST) equipment
was being used to evaluate geo-thermal wells in the Salton Sea.
On one of these open hole tests the bottom hole temperature was
450° F at 5000 feet. Sub-surface samplers were not available at
that time; however, the desired data was gathered. The Key to this
97
-------
test and other deep hot tests was a new (in 1963) rubber compound
for packer elements. Laboratory tested to 10,000 psi differential
at 500° F this compound is standard today in oil field applications.
Many deep-hot-high pressure DST's have been run since 1963.
Conditions in some instances have been extremely severe. Equipment
and techniques used for these tests will be discussed.
EQUIPMENT
Four options are available today to those wishing to test
and/or sample geo-pressured or geo-thermal reservoirs. Well condi-
tions and the type data desired usually dictate the type equipment
used.
Figure 1 illustrates a typical string of 'open-hole' tools
used for Drill Stem Testing geo-pressured or geo-thermal wells in
the open-hole (uncased hole). This is standard equipment for Drill
Stem Testing in oil field applications. For geo-thermal wells
where the Bottom Hole Temperature is expected to be in excess of
350° F, the tools are dressed with special high temperature seals
and packer elements. Reference number one gives a detailed des-
cription of each item in this string.
Figure 2 shows the equipment schematically. Since the packer
is larger in diameter than the other tools, a portion of the well-
bore fluids enters the string through the anchor perforations and
98
-------
FIGURE 1
OPEN HOLE SINGLE PACKER TEST
-DRILL PIPE
*HOLLOW PIN IMPACT
REVERSE SUB
(RILL PIPE OR DRILL COLLARS
DUAL CLOSED IN
PRESSURE VALVE
-REVERSE CIRCULATION PORTS
-HANDLING SUB 8 CHOKE
ASSEMBLY (OPTIONAL)
-«HYDROSPRING TESTER
_TB
lY-PASS PORTS
IT. PRESSURE RECORDER
i (A P TYPE)
HHYDRAULIC JAR
VR SAFETY JOINT
BY-PASS PORTS
AXPAND SHOE PACKER
SSEMBLY
-ANCHOR PIPE SAFETY JOINT
-FLUSH JOINT ANCHOR
-HT 500 TEMPERATURE
RECORDER
IT. PRESSURE RECORDER
(BLANKED OFF)
99
FA
-------
passes through the center of the packer. This fluid then exits
through bypass ports in the VR Safety Joint and Hydro Spring Test-
er as shown in Figure 2-a.
While going in the hole, the Hydrospring's valve is closed
so the drill pipe will be empty when it reaches bottom. When the
testing string reaches the bottom of the hole, a portion of the
drill pipe weight is applied to the string. This weight expands
the packer element out against the wall of the hole, isolating
the well bore fluids from the interval to be' tested. As shown in
Figure 2-b, this also closes the bypass ports; and after a brief
time delay(normally 3-5 minutes), opens the valve in the Hydro-
Spring Tester. The formation effluents then enter the test string
through the perforations in the anchor pipe.
After the formation is flowed the desired length of time,
clockwise rotation of the drill pipe closed the Dual CIP Valve,
as shown in Figure 2-c, allowing the formation to repressure the
area around the well bore.
When the formation has had time to develop a build-up pressure
curve, the drill pipe is again rotated clock-wise. This moves the
Dual CIP Valve to the second flow position, as shown in Figure 2-b.
Following the second flow period the drill pipe is rotated
clockwise to close the Dual CIP Valve to develop another build-up
curve. Figure 2-c illustrates this position.
100
-------
<
u.
h-
co
HI
h-
UJ
I-
co
cc
D
LLJ
-J
o
I
t;x>>Xx>Xx<>^xWXx<^/^V/ANW^>^
^^^/^^/,^^/^^Z^^
^^y^:^>^<^>^^;%<<^^
LJLJ LU
CC Q.
z> a
o ,
cr o
o i
u- o
CC
(D
S
u- P
LLJ
O
<
CO
CO
O
D
CD .
5 o
101
-------
After the final closed-in-pressure period, the drill pipe
is raised, closing the Hydrospring's Valve and opening the bypasses
to equalize the pressure differential previously existing across
the packer as shown in Figure 2-d. After a brief pause the pipe
is raised further releasing the packer. A bar is then dropped
through the drill pipe to open the Hollow Pin Impact Reversing Sub
as shown in Figure 2-e. With the reversing valve, drilling fluid
is then pumped down the annulus and back up the inside of the drill
pipe to safely remove the formation effluents.
As shown in Figure 2-f, the reversing valve drains the drill
pipe on the trip out of the hole. Fluid enters the bypass ports
and exits through the Anchor Pipe perforations to bypass the pack-
er.
When the formation is unconsolidated, or the hole has already
been cased, a string of tools similar to those shown in Figure 3
are used for Drill Stem Testing a geo-thermal or geo-pressured re-
servoir. Tools from the VR Safety Joint upward are the same as
those illustrated in Figure 2 for testing in the open hole. The
prime difference when testing inside casing (commonly called a
'Hookwall Test') is the Packer.
The packer for Hookwall Testing is a casing packer having
slips that grip the casing (when activated) to support the drill
pipe weight necessary to expand the packer elements and operate
102
-------
FIGURE 3 - HOOK WALL PACKER TEST
TUBING
IMPACT REVERSE SUB
TUBING
DUAL CLOSED IN PRESSURE
VALVE
REVERSE CIRCULATION PORTS
HANDLING SUB 8 CHOKE
ASSEMBLY
-HYDROSPRING TESTER
-BY-PASS PORTS
-B.T. PRESSURE RECORDER
(AP TYPE)
-BIG JOHN HYDRAULIC JAR
-VR SAFETY JOINT
BY-PASS PORTS
H-RTTS TESTING PACKER
t±
-COLLAR
-PERFORATED TAILPIPE
-B.T. PRESSURE RECORDER
(BLANKED OFF)
H.T.-500 TEMPERATURE
RECORDER
103
FA
-------
the other tools. With the casing packer the slips make it un-
necessary to set the string on bottom. Operating sequence is
similar to that described above for an open hole test.
Effluent samples can be trapped down hole at final flow
pressure and returned to the surface for analysis. This analysis
can either be done at the wellsite (in oil field applications)
or the tool can be taken to the laboratory for a more exacting
analysis. The sample chamber is simply attached to the lower end
of the Dual CIP Valve and is operated by the.Dual CIP Valve. As
shown schematically in Figure 4, the sample is trapped when the
Dual CIP is rotated to the final closed-in-pressure position.
A new tool has opened up two new types of formation evalua-
tion, a Limited Entry Type Open Hole Test and a Limited Entry Type
Hookwall Test. Operationally the two types of tests are similar.
Equipment wise the only difference is in the type packer used.
The Limited Entry Type Test was developed as a sampling tech-
2 "?
nique for deep, hot, hydrogen-sulfide environments. » Consider-
able time was spent in seeking special seals for this hostile en-
vironment. ^ Design criteria for this tool was to sample hydrogen
sulfide reservoirs where the bottom hole pressure could be as high
as 24,000 psi and the bottom hole temperature might be 450° F. To
meet these specifications the system shown in Figure 5 was developed,
104
-------
oc
111
o
CO
_
> ID
UJ O
cc a:
a.
o
Z CO
LZ o
a>
DC
O
<
Z
o
cr
LU
CL
O
LU
DC
D
g
u.
Li CO
Z .O
o
o
'05
-------
Page Six - Drill Stem Testing and Sampling of Geo-Pressured Brines
To date this equipment has been used on two tests, one offshore
and one inland. These tests indicated that this is an operation-
al system.
For testing geo-thermal or geo-pressured reservoirs, all
tools, except the Expansion Type Double Sampler, will be standard
tools dressed for high temperature applications.
Key to the limited entry sampling technique is the new Ex-
pansion Type Double Sampler. This tool is capable of trapping
two bottom hole samples and bringing them back to the surface for
analysis. The sample chambers expand as the tool is brought out
of the hole in order to reduce the sample pressures for safer hand-
ling at the surface. Expansion of the chambers, plus the reduction
in pressure due to the temperature change, normally provides a sur-
face pressure in the chambers equal to 40 to 60% of bottom hole
pressure. Figure 5 is a schematic of the Expansion Type Double Sam-
pler. The tool is open so that the drill pipe fills with drilling
fluid as the tools are run in the hole. A description of a typical
operating sequence describes the other advantages of the system.
Since the tool string is open ended, the drill pipe fills with
well bore fluid as the tools are run in the hole. Once on bottom,
prior to setting the packer, the drill pipe is partially displaced
with a light fluid such as water or diesel, as shown in Figure 5-a.
The amount of water or diesel placed in the drill pipe is just
106
-------
IO
LJJ
cc
o
S
tf
o
o
Q
O
O
g
5
OL,
CO
»7
-------
enough to clean out the rat hole (below the packer) and get the
formation material up to the sampler. If the formation pres-
sure is low, it may be necessary to put a greater amount of cushion
in the drill pipe to further reduce the hydrostatic head. Nor-
mally 10 to 20 barrels of cushion is adequate. With the cushion
in the drill pipe the packer is set. The weight applied to the
packer activates the hydraulic time delay in the Sampler. After
a brief time delay, the Sampler closes off the inside of the tool
and opens the two sample chambers so that all the flow passes
through the chambers. The surface equipment valves can then be
opened and the cushion slowly flowed out of the drill pipe as shown
in Figure 5-b. The cushion is directed to a measuring tank so that
the amount of formation material entering the system is carefully
controlled. When all the cushion has been recovered, the surface
equipment valve is closed and the drill pipe is raised to the free
point to close the sample chambers. Raising the drill pipe auto-
matically moves an 'Indexing J1 within the tool to the next position
so that the drill pipe weight can be placed back on the packer
while the sample chambers remain closed. With the sample chambers
closed, the center of the tool is open as shown in Figure 5-c.
While the surface equipment is closed, a closed-in-pressure can be
taken to obtain additional reservoir data. With the special hy-
draulic hold down packer the formation fluids can be pumped back
into the formation. If the formation material cannot be pumped
108
-------
back into the formation, the packer can be unseated and the re-
covery reversed out. As the tools are withdrawn from the hole,
the sample chambers expand, reducing the pressure inside to approx-
imately 40-60% of bottom hole pressure. These expanded sample
chambers can then be separated from the tool string and one drain-
ed on location. This gives a quick preliminary idea of the type
materials present in the formation while the other sample is being
transported to the laboratory for a more thorough analysis.
The Expanding Type Double Sampler permits the operator to use
the limited entry type technique to obtain samples of the formation
material early in the program. In hardrock formations this tech-
nique can be used to determine if standard material liners can be
used, or if exotic liners will be required because of hydrogen sul-
fide. In unconsolidated formations, the tool allows the operator
to obtain samples of the formation materials so he can determine
if special tubing will be required because of hydrogen sulfide or
if it will be safe to test with the drill pipe.
TEST PROCEDURE
The Key to obtaining maximum data from the Drill Stem Test
is the procedure. Each test will react differently due to pre-
vious activity during the drilling operation, formation charac-
teristics ,etc. Because of these differences, 'cook-book1 test
procedures are not practical. The length of the flow and closed-
109
-------
in-pressure periods should be selected at the well site, based
on surface reactions of the well during the test. As a general
rule-of-thumb, low permeability formations require longer flow
and closed-in-pressure times than high permeability formations.
In either case, the flow periods must be long enough for the well
to clean up and to draw the reservoir down. The first flow
period must be long enough to remove the 'super-charge' so that
meaningful data can be collected. Wells that flow to the surface
should be flowed long enough to reach a semi-steady flow rate
and that rate should be measured. If reservoir data is desired
from wells that do not flow to the surface, the well should not
be permitted to die; i.e., it should not be flowed until the back
pressure created by the hydrostatic head is approaching reservoir
pressure. Surface indications of a well killing itself is ob-
vious in the bubble bucket.
On deep Drill Stem Tests it is sometimes necessary to put
a light fluid(commonly called a cushion), such as fresh water or
diesel oil, in the drill pipe to help protect the drill pipe
against the collapse pressure created by the drilling fluid in
the annulus. These cushions can offer both advantages and dis-
advantages depending on the type formation being evaluated.
On high permeability wells capable of producing at high
rates, a full cushion helps provide better well control, assuming
the formation pressure is adequate to overcome the hydrostatic
110
-------
head of the cushion. When testing very hot formations where salt
water is used as the drilling fluid, cushions can provide a back
pressure to help prevent flash separation of the drilling fluid
(rat-hole fluid) as it passes through the reduced diameters of
the tools. Should this flash separation occur, the precipitants
could plug the tools aborting the test. Where a cushion is not
required to protect the drill pipe, but flash separation is a
possibility, a gas cushion can be used. The gas cushion, usually
nitrogen, can be slowly bled-off at the surface permitting forma-
tions not capable of producing against a liquid cushion to flow.
When testing unconsolidated formations a cushion is bene-
ficial as it will create a back pressure on the formation the in-
stant the Hydrospring opens. With out the back pressure, the face
of these formations tend to 'explode', when the well bore pressure
is reduced as the Hydrospring opens, plugging the tools and some-
times the hole.
Low permeability formation tests, on the other hand, are
usually hindered by cushions. This type formation normally re-
quires longer flow and closed-in-pressure periods. The addition-
al back pressure of the cushion will require even more time to
clean up the well and sufficiently draw the reservoir down.
CONCLUSIONS
Equipment and technology are available today to Drill Stem
111
-------
Test and sample most geo-pressured and geo-thermal wells. Proper
testing procedures and techniques will gather the data to make
an evaluation of the economic feasibility of a well.
Preplanning of the test should include adequate lead time for
the service contractor to obtain the special high temperature seals
and packer elements. These items are not normally stocked so that
they will be fresh for each application.
112
-------
REFERENCES
1. Edwards, A. G.; Winn, R. H. ; "A Summary of Modern Tools and
Techniques Used in Drill Stem Testing", presented at the
Dedication of the U. S. East-West Trade Center, June 14-17,
1973; in Vienna, Austria.
2. Montgomery, J. M.;"Drill Stem Testing and Sampling of Deep
Frio and Wilcox Reservoirs"; presented at the "Geopressured
Geothermal Resources Conference"; June 2-4, 1975 in Austin,
Texas.
3. Edwards, A. G.; "Offshore Evaluation of Hydrogen Sulfide
Reservoirs", presented at the Fourth Annual Convention of
the Indonesian Petroleum Association", June 2-3, 1975; in
Jakarta, Indonesia.
4. Montgomery, J. M.; Gaskins, J.F.; and Coleman, J. E.;
"Exploratory Reservoir Evaluation for Sour Gas or Crude";
Presented at the Sour Gas and Crude Sumposium, November
10-12, 1974; in Tyler, Texas.
5. Edwards, A. G. ; West, E. R.; "Obtain Accurate Data From
Deep Formation Tests"; World Oil, October, 1974, pg.104-109.
6. Olson, C. C.; "Technical Advancement-Four Decades of DST",
presented at the Eighth Annual Logging Symposium of the
Society of Professional Well Log Analysts, June 12-14, 1967,
at Denver, Colorado.
113
-------
ALliim PROFILE OF THE EAST MESA FT ELL-
AS DEIERMTVED FROM P/uAL INDUCTION RESISTIVITY AND SP LOGS
Ey R. T. Littleton and E. E. Burnett
U.S. Bureau of Reclamation
Boulder City, Nevada
October 1975
In a water dominated geothernal reservoir such as the East Mesa Field or,
if you prefer, a high temperature ground-water reservoir, it is practical
to obtain the general dimensions of dissolved mineral concentration by
interpretation of dual induction resistivity logs and self-potential logs.
Values so obtained may be substantially greater or less than those obtained
by laboratory analyses of water samples from the same zones. Therefore, it
is necessary to establish calibrating coefficients by borehole sampling and
laboratory analyses.
In the East Mesa Field, a useful method of obtaining representative fluid
samples is by drill stem testing during drilling. The saturated sand
teds as deep as 7,590 feet in the East Mesa Geotherinal Reservoir produce
readily under drill stem testing. In a 35-minute period, we may obtain a
column of water that extends to within a few hundred feet of land surface;
in fact, we would flow vater at the surface on most drill stem tests but
for blocking the tool with sand. We obtain a good flush of the sand for-
mation tested as indicated by the bottom hole temperature obtained by
maximum recording thermometers run with the drill stem testing tool. The
temperature we obtain is very near equilibrium and invasion has been over-
come. Determination of the temperature profile several weeks after com-
pletion of wells shows that near equilibrium conditions were reached during
the drill stem test by the production of formation fluid.
Our procedure for deriving salinities from logs is as follows: We first
make a lithologic interpretation using the natural gamma ray log, the SP
log, and the resistivity log (see Figure 1). Using these three logs, we
are able to pick the tops and bottoms of individual sand beds with confi-
dence. The next step is to draw the clay or clay-shale line on the SP log.
Then we draw a sandline wherein we give greater weight to the thick sand
bed deflections than to the thin sand bed deflections. We then read a
value of static SP deflection for each sand bed which is the difference
between the clay-shale baseline and the sandline in millivolts. Next we
pick resistivity values in OHM-meters for each permeable bed, which is sand
or sandstone in the geologic environment of East Mesa. Our methodology is
to pick a shallow resistivity value and a deep resistivity value. In
picking these values, we give considerable weight to that portion of the
curve which suggests the most permeable part of the sand. Our shallow
reading presents the flushed zone and our deep reading represents the
unflushed zone. Values of SP and resistivity obtained are inserted in
formulae described in logging company manuals.
-------
GEOTHERMAL TEST WELL
MESA 6H
JP CURVE LOST CIRCULATION ZONE. 7300-7450
->j k-
10 MLLIVQIJS GEOLOGY
Figure I
RESISTIVITY CURVES OHM-METERS
KEY
SANU
c1 CLAY-SHALE
SHALLOW RESISTIVITY
DEEP RESBTIVITY
MEDIUM RESISTIVITY
-------
The results of oar computations of total salinity on the logs of two shal-
low wells, one in the. Yum a Valley and one at the Lasc Mesa Heothermal
Station, are shown in Figures 2 and 3. The well in Yuma Valley was
drilled to test foundation conditions for a desalting plant site. The
logs depicted or. the slide include a geologiTl's log based upon drilling
character and an SP curve and two resistivity curves. This well is
screened from 50 to 85 feet and was pump tested and water samples were
obtained. Electrical conductivity of the sampled interval according to
laboratory analysis is 2,270 p/m, according to the SP curve is 3,660 p/is,
and according to the two resistivity curves is 2,000 p/n.
An SP log of a water supply well at the geothermal site on East Mesa indi-
cates a high mineral concentration in sand at the water table and a sharp
decrease in concentration in the thin sand bed below the first thick clay
bed (Figure 3). The higher salinity in the first sand is indicated by the
stronger SP deflection in the first sand than in the second sand. Pumped
water from two shallow sand-point observation wells nearby, one completed
in the upper sand and the second completed in the lower sand, confirm what
the SP log indicated even though the values computed by the conventional
formula differ substantially from the laboratory values. In order to
obtain a perfect match, it would be necessary to account for all the fac-
tors that influence the SP curve including differences in the basic chemi-
cal makeup of the formation water and the drilling inud filtrate and stray
electrical currents emanating from forces outside the borehole.
A general trend of the salinity profile throughout some 8,000 feet of
strata is computed from the dual induction logs and SP logs of geothernal
Well Mesa 6-1 (see Figure 4). Both the self-potential and resistivity
curves show broad zonations through the interval measured. The resistivity
curve shows mineralization of less than 5,000 p/n down to a depth of
1,450 feet. From 1,450 feet to 6,100 feet + is 5,000 to 6,000 p/ra gener-
ally but locally approaching 9,000 p/m.
Four drill stem tests were made during the drilling of Mesa 6-1.
Laboratory analyses of fluid samples obtained indicated the following total
dissolved solids: 7,620 p/n fron 2,505 to 2,601 feet depth; 5,720 p/n in
the interval 4,445 to 4,480 feet; 1,850 p/m from 5,557 to 5,607 feet depth;
and 12,620 p/n from 7,292 to 8,030 feet depth. With respect to the latter
interval, we have considerable evidence that most of the fluid probably
came from the lost circulation zone between 7,300 and 7,400 feet depth.
Below 7,300 feet we cannot corroborate our methods and computations with
drill stein test data.
We have no airtight explanation for the reversal of the curves but are dis-
posed to state that there was such heavy mud loss that neither the resis-
tivity nor the SP curves sensed the in situ formation water. The geologic
terrane penetrated apparently has fracture permeability which permits deep
mud invasion. Drilling below 7,100 feet, the hole took mud continuously.
Large quantities were lost between 7,300 and 7,400 feet requiring the set-
ting of a cement plug to regain circulation so that we could continue
116
-------
YUCCA DESALTING PLANT SITE
YUMA, ARIZONA
PR-12 Y TEST HOLE
RESISTIVITY CURVES
Figure 2
4O
90
- 190
-200
- 280
- MO
-3SO
-400
MOTE'- REaSTMTY OF
MUD FILTRATE AT 77°F.
* 4.44 OHM-METERS
SLOTTED/
SECTION >
50'- W1 \
40
90
PUMPED
WATER
EC '2270
MICROMH08/CM
AT 77°F
(LAB.)
COMPUTED EC AT
so-ae' -2000 MICROMHOS
f*R CEWT1METER AT 77* f.
(FROM RESISTIVITY CURVES)
COMPUTED EC AT
gO-S5'» 3MO MKMOMH08
PfR CENTMETEN AT 77* F.
(PMM » CURVE)
00 -
100 -
200 -
ISOO-
380 -
4OO-
17
-------
- o
WATER SUPPLY WELL 6-1S2
EAST MESA GEOTHERMAL WELL SITE
SP LOG WATER QUALITIES COMPARED TO PUMPED WATER QUALITIES
FROM OBSERVATION WELLS 6-IMI AND 6-IM2
LOG
Figure 3
SP CURVE
10 MV-»|
- ro
40
- 90
60
COMPUTED EC=69O4 MlCROMHOS/CM
COMPUTED EC
»3700 MlCROMHOS/CM
0-LAND SURFACE
:D
OB. WELL
6-IMI
LAB EC= 19,130
LAB PPM=II,295
39
43
DOB. WELL
6-IM2
LAB EC*«70
LAB PPM-3496
651/2
70
GROUND-WATER TEMPERATURE - 90° F. TO I008F.
NOTE. Although eh* BQQfotH *Mlltloo differ eeMi4«r«bly
tha p«MB«4 laboratory oaalysod o.oalltlo*, ebo oooD«roClv«
tho «poor MM is ladleoe«l by tho CT log to bo ooMldorobly
oro Mlloo than tho lovor BOM, *hleh im eoafinod by tho
loborotory
118
-------
WATER
QUALITY
MESA 6-1
PROFILE
Figw*
WATER QUALITY IN P P
0 5,000 IQooo 15.000
WATER QUALITY tN PPM
0 5«o IQnoo
WATER QUALITY IN PPM
-2.='
6*°° r
OST
4443-4480
SELF POTENTIAL
EZ3
- 5.000
DST
7292-8030
.3 t2,«2OPPhL
0 ST
Z5O9-2«OI
TOST52O PPH
0 5ooo ICXoo I5.ooo
WATER OUAL1TY IN PPM
WATER (
KEY
SAND
0 5x»o IOU ISooo
WATER QUALITY IN PPM
11'
-------
vlrillinp, to the r.arpet depth of 3,000 feet. Wr ended up with fresh :nud
not only in thi: bore-hole but al.so with a s-.-bstannial »,u3i:-t.ir.v In the frac-
tured terrane. Wt_ are unable to place nuch confidence in the values
computed below 7,300 feet for two reasons. One, the value of the raid
filtrate resistivity used for the calculation may not be co~rect; and, tw, ,
deep mud invasion of the fractured terrane may have biased the computations.
The computed salinities may reflect a mud-soaked formation rather than the.
saline character of the natural water that was indicated by the drill stem
test.
With respect to the low salinity of 1,850 p/m in the interval 5,557 to
5,607 feet, we computed a total salinity of about 4,009 p/m from the resis-
tivity curves of the dual induction log. V.'e hava no satisfactory explana-
tion for this difference.
Ve have also computed the salinity profile throughout about 6,200 feet of
strata in Mesa 31-1 which was drilled in the cooler part of the East Mesa
geothermal field (see Figure 5). The two salinity curves computed fvom the
resistivity and self-potential logs more or less parallel each other as was
the case in Mesa 6-1. Also, the resistivity values computed are lower than
the self-potential values and more nearly fit actual formation conditions
based upon our meager data from drill stem tests. The match between the
drill stem test and results computed from the resistivity curves of the
dual induction log are good. Laboratory analyses of water from drill stem
test interval at depth 4,333 to 4,395 feet gave a total dissolved solids
ranging from 2,000 to 2,500 p/m. We computed 2,000 p/m from the resistiv-
ity curves. In the drill stem test interval 5,656-5,696 feet, not shown on
Figure 5, laboratory analyses of samples gave total dissolved solids rang-
ing from about 1,900 to about 2,220 p/m. We computed about 1,300 p/m from
the resistivity curves. The levels of mineral concentration in the water
throughout the reservoir profile are substantially less at Mesa 31-1 than
at Mesa 6-1.
A comparison of salinity profiles as computed from geophysical logs of
three East Mesa geothernal wel]s shows a common pattern (see Figures 7, 8,
and 9). To a depth of 6,000 feet, three salinity zones seem recognizable.
Little information is available above 1,900 feet. Zone 1 is from 1,000 to
about 2,000 feet in which salinities of 2,000-3,000 p/m seem to prevail.
Zone 2 is an irregular zone from 2,000 to about 4,000 feet in which salini-
ties of 2,000-5,000 p/m seem to occur. Zone 3 is from 4,000 to 6,000 feet
in which water of about 2,000 p/m prevails. Mesa 6-1 (Figure 4) seems to
indicate a fourth zone beginning shortly below 6,000 feet in which water in
the 10,000 p/m range occurs, although there is confusion about the water
salinity below 7,300 feet. A comparison of the Bureau of Reclamation wells
with apparent salinity profiles of other deep wells in the area but outside
the East Mesa Anomaly suggests similar zonations in the water quality
profile.
120
-------
QUALITY
MESA 31-1
; lau*Lirv M FPN
Zooq Jowl ^ooo Oooo Owe Tooo Ho
PROFILE
£»,
tv *-£
c" r:
I000 __!.. _
3BOO-- -. -
4 BOO _^ _ -_ _ ^
I BOO- ~
S'
S£LF POTENTIAL PPM
Jy"0; .
£4,-*.-
5900 r
/
Oooo 4ooo dooo Dooo 1
ma OIMUTV IN PPM
I . I I I I ' I I
Oooo looo 2_ooo 3ooo *Tood Oooo Oooo
I Mm *IMLITf ih PPM
121
-------
SP CURVE
GEOTHERMAL TEST WELL
MESA 31-1
Flgurt t
HP-
8 MILLIVOLTS
+ DRILL STEM TEST 4333-4397
OCOL09Y
4300
w»
2
>3
d-
X
4390
4400
RESISTIVITY CURVES
OHM-METERS-*
KEY
9ANO
CLAY-SHALE
SHALLOW NCMTtVITY
DEEP nuttnvmr
122
-------
COMPUTED CURVES OF LOG
MESA 5-1
Figure 7
WATER
QUALITY
FORMATION
FACTOR
POROSITY
SONIC
DENSITY
CEMENTATION
FACTOR
1000
5000
GOOD
123
-------
COMPUTED CURVES OF LOG
MESA 6-2
Figure 8
WATER
QUALITY
FORMATION
FACTOR
PflROSIH
SONIC
DENSITY
CEMENTATION
FACTOR
3000
1000
2000
* 3000
4000-
5000
6000
20
FiCIOl
-------
COMPUTED CURVES OF LOG
MESA 8-1
Figure 9
WATER
QUALITY
FORMATION
FACTOR
POROSITY
SONIC
DENSITY
CEMENTATION
FACTOR
6000
125
-------
Environmental Analysis
Laboratories
CORPORATION
TLW-6147
FIELD SAMPLING OF RADIOACTIVE GEOTHERMAL EFFLUENTS*
by
Arthur J. Soinski
David E. Claridge
Rodney Melgard
LFE Corporation
Environmental Analysis Laboratories
2030 Wright Avenue
Richmond, California 94804
ABSTRACT
A sampling program for radioactive effluents from The Geysers geothermal
power plant is described. Radon-222 was sampled both in the non-condensable
fraction of geothermal steam and in the atmosphere. A variety of solid and liquid
matrices, including steam condensate, cooling tower sludge, soil, and grass, were
sampled and analyzed for 226Ra and 2i°Pb. The three radioactive isotopes 222Rn,
22%a, and 2-^^Pb are members of the naturally-occurring 238y radioactive decay
chain.
Stack sampling techniques were applied to the collection of steam, and a
simple sampling train was constructed to separate and collect the condensable and
non-condensable fractions. Collection techniques for selected solid and liquid
matrices are described. Two complicating factors in the sampling and analysis
program are addressed: the collection of atmospheric samples for radon from a
local source in the presence of the natural background and the long term temporal
variation in the emission rate of contaminants from a geothermal field.
The sampling and analytical methods used are capable of detecting 222Rn,
and 2-^Pb at environmental concentrations that are below the allowable
maximums set in State of California regulations.
126
-------
Environmental Analysis
Laboratories
CORPORATION
TLW-6147
FIELD SAMPLING OF RADIOACTIVE GEOTHERMAL EFFLUENTS
I. INTRODUCTION
A monitoring program lor radioactive effluents from The Geysers Geothermal
Power Plant, Sonoma County, California, was conducted during the summer and
fall of 1974. The State of California has set maximum allowable concentrations
for the naturally-occurring radionuclides 222Rn in air and 226Ra and 210Pb in
water. The purpose of the program was to determine if these allowable concen-
trations are exceeded. The primary goal was to measure the concentration of
222Rn in power plant emissions. Secondary goals were to measure atmospheric
Rn concentrations and to measure the concentrations of 22^Ra and 2l^Pb in
solid and liquid matrices including steam condensate, drilling mud, cooling tower
sludge, surface water, soil, and vegetation. Soil, water, and vegetation samples
were collected both at The Geysers and in surrounding communities.
Radon-222, 226Ra, and 210Pb are members of the 238U decay chain which
is shown in Table 1. Uranium-238 is present in igneous and sedimentary rocks
(2)
at average concentrations of 1.2 to 3.9 ppm (parts per million); v ' however, values
(3) 9*^8
between 0. 03 and 120 ppm have been reported for certain samples. v ' The ^°°u
decay chain involves eight alpha decays and six beta decays (weak branches to other
products are not shown) before the stable element 206pb terminates the chain.
Radon-222, historically called radon, is unique in that it is the only gas
Q O Q
in the ^00u decay chain. It can diffuse out of soil minerals into the soil gas and
then across the soil surface-air interface into the atmosphere. The mean radon
exhalation rate is 0. 70 atoms/cm2- sec or 40. 0 x 10~18 Ci/cm2- sec.' ' The
concentration of radon in ambient air over the continents ranges from 0. 03 to 3. 0
pCi/1 with an average value of 0.1 pCi/1. Radon in ambient air at The Geysers
has three major sources: the surface soil, geothermal steam, and natural fumaroles.
Our study emphasized the latter two sources of radon.
Radon itself presents minimal health hazards; but its daughter products
are chemically reactive when formed, and they readily attach to other particles
in the atmosphere. Some of these particles can be retained in the lung following
127
-------
Environmental Analysis
Laboratories
CORPORATION
inhalation. When the attached radioactive isotopes decay, their decay energy is
deposited in lung tissue. Of special concern are the short-lived daughters 218Po,
214Eb, 214Bi,and214p0.<6>
Because of its relatively long half-life of 1600 years, 22%a in soil can be
regarded as an essentially constant source of radon to the air. Lead-210, with a
22 year half-life, blocks the decay chain, at least over the time scale both of our
sampling and of developed emissions at The Geysers. The presence of 210Pb in
environmental samples at a level above natural background is an indicator of the
long term radiological impact of a geothermal power plant.
Site selection techniques are discussed in the next section. Sample collection
techniques are presented in Section III. Problems unique to geothermal fields are
described in Section IV. The sampling methodology cannot be separated from the
analytical methodology; therefore, selected methods are described briefly in an
appendix.
II. SITE SELECTION RATIONALE
Selecting the sampling points for 222Rn emission measurements was straight
forward. A schematic of a typical power generation system at The Geysers is
shown in Figure 1. The well bore is vented directly to the atmosphere both during
the drilling stage and during power plant shut down. During normal plant operation,
the particle separators, the pressure relief valve, the off gas ejector above the
condenser, and the cooling tower are sources of 222Rn emissions. Steam flows
through the turbine and into the condenser where it is condensed by direct washing
with cooling water The condensate is then used to wash additional steam. Essentially
all of the 222Rn,along with other non-condensable gases, is removed from the con-
denser by the off gas ejectors, and these gases are released to the atmosphere
through a stack approximately 20 to 30 m (60 to 90 ft) high. Therefore, the off gas
ejector stacks were expected to be the most significant emission source. A small
fraction of the 222Rn dissolves in the scrubbing water and is transported to the
k
cooling towers where the large volume flow of cooling air strips out the radon and
then dilutes it upon release to the atmosphere. The radon concentrations both in
the air above the cooling towers and within the cooling tower therefore were expected
to be low.
128
-------
fcsS^ Environmental Analysis
9* ' Laboratories
CORPORATION
Ambient air samples were collected where meteorological conditions and/or
topographical features indicated that higher than average radon concentrations might
exist. Locations frequently occupied by plant personnel and off-site population
centers were also sampled. The on-site and surrounding area sampling locations
are shown in Figure 2. Because the winds are predominantly westerly, sampling
sites were selected at low points along the Mayacmas Mountains ridge line east
of The Geysers. On-site sampling locations included the Union Oil Shop, the
Pacific Gas and Electric Company Camp, points downwind of Thermal 4 (a free-
venting well), power plant control panels, 'and the inside of the condensers (which
are entered occasionally during plant shutdowns). The closest population centers
are the sparsely populated communities in Cobb Valley, Lake County, which is
below and east of the Sonoma-Lake County line. Air samples were collected in
several of these communities as well as in Cloverdale which is west of The Geysers
and served as a control.
Most of the steam condensate is reteasedas water vapor from the cooling
towers, and approximately 20% is re-injected into wells. A much smaller fraction
is discharged to the ground from particle separators and condensate traps. Con-
densate samples were collected below particle separators, below condensate traps,
and in both the cooling towers and off gas ejectors. Water samples were collected
in creeks at a time of the year when flows are low but when the relative contri-
butions of fumaroles and steam condensate to the flows are high. Drinking water
supplies were sampled both at The Geysers and in surrounding communities.
Soil and vegetation samples were collected in the same general locations
as the ambient air samples described previously. Vegetation samples were obtained
of the predominant growth types, usually introduced wild grasses, encountered at
each location. Soil samples were collected nearby from sparsely vegetated areas
in order to facilitate separation of soil from vegetation. Sludge samples were
collected in the cooling tower basins.
HI. SAMPLE COLLECTION
The sampling train used for collection of steam is shown in Figure 3. The
total volume of the tubing is approximately 1 liter. The stainless steel probe was
inserted into the flow stream in such a way that only steam passed through the
collection train; the high steam flow rates made this simple to achieve. The ice
chest was filled with wet ice to condense the steam in the glass
coil. The condensate trap forces non-condensable gases up into
the collection bag. The system was purged and equilibrated
129
-------
Environmental Analysis
Laboratories
CORPORATION
without a collection bag attached for approximately 10 min. or until 10 to 20 cm^ of
condensate had passed through the system. Collection was then initiated and continued
until 1 to 2 liter of non-condensable gas had been collected in a Tedlar bag. The con-
densate, which had been collected in a bottle, was returned to our Richmond facility
for ^lOpb and 226^a analyses. The gas was analyzed in a field laboratory for the 3.8
day half-life 222]^ by means of counting in a Lucas cell. The emission concentration
of radon was calculated from the known volume of gas counted, the volume of gas collected,
and the steam volume sampled (which was calculated from the volume of condensate
collected).
Several natural fumaroles were sampled for radon using a modification of the
sampling method used for steam. A 20 cm (8 inch) pyrex glass funnel was placed over an
area at each fumarole where natural gaseous venting was active. .The funnel was "sealed"
by placing mud around the bottom edge. From the top of the funnel's stem a polyethylene
tube was attached which connected directly to the condensing coil from which the non-
condensables were collected in the usual manner. The collection system was equilibrated
prior to sampling. Water condensate collected on the walls of the polyethylene tubing
and refluxed continually back into the glass funnel, but it was assumed that equilibrium
was achieved and that the condensables collected were properly ratioed to the non-
condensables.
Ambient air samples were collected at a height of between 1 and 2 m (3 to 6 ft)
above ground level. Ambient air was pumped into a Tedlar bag using either a hand
pump or a portable mine safety type personal pump. In most cases an integrated sample
of 2 liter volume was collected over a period of 10 minutes.
Cooling tower exhaust was sampled using the ambient air technique. Off gas ejector
/Q \
exhaust was sampled using standard stack sampling techniques/ ' Samples were collected
through a standard sampling port in the ejector exhaust pipe.
For soil samples a circle of earth of 2 to 4 m (6 to 12 ft) in diameter was selected.
The top 2. 5 cm (1 inch) of soil at several locations in the shape of a "T" within the circle
was removed with a flat-bottomed shovel. The soil was placed in wide-mouth plastic
bottles. At the laboratory, the soil was dried at 110° C, the gravel fraction was sieved
out, and an aliquot was taken by successively removing alternate sections of the soil
pile.
130
-------
Environmental Analysis
Laboratories
CORPORATION
Grasses were cut off at ground level and placed into large plastic bags. In
sampling trees and bushes, only the outer leaves were removed.
Water samples were collected in previously cleaned 1 liter polyethylene
bottles. No special preservation techniques are required for Ra or 210Pb in
water. The collection bottle was rinsed a minimum of three times with the water
prior to actual sample collection. Surface water samples were taken by immersing
the entire bottle. For low flow systems, care was taken not to disturb mud or
sediment. Drinking water and steam line condensate were collected by holding a
bottle within the flow. Instructions for sampling water are given in both an
Environmental Protection Agency Publication^ ' and an ASTM Book of Standards. * '
The EPA publication is a handbook for monitoring industrial wastewater, but it is
an excellent introduction to water sampling and analysis, especially over an
extended time frame.
IV. SAMPLING CONSIDERATIONS
The first problem that we will discuss is the sampling of ambient air for
radon. The atmospheric 222Rn concentration away from point sources
is a function both of meteorological factors such as wind velocity, the temperature
profile, and barometric pressure and of soil conditions such as temperature and
moisture content. A 1 to 2% change in barometric pressure produces changes in
222 (11)
the Rn flux of from 20 to 60%.v ' A drop in soil temperature to below freezing
(12 13)
or saturation of the soil with water will decrease the emanation rate.v ' The
importance of these factors to the design of the sampling program depends upon
opo
the relative importance of Rn released from geothermal fluid compared to that
released from soil within the air basin, the purpose of the monitoring program,
and the accuracy of the results desired. If the program purpose is to assess the
environmental impact of a geothermal development, then a methodology to dis-
tinguish a local source in the presence of a variable background must be designed.
The purpose of LFE Environmental's program at The Geysers was to
determine if California state standards for 222Rn in air and for 226Ra and 21°Pb
(14)
in water were being exceeded due to the operation of the generating station.
The maximum allowable environmental concentrations for these radionuclides in
air and water are shown in Table 2 (there are no concentration limits set for soil
131
-------
Environmental Analysis
Laboratories
CORPORATION
by the State of California). These are total concentrations consisting of the natural
background component plus the contribution of any local source. These allowable
concentrations can be compared with our detection limits which are given in Table 3.
For the purposes of this table, the detection limit is defined as that concentration
which will yield an integrated number of radioactive decay counts
that has a counting error of ± 10%. Numerous definitions of detection limit exist, * '
and therefore one must be careful in making comparisons between the results from
different analytical laboratories. Our detection limits are well below the California
standards,and therefore the sample sizes and analysis methods used were suitable
for the intended purpose. The analysis methods used are capable of counting these
radionuclides at concentrations several orders of magnitude above the detection
limits.
The variability in the soil emanation rate and the effect of meteorological
factors on the radon concentration in an air basin present the possibility of collecting
samples at times and places such that samples that are not representative of
average concentrations and normal variations are collected. In order to assure
state officials that our results would not be below the annual average concentrations,
our sampling program was conducted at a time of the year when soil emanation rates
would be expected to be high; that is, in the late summer and early fall when the
soil is dry and winds are calm. The conditions prevailing at the time of sampling
were documented in order to support the validity of the results.
The second consideration is the long term temporal variation in the emission
rate of contaminants from a geothermal field. Geothermal effluent rates at Warakei,
New Zealand, have been relatively constant over the short term, but they have been
/l c\
variable over the long term exhibiting first increasing and then decreasing behavior.
We are not aware of any emission rate data in the literature for The Geysers geo-
thermal field. Therefore, it is difficult to determine the frequency at which an
emission source should be sampled for either radon or other potentially hazardous
pollutants such as hydrogen sulfide. The recommended frequency of sampling
over the short term is a question that is best answered after a preliminary survey
has been conducted. A cost-effective field sampling protocol that will produce
defensible results can then be formulated.
132
-------
Environmental Analysis
Laboratories
CORPORATION
V. CONCLUSION
Three radionuclides in the 238U decay chain, 222Rn, 226Ra, and 210Pb,
are possible environmental contaminants resulting from the utilization of geothermal
resources for electricity production. A monitoring program was conducted to assure
that the concentrations of these three radionuclides in the environment at and near
The Geysers do not exceed the maximum allowable concentrations set by State
of California regulations. Methodologies for the sampling of 222Rn in
both steam and the atmosphere and for the sampling of 226Ra and 210Pb in solid
and liquid matrices were developed. Standard stack, ambient air, water, vege-
tation, and soil sampling techniques are applicable for the intended purpose at
The Geysers Field. It should be kept in mind, however, that steam from this field
is relatively free of contaminants. Effluent sampling elsewhere-probably will
require modifications of the sampling train described.
Atmospheric radon sampling presents special problems because both the
geothermal resource and the soil are sources of radon. The background atmospheric
radon concentration is a function of a number of topographical, meteorological, and
soil variables. The relative importance of these factors depends upon the purpose
of the monitoring program. The conditions prevailing at the time of sampling should
be documented both as a standard operating procedure and in order to facilitate the
interpretation of possible anomalous data.
133
-------
Environmental Analysis
Laboratories
CORPORATION
APPENDIX I
ANALYTICAL METHODOLOGY
A. RADON ANALYSIS
Radon, in either non-condensable gas or ambient air, was determined by
(17)
means of counting in a Lucas cell.v ' A Lucas cell is a simple instrument consisting
of a glass or metal chamber of approximately 100 cm3 volume that is coated on the
inside with ZnS scintillator and a flat glass bottom. Scintillations caused by alpha
particles are viewed and counted through the uncoated flat window by means of a
photomultiplier tube. The Lucas cell and the associated electronics are of sufficient
durability to be used in the field in the back of a pick-up truck.
A potential interference in the radon determination is the. presence of Rn,
historically known as thoron, which is a member of the 232^ decay chain, another
naturally-occurring decay chain. Thoron has a short half life of only 55 seconds;
therefore if the filling of the Lucas cell is delayed for at least 20 minutes following
collection, the thoron has decayed to a negligible quantity. The air is filtered prior
to introduction into the Lucas cell in order to remove radon's particulate daughters
some of which are also alpha emitters. The cell is filled with a known volume of
gas, the scintillations counted, and the data analyzed by a computer program which
subtracts background and converts the net counts of radon plus radon daughters to
the radon decay rate at the time of collection.
The Lucas cell detection efficiency was determined by de-emanating radon
from an aliquot of a National Bureau of Standards 226Ra standard. Carrier gases
used were helium and also a mixture of non-condensable gases obtained from a well
at The Geysers. Counting efficiencies for alpha emitters are typically 70%.
B. RADIUM-226 ANALYSIS
Radium-226 in water is determined by counting the daughter radon in a
Lucas cell. Water is filtered to remove solids, acidified,concentrated, de-
emanated, sealed and stored for 10 to 20 days to permit the radon to grow in.
The radon is then de-emanated into a Lucas cell for counting.
-------
Environmental Analysis
Laboratories
CORPORATION
Soil and vegetation samples are analyzed in the same manner as water
samples after dissolution steps.
C. LEAD-210 ANALYSIS
The solution from the de-emanation apparatus is acidified with nitric acid
and equilibrated with lead carrier. Lead is precipitated and redissolved in a series
of steps in order to remove various impurities. Lead is finally precipitated as
lead sulfate, mounted on a counting planchet, and the 210Bi daughter is beta counted
periodically over a period of one month. Corrections for counting efficiency and
source thickness are made.
Environmental Analysis
Laboratories
ACKNOWLEDGEMENT to***
The authors are grateful to Mr. William Beeman who directed the field sampling
program and modified both sampling equipment and sampling procedures in the field
to the specific requirements of our monitoring program.
Mr. Joel Robinson of Union Oil Company and Dr. Doug P. Serpa of Pacific Gas
and Electric Company provided helpful suggestions during the course of our sampling
program.
135
-------
Environmental Analysis
Laboratories
CORPORATION
REFERENCES AND FOOTNOTES
*Sampling and analyses funded by the Union Oil Company, Los Angeles, California.
1. Adapted from C. M. Lederer, J. M. Hollander, and I. Perlman, Table of
Isotopes, Sixth Edition (New York: John Wiley & Sons, Inc., 1967).
2. M. Eisenbud, Environmental Radioactivity, Second Edition (New York: Academic
Press, 1973), p. 170.
3. W. M. Lowder and L. R. Solon, "Background Radiation, " U.S. Atomic Energy
Commission Report NYO-4712 (1956).
4. M. H. Wilkening, W. E. Clements, and D. Stanley, "Radon-222 Flux Measurements
in Widely Separated Regions, " in The Natural Radiation Environment, edited by
J.A.S. Adams, W.M. Lowder, and T. F. Gesell, Report CONF-720805 (1972),
p. 717.
5. W. Jacobi and K. Andre, "The Vertical Distribution of 222Rn, 220Rn and Their
Decay Products in the Atmosphere, " J. Geophys. Res. 68, 3799 (1963).
6. International Commission on Radiological Protection, ICRP Publication 6
(Oxford: Pergamon Press, 1964).
7. E. Segre, "Radioactive Decay, " in Experimental Nuclear Physics, edited by
E. Segre" (New York: John Wiley & Sons, Inc., 1959), Volume III, p. 1.
8. cf, e. g., I. H. Davis, "Sampling Probes for Duct or Stack Sampling, " in Air-
Sampling Instruments, Fourth Edition (Cincinnati: American Conference of
Governmental Industrial Hygienists, 1972), p. L-l.
9. U. S. Environmental Protection Agency, Handbook for Monitoring Industrial
Wastewater (1973). (Available from Technology Transfer, U. S. EPA,
Cincinnati, OH 45268).
10. Standard D3370-75T: Sampling Water in 1975 Annual Book of ASTM Standards,
Part 31, Water (Philadelphia: American Society for Testing and Materials, 1975).
11. W. E. Clements and M. H. Wilkening, "Atmospheric Pressure Effects on 222Rn
Transport Across the Earth-Air Interface." J. Geophys. Res. 7J), 5025 (1974).
12. J. E. Pearson and G.E. Jones, "Emanation of Radon-222 from Soils and its Use
as a Tracer," J. Geophys. Res. 7_0, 5279 (1965).
13. H. Israel, M. Herbert, and G. W. Israel, "Results of Continuous Measurements
of Radon and its Decay Products in the Lower Atmosphere, " Tellus 18, 638 (1966).
14. California Administrative Code, Title 17, Chapter 5, Subchapter 4.
136
-------
Environmental Analysis
Laboratories
CORPORATION
15. L. A. Currie, "Limits for Qualitative Detection and Quantitative Determination, "
Anal. Chem. 4£, 586 (1968).
16. R. C. Axtmann, "Environmental Impact of a Geothermal Power Plant, " Science
187, 795 (1975).
17. H. F. Lucas, "Improved Low-Level Alpha-Scintillation Counter for Radon, "
Rev. Sci. Instrum. 28, 680 (1957).
137
-------
Environmental Analysis
Laboratories
CORPORATION
TABLE 1 URANIUM-238 DECAY CHAIN
Isotope
Uranium-238
1 «
Thorium-234
1 "
Protactinium-234
1
Uranium-234
1 «
Thorium-230
Radium-226
1 "
Radon- 222
1 "
Polonium-218
1 «
Lead-214
1 »
Bismuth-214
I »
Polonium- 2 14
1 <*
Lead-210
1 >
Bismuth-210
I ^
Polonium-210
1 "
Lead-206
Half- Life
4.5x 109 yr
24.1 d
1.17 min
2.48 x 105 yr
8. Ox 104yr
1.6x 103 yr
3.825 d
3 . 05 min
26.8 min
19 . 7 min
164 usec
22 yr
5.02 d
138.3 d
Stable
138
-------
Environmental Analysis
Laboratories
TABLE 2 MAXIMUM ALLOWABLE CONCENTRATIONS FOR 222Rn, 226Ra, AND
210pb IN AIR AND WATER AS SET BY STATE OF CALIFORNIA STANDARDS
Concentration Limit
(pCi/1)
Air Water
222^ 3.0
226Ra 0.003 30
210pD 0.004 100
I* Environmental Analysis
Laboratories
TABLE 3 DETECTION LIMITS FOR 222Rn, 226Ra, AND 210Pb IN AIR, WATER
AND SOIL
Detection Limit
(pCi/1 or pCi/kg)
Air Water Soil
222Kn
226^
210pb
0.1 3.0
0.1
4.0
10
200
139
-------
FIGURE i SCHEMATIC DIAGRAM OF THE TYPICAL SYSTEM
AT THE GEYSERS.
OFF GAS EJECTOR
COOLING
TOWER
PARTICLE SEPARATORS,
EXHAUST
I AIR 8
WATER
VAPOR
GROUND
.WELL BORE
REINJECTION
WELL
RESERVOIR
RESERVOIR
-------
-------
-------
SAMPLING HOT SPRINGS FOR RADIOACTIVE AND TRACE ELEMENTS*
Harold A. Wollenberg
Lawrence Berkeley Laboratory
University of California
Berkeley, California 94720
INTRODUCTION
The Lawrence Berkeley Laboratory is conducting a program to define
parameters for assessment of geothermal resources, and to develop and
evaluate techniques to measure these parameters. Field activities, pre-
sently underway, combine interrelating geological, geophysical and geo-
chemical studies, leading eventually to choices of sites for deep test
holes. As well as furnishing valuable information on the nature of a
potential resource, geochemical data provides a baseline upon which the
effects of future geothermal developments may be compared.
To date, most of our studies have been centered in northern Nevada
where high regional heat flow, numerous hot springs, and available govern-
ment land combine to furnish satisfactory field test sites. A regional
heat flow map, Fig. 1, shows the Battle Mountain High, an area where heat
-2 -1
flow exceeds 2.5 ycal cm sec . Figure 2 illustrates a cutaway model of
a geothermal system considered typical of those associated with basin-and-
range fault zones. The fault zone furnishes a pathway for meteoric water
to percolate deeply into a region of high geothermal gradient, forming a
convecting system which occasionally surfaces as a hot spring.
Work performed under the auspices of the U. S. Energy Research and
Development Administration.
143
-------
Highest regional heat flow
Scale
0 500km
Heat flow of 1.5 or less
/* 1.5 HFU contours
|§ Heat flow of 2.5 or greater
XBL735677
Figure 1.
-------
Figure 2.
145
-------
SAMPLING FOR MAJOR AND TRACE ELEMENTS
In our geochemical program, water samples are obtained for laboratory
radiometry, x-ray fluorescence analysis for major elements (Si, Na, K, Ca,
Al, Mg, and S), and neutron activation analyses for trace elements. Collec-
tion methods were devised to retain all solid material, including that
which precipitates. Major-element data furnishes chemical geothermometry,
based on silica-and alkali-element ratios. Besides establishing natural-
background baselines, radio- and trace-element contents of hot and cold
spring waters, as well as of country rock, may help illuminate the pathways
of meteoric water as it flows from its terrestrial origin into hydrothermal
systems, and eventually into springs and wells.
Various types of springs sampled are illustrated on Figs. 3, 4, and 5:
a hot pool at Big Sulfur Hot Springs, a warm pool at Leach Hot Springs, and
a pool below a cold spring east of Kyle, respectively. (Cold springs are
sampled because they may represent the groundwater which enters the fault-
zone hydrothermal systems.) To a limited extent we have attempted to
directly sample blowing wells, as shown on Fig. 6. Chemical geothermometer
temperatures from these samples have compared well with reported measured
subsurface temperatures.
Frequently, we sample muddy seeps, where only a small flow
of water wells up between the cattle hoofprints. At these springs, a 1/4"
diameter tygon tube is inserted directly into the flow, and water is drawn
with a hand-operated vacuum pump as shown on Fig. 7. Instead of passing
into a bottle, the water can also be drawn directly through a 0.45 micron
cellulose acetate filter, whose apparatus is shown on Fig. 8. Therefore,
water can be introduced to the filter either directly from the spring, or
by pumping from a bottle in the field or laboratory. Normally,- 500 ml
-------
CO
(-1
D
60
H
147
-------
Figure 4.
148
-------
149
-------
t>0
H
Pn
150
-------
r^
0)
bO
rH
151
-------
oo
0)
^
D
W>
r-l
152
-------
-3-
Nalgene bottles are used to collect and store the samples. These field
sampling techniques, and laboratory analytical methods and results, are
described in detail in papers by Bowman et al. (1974, 1975), and Hebert
and Bowman (1975).
In the field or laboratory, drops of filtered water are evaporated
onto a lexan disc, with a fixing solution, for subsequent x-ray fluorescence
analysis. (After the x-ray fluorescence analysis, the lexan can be irradi-
ated, cleaned and etched for determination of the water's uranium content.)
Evaporation in the field is shown on Fig. 9, and the resulting disc on
Fig. 10.
For H2S determinations, a silver disc is placed in an unfiltered ali-
quot of each water sample. The disc is later analyzed for sulfur by x-ray
fluorescence. Figure 11 shows the response of x-ray intensity to FLS by
this method.
Filtered samples for neutron activation analysis are obtained by eva-
porating the water directly from the Nalgene bottles (at 80°C) in the labor-
atory. The resulting residue is incorporated with a plastic binder into a
pellet, and irradiated along with standards in a research reactor at the
University of California, Berkeley.
Some results of the neutron activation method are illustrated on Figs.
12 and 13; Fig. 12 illustrates the contrast in uranium contents of hot and
cold spring waters, and Fig. 13 the levels of some trace elements in pools
of differing temperature at Buffalo Valley Hot Springs.
RADIOACTIVE EFFLUENTS
Prior to sampling, a gamma survey of the spring area is conducted
using a portable NaI(T£) detector, shown on Fig. 14. Samples for
153
-------
-------
60
H
155
-------
1000
100
10
Q_
Q_
CVJ
O.I
0.01
BEOWAWE
KYLE
LEACH
BUFFALO
1000 100 10 I
X-Ray intensity
nn
O.I
XBL747-3624
Figure 11.
156
-------
URANIUM (PPB)
Hot and Cold Springs
5
4
£
CL
&3
£
3
"c
o
Z>
2
1
h
-K,,,
1 &m
». L.CU1
C
&
C
M»
H
"
m
^
1
\
UL
1
u.
C
C
i
i
H
H H T
T T i
Z\
C
I
%
1
C
s
, m , Rf>n\A/n\A/f
DC\/WwwV\
l_
C
C
E
H
HT!C
T i
"
f- Buffalo *i
C
I
H H H H C
Figure 12.
1
57
-------
BUFFALO HOT SPRINGS
300
200
100
Sodium (PPM)
6?
150
100
50
Barium (PPB)
Lit
Antimony (PPB)
200
150
100
50
Cesium (PPB)
65°
30
20
10
40
30
20
10
Chloride (PPM)
rrrr,
65'
HZ,
Tungsten (PPB)
277
60
40
20
-
-
777
i-t-U
.
J777
rap^^
777
I2ii^)
60
40
20
Bromide
200
150
100
50
Rubidium (PPB)
65'
73f
Figure 13.
158
-------
Figure 14.
159
-------
laboratory radiometry are usually collected by scooping the spring water
directly from the pools into Nalgene bottles. This minimizes radon loss
which might occur if the water were drawn through the filter system.
Bottle lids are immediately taped, and samples transported to the labor-
atory for gamma-ray pulse-height analyses. The time of sampling is care-
222
fully noted, to account for the radioactive decay of Rn (3.8-day half-
life) between sampling and gamma counting. With a reasonably short
interval between sampling and counting, sensitivity of this method is of
222
the order of a few tens of pCi per liter of Rn. Along with spring
waters, spring wall sinter, tufa, and muck are collected, for subsequent
laboratory gamma-ray analyses. This provides comparison of the contents
222
of radium and other radioelements with the Rn content of the water.
A sampling system for radon emanating in and around a spring system
utilizes alpha-track detectors. This method integrates radon emanation
over a long time period, minimizing short-term fluctuations in response
to changes in atmospheric conditions. The detectors are inverted plastic
cups with specially treated dielectric alpha-sensitive plastic wafers
attached inside, as shown on Fig. 15. They are placed, each in an approx-
imately 0.5 meter deep hole, then covered. After several weeks' exposure,
the cups are retrieved, detectors removed, etched, tracks counted, and
normalized track densities calculated. This service, used primarily by
the uranium industry, is provided by Terradex Company in Walnut Creek,
California. Resulting track densities in the vicinity of Buffalo Valley
Hot Springs, are shown on Fig. 16, and point out the sharp variations in
radon emanation at that site. Figure 17 illustrates the contours of radon
emanation over a broader area of Buffalo Valley. More detailed descriptions
160
-------
TRACK ETCH RADON DETECTOR
Figure 15.
161
-------
30
X
High Rn
zone
Hot Springs
\Area
\
N
49°
1
0
1
0
1
500
Feet
i
150
Meters
1
1000
1
300
30
X
USGS
Drillhole
49
225
-Spring, temperature °C
x - Detector, tracks/mm2
Buffalo Valley Hot Springs
Radon Detector Array
XBL74I2-8379
Figure 16.
162
-------
*«
\
Figure 17.
163
-------
of radiometric methods and results are provided in papers by Wollenberg
(1974a and b, 1975).
SUMMARY
The techniques described briefly here have proved successful in obtain-
ing samples of hot and cold spring waters for x-ray fluorescence, neutron
activation, and radiometric analyses. These sampling methods require
only lightweight, portable field apparatus, and do not involve lengthy
collection procedures. Good flexibility in field operations is necessary
to accommodate the widely varying conditions of temperature, flow, and
accessibility encountered at the different spring sites.
REFERENCES
Bowman, H., A. Hebert, H. Wollenberg and F. Asaro, 1974, A detailed chemical
and radiometric study of geothermal waters and associated rock forma-
tions, with environmental implications; Lawrence Berkeley Laboratory
Report LBL-2966.
Bowman, H., A. Hebert, H. Wollenberg, and F. Asaro, 1975, Trace, minor, and
major elements in geothermal waters and associated rock formations
(north-central Nevada); for Proceedings of Second United Nations Geo-
thermal Symposium, San Francisco.
Hebert, A., and H. Bowman, 1975, Non-dispersive soft x-ray fluorescence
analyses of rocks and waters; for Proceedings of Second United Nations
Geothermal Symposium, San Francisco.
Wollenberg, H., 1974a, Radioactivity of Nevada hot-spring systems; Geo-
physical Research Letters, v. 1, no. 8, p. 359-362.
Wollenberg, H., 1974b, Radon alpha-track survey of a potential geothermal
resource area; Lawrence Berkeley Laboratory Report LBL-3225.
-------
UNION OIL COMPANY OF CALIFORNIA'S GEOTHERMAL
SAMPLING TECHNIQUES
by
D.J. Christoffersen, R.N. Wheatley, and J.A. Baur
Introduction
The purpose of this paper is to describe the procedures used by Union
Oil Company to sample and analyze produced geothermal steam for total compo-
sition.
As is the case in any analytical problem, representative, accurate
sampling is essential to the problem solution. Sampling produced geothernal
steam is no exception to this rule. We are dealing with a sample matrix
which is mainly water and analyzing for gaseous and other compounds some
of which are highly soluble in water and some of which are relatively
insoluble. The usual components of geothermal steam, in addition to water,
are shown in Figure 1 and the sampling and analysis for these compounds is
described below.
Sampling Apparatus
Our sampling train is shown schematically in Figure 2 and photographically
in Figure 3. This system is used for two purposes. First of all it is used
to obtain an accurate measure of the "non-condensible" gas content necessary
for total composition calculations. Non-condensible gases being defined in
this case as those compounds which remain in the vapor phase after passing
through a 0°C condenser and water contact. Included are small amounts of relatively
water insoluble compounds such as hydrogen, nitrogen and methane and also a
portion of water soluble gases such as carbon dioxide, hydrogen sulfide and
ammonia. Secondly, the train is used to obtain representative samples for
analysis.
For determining the non-condensible gas content of steam, the sample
system is set up on location and the insulated container holding the
165
-------
Union Oil's ... Techniques
condenser coil is filled with ice water and the condenser disconnected from
the sample train. The condenser coil is connected to the steam line via a
by-pass valve by means of a 25-foot flexible, stainless steel tubing using
appropriate fittings at both ends. The by-pass valve is positioned to vent
with no flow to the condenser and the sample port valve on the steam line is
fully opened. After flowing steam through the connecting tubing for sufficient
time (usually 10 to 15 minutes) for the tubing to come to temperature, steam
is valved to the disconnected condenser until a condensate flow rate of about
50 ml/min is obtained. Approximately one liter of condensate is purged
through the condenser. This allows the condenser system to come to an
equilibrated condition and also cleans the tubing of any possible residue
or contamination from prior samplings.
Wet test meters must be properly equilibrated to prevent absorption of
gases such as carbon dioxide for accurate volume measurements. This is done
by attaching the equilibrated condenser to the sample train and< passing the
non-condensible gases through the wet test meter for 10 to 15 minutes. The
condenser is disconnected from the sample train. The collection bottle is
replaced by a tared one-quart bottle and the wet test meter reading is noted.
The condenser is quickly attached to the sample train by neans of rubber
tubing and actual sampling begins. Sampling continues until the collection
bottle is filled or 3 to 4 liters of non-condensible gas volume is reached.
The condenser is then disconnected from the train and the non-condensible
gas content of the steam determined from the weight of water collected and
the volume of gas measured after appropriate corrections for temperature
and barometric pressure.
166
-------
Union Oil's ... Techniques
The condenser system is then maintained in equilibrium and used for
collection of analytical samples. Sampling bottles containing appropriate
chemical fixing reagents for collection of reactive compounds are inserted
in place of the condensate collection bottle. A gas sampling cylinder is
placed between the bottles and the wet test meter for sampling non-reactive
gases. Analysis samples are described in detail below.
There have been cases where equipment such as described here has not
been available for field use. In these instances we have used an improvised
method substituting measurement of the volume of non-condensible gases by
water displacement in a graduated cylinder and collection of non-condensible
gases using the device shown in Figure 4.
Collection of Analytical Samples and Their Analysis
Gas samples are collected in 250-ml glass cylinders by the procedure
described above. The composition of the gas is determined by a Union Oil
mass spectrometry- method accurate to ±2% relative or ±0.1% absolute.,
whichever is greater.
For the determination of ammonia a quart bottle containing a weighed
volume (approximately 200 ml) of 0.1N hydrochloric acid and 4 drops of
methyl orange is inserted in the sample train in place of the condensate
bottle. Condensate and gas pass through a glass frit immersed in the
solution quantitatively removing ammonia from both phases. Sampling continues
until approximately 200 ml of condensate has been collected and the bottle
is then removed from the train for weighing and analysis. The analysis
procedure consists of making the solution alkaline, distillation into boric
acid followed by titration with standard hydrochloric acid.
167
-------
Union Oil's ... Techniques
For the determination of hydrogen sulfide, a 1-quart bottle containing
a weighed volume (approximately 100 ml) of 15% neutral cadmium sulfate is
placed in the sample train. Condensate and gas contact this solution after
passing through an immersed glass frit and hydrogen sulfide is quantitatively
removed from both phases. Sampling continues until approximately 100 ml of
condensate have been collected. (A second "check flask" of cadmium sulfate
in series can be used to make certain there is no H2S carryover.) The
/
bottles are then removed from the train, weighed, and analyzed. The analysis
procedure consists of adding excess standard iodine, acidifying and back
titrating with standard thiosulfate. Take care to include any cadmium
sulfide adhering to the glass frit in the analysis.
Samples for carbon dioxide analysis are obtained by placing two one-quart
bottles containing weighed volumes (approximately 200 ml each) of IN sodium
hydroxide in series in the sample train to quantitatively remove carboti
dioxide from condensate and gas. The inlet to each bottle is a glass frit
immersed in the solution. After collection of approximately 100 ml of
condensate, remove the bottles from the sample train, weigh and analyze.
The analysis procedure determines carbon dioxide by acidification with
sulfuric acid, evolution, absorption on ascarite, and weighing the ascarite.
A summary of the analytical methods used is given in Figure 5.
Calculations
The calculations required are relatively straightforward and proceed
as follows: The weight percent of collected non-condensible gases is
derived from the measured volume of gas, average molecular weight of the
gas as determined by mass spectrometry and appropriate temperature,
168
-------
Union Oil's ... Techniques
pressure corrections (i.e., the actual weight of the non-condensible volume)
compared to the same number plus the condensate weight. Similarly the
volume percent of the non-condensibles is calculated by comparing the
corrected volume of non-condensibles to this same volume plus the calculated
gaseous volume of the condensate.
The weight % of inert (as opposed to collected) non-condensibles
methane, hydrogen, nitrogen and argon is calculated from their mass spec-
trometrically determined concentrations in the gas phase, measured gas volume,
and known molecular weights compared to the determined total sample weight.
Similarly the total sample basis concentrations of chemically analyzed
absorbed ammonia, hydrogen sulfide and carbon dioxide are calculated from
determined concentrations, known sample weights and relationship of condensate
to total sample weight from above.
Additional Comments
The sampling scheme described is one which is used on a "dry" brine-free
steam. In those cases where the produced steam has a high dissolved solids
content, exactly the same procedure is used except that a separator is used
to remove brine ahead of the sampling train. Figure 6 is a photograph of
V
such a separator in action in the field.
Figures 7 and 8 are photographs of our sampling vehicle and a field
laboratory at The Geysers field in California's Sonoma County.
169
-------
FIGURE 1 - Components Sought in Steam Compositional Analysis
COMPONENTS SOUGHT IN
STEAM COMPOSITIONAL ANALYSIS
AMMONIA
CARBON DIOXIDE
HYDROGEN
HYDROGEN SULFIDE
"INERT" GASES
METHANE
NITROGEN
FIGURE 2 - Steam Sampling Apparatus
STEAM SAMPLING APPARATUS
FLEXIBLE
LINE
EXCESS STEAM
VENT
NON-CONDENSIBIE
GAS COLLECTOR
CONDENSER
COLLECTORS
CONTAINING
REAGENTS
TO
ATMOSPHERE
WET TEST
METER
170
-------
FIGURE 3 - Steam Sampling Apparatus
FIGURE 4 - Collection of Air Free Gas Samples
COLLECTION OF AIR FREE GAS SAMPLE
FROM
CONDENSER
GAS COLLECTOR
TO METER
OR
GAS BURET
WATER
DISCHARGE
GAS SEPARATOR
171
-------
FIG. 5 - Analytical Methods - Summary
ANALYTICAL METHODS
NON-CONDENSIBLE GASES - MASS SPECTROMETRY
AMMONIA - DISTILLATION
HYDROGEN SULFIDE - IODIMETRIC
CARBON DIOXIDE - ACID EVOLUTION
FIG. 6 - Baca, N.M., Steam Well
172
-------
FIG. 7 - Sample Truck & Gear
FIG. 8 - Field Laboratory at Big Geysers
173
-------
FREDERIKSEN ENGINEERING
Atmospheric Discharge Sampling While
Drilling Geothermal Steam Wells
by M. H. Hyraan and G. R. Fox, Frederiksen
Engineering Co., Consultants, Oakland, Calif.
Presented at "Workshop on Sampling Geothermal Effluents"
U.S. Environmental Protection Agency
Environmental Monitoring & Support Laboratory
Las Vegas, Nevada
October 21, 1975
ABSTRACT
During the final phases of drilling live geothennal steam wells, considerable
amounts of particulate (rock dust) and vapors are discharged. Abatement of
particulates, gases and noise has been accomplished with water injection ?.nc
utilization of centrifugal separators. A special method of sampling for
particulates while drilling, even when active steam formations are encountered,
has been proved to be successful. The sampling train employed is also useful
for source testing of hydrogen sulfide, mercury vapor and ammonia.
ABATEMENT OF PARTICULATE EMISSIONS AND NOISE
At a depth determined by the geologist for each steam well, a switch is made by
the drilling operator from drilling with mud to drilling with compressed air.
Air at several hundred psig pressure is forced down into the well and helps bring
up cuttings which travel (with any steam that is encountered) to the surface and
then horizontally in a blooie line. The period of air drilling may be frrm a few
days to a few weeks depending on drilling problems and how uucb steam is
-------
FREDERIKSEN ENGINEERING
encountered. During that time, a heavy emission to the atmosphere of rock dust
would occur if the blooie line were left open and the emission were not abated.
However, the particulate can be easily abated by injection of water into the
blooie line and the addition of a water/gas separator at the end of the blooie
line. This separator also serves as a noise muffler.
Figure 1 shows how the well, blooie line and separator are connected. Currently
at The Geysers area in northern California, the separator/muffler design
frequently employed has a tangential inlet, thereby providing a centrifugal
separation of particulate and water droplets from the steam and nonccntlerisable
gases. At some point in the blooie line the velocity of the steam and gases is
similar to that in the throat of a Venturi scrubber. Thus, the blooie line, with
water injection, and the muffler act similar to a Venturi scrubber with a cyclonic
separator as used for industrial stack scrubbing.
Currently experiments are underway to find the best water injection point(s) in
the blooie line; to determine the optimum water injection rate; to prove whether
water can be recycled from the sump normally adjacent to the drilling rig. Water
injection affects particulate abatement and abatement of certain gases such as
hydrogen sulfide. In general, the higher the water rate, the better the noise
reduction.
PARTICULATE SAMPLING
Our first testing of a geothermal steam well to be undertaken while the well was
being drilled using compressed air injection was done by EPA Method No. 5. In
this method a sampling probe is inserted into the stack, and the sample is with-
drawn into a small cyclone separator followed by a flat, paper-thin filter in a
175
-------
FREDERIKSEN ENGINEERING
heated box. The filter is followed by a series of wet impingers, a dry trap, a
vacuum pump and a gas flow meter. This method did not work very well, because
the filter tended to plug.
More recent testing has been successful using a method similar to that of the
San Francisco Bay Area Air Pollution Control District (BAAPCD). In this method
the filter thimbles utilized are made of glass packed with plugs of fiberglas
wool, and inserted into the inside of the stack. This type of filtering has bt?^n
demonstrated to be generally non-plugging. By having the thimbles in the stack,
we are assured that the filtering is occurring at stack condit?.ons of temperature
and moisture content. Figure 2 shows a sampling train assembly that can be used.
Achieving isokinetic sampling conditions with a cyclonic separator/muffler is
very difficult. Ideal isokinetic conditions are where the velocity of the. gases
rising in the stack is equal to the velocity entering the sampling nozzle, which
is oriented to face upstream against the gas flow. Hov;ever, when a cyclonic
separator is employed the gas has a spiral flow pattern. Therefore, we have
suggested to one of the well developers at The Geysers area that their new
separator/muffler design includes straightening vanes within the stack. If this
installation is accomplished accordingly, subsequent tests will show whether the
velocity of the gases in the stack can be monitored more accurately. If so, the
rate of sampling can then be. adjusted to match the stack velocity at each point
in the stack where the sample probe is inserted.
SEPARATOR/MUFFLER DESIGN
The first separator/muffler utilized had a main body diameter of.7 feet. This
proved inadequate for wells discharging hundreds of thousands of pounds per hour
176
-------
fREOERIKSEN ENGINEERING
of steam, and entrainment of mud droplets out the top of the stack was very
heavy. Recent designs, using a 10-foot main body diameter, are proving to be
very good for de-entraininent for the wells most recently tested.
In the design of separator/mufflets, consideration also has to be given to the.
ability to dismantle and easily move the apparatus from well to well as the
drilling of new steam wells progresses. The separator/muffler also has to have
provisions for resisting erosion caused by fast-flowing steam and rock dust. The
worst wear points should have extra thick steel. Provisions should also be made
for keeping the separator/muffler clean, since drilling mud is sticky and tends
to build up. Normally if sufficient quantities are always injected into the
blooic line while drilling is in process, the apparatus will stay clean on the
inside. If the water injection is interrupted from time to time, then it. is
advisable to have a second source of water injected into the bottom area of the
separator/muffler.
SAMPLING OF GASES AND OTHER CONSTITUENTS
In some wells, in addition to particulate sampling we have sampled for organic
vapors, hydrogen sulfide and radionuclides. Forthcoming tests are scheduled for
sampling gas phase emissions for ammonia and mercury vapor, with particulate to be
analyzed for arsenic, lead, cadmium and sulfate. The condensate collected in the
impingers will be analyzed in our lab for ammonia, bicarbonate, sulfates, chlorides,
nitrates, calcium, magnesium, sodium, potassium, boron, sulfide, fluoride, iron,
silicon dioxide, mercury, aluminum and conductance.
Gases such ar hydrogen sulfide, ammonia and mercury vapor can be sampled directly
into impingars as shown in part of Figure 2, by partially filling the impingers
with the appropriate chemical reagent solutions. Organic vapors and radionuclides
are normally sampled into stainless steel tanks.
177
-------
FREDERIKSEN ENGINEERING
About the Authors
M. H. Hyman has been Technical Director for Frederiksen Engineering Go.,
Consultants, Oakland, California since 1969, specializing in plant design,
pollution control, testing, and computer applications. He previously had
fifteen years of engineering experience in industry. Ke received a B.S. in
Applied Chemistry from Caltech in 1952 and an M.S. in Chemical Engineering
from the University of California (Berkeley) in 1967. He lectures at local
colleges to engineers on pollution control and on computer applications, and
until recently broadcasted a regular FM series on environmental technology.
Mr. Hyman is registered as a chemical engineer and as a mechanical engineer
and is a member of the American Institute of Chemical Engineers, American
Chemical Society, National Society of Professional Engineers, and several air
and water pollution control associations.
G. R. Fox is Chief Process Engineer for Frederiksen Engineering where he has
been responsible for process design, environmental control, safety and corrosion
problems since 1971. He previously had 12 years experience in the oil refining
industry. Mr. Fox received his B.S. degree from Washington State University
in 1959, and is a member of the Western Gas Processor & Oil Refiners Association.
178
-------
-, rt;i3
LJLJ -
OOo
o
Ck? z
LU a.
Q "
LU 5
Qd °
U_ ^
W
H
iJ
w
ii
o
3
t*
I
179
-------
o,
a . N
w- §
2. §
o
CO
w
o
o c;
^IL,
O *H » * K'4 M/ *^ r-^% i.
54 C l*. f. -C nj VH r
O (, t, c. (_. ^^ ,-s r
r.
o
o
u
o
rj
«-l
(D
o p.
U
.
o in
\t- in
co c\ o ^ 01
»." 10
co
-------
SAMPLING A TWO-PHASE GEOTHERMAL BRINE FLOW
FOR CHEMICAL ANALYSIS*
J. H. Hill
C. J. Morris
Lawrence Livermore Laboratory, University of California
Livermore, California 94550
ABSTRACT
This report describes an experiment designed primarily to define the
problems associated with sampling the two-phase flow in a pipeline of
geothermal brine. Analyses reported for 26 samples include chemical composition,
oxidation potential, pH, density, and total solids. Changes in brine
composition as the well operated during a four-week period are evaluated.
The apparatus and techniques used for sampling are described and evaluated.
*This work was performed under the auspices of the U.S. Energy Research &
Development Administration under contract No. W-7405-Eng-48.
181
-------
INTRODUCTION
When hot, highly saline, geothermal brines are used to generate
electrical power, corrosion and scaling cause serious problems. Because we
must understand brine chemistry to solve these problems, reliable techniques
are needed for sampling and analyzing the brine.
At Sinclair //4 well, brine is produced by a flashing mechanism which
gives a two-phase flow in the brine pipeline. As shown in Fig. 1, several
such systems could exist. At the time this sampling experiment was designed,
it was not known which system predominated. Subsequent observations, made
through a sight glass on a two-inch line, indicate tha't there is probably a
boundary between the two phases but that the "liquid" phase contains entrained
bubbles of vapor and the "vapor" phase contains entrained drops of liquid.
This report describes a sampling experiment conducted at Sinclair #4
well in April 1975. Its purpose was to define the sampling problems, to
evaluate the reliability of samples taken from the pipeline, and to develop
sampling apparatus and techniques.
SAMPLING
Samples were taken from the 6-inch line about 25 feet downstream from
the wellhead using a sample probe as shown in Fig. 2. This probe was inserted
into the brine stream through a one-inch valve on top of the pipeline. The
tip of the probe was positioned approximately 3/4 inch below the top of the
pipe to obtain top ("vapor" phase) samples and about 3/4 inch above the
bottom of the pipe to obtain bottom ("liquid" phase) samples.
One-liter stainless steel sample bottles were used to collect the
samples. The bottles were coated with Teflon on the inside to minimize
182
-------
contact of the brine with the steel. Bellows-sealed stainless steel valves
were used to obtain a gas-tight seal on the bottles. A pressure gauge and
throttling valve were connected to the outlet of each sample bottle to
control the pressure drop through the apparatus. All bottles were flushed
with N- prior to sampling.
Sinclair #4 well was started up on March 31, 1975. So as to permit
stabilization of flow conditions, the first samples were not taken until
April 3. Two samples were taken with the full flow of the 6" line
discharging into the brine pond. The other samples, consisting of three
sets of eight samples each, were taken at weekly intervals with the well
under restricted flow to investigate the effect of four sampling parameters.
The sample numbers are coded to these parameters as follows:
Date of Sampling: 6" - Samples taken April 3, 1975
1 - Samples taken April 9, 1975
2 - Samples taken April 14, 1975
3 - Samples taken April 23, 1975
Position of the probe in the pipe: T - Top
B - Bottom
Effect of cooling: H - Sample was valved off while hot
C - Sample was quenched with water
Orientation of the sample bottle: S - Bottle was sideways (horizontal)
U - Bottle was upright (vertical)
Thus, sample number 1THS denotes a sample taken on April 9 from the top of
the pipe with the bottle oriented sideways (lying horizontally) and valved
off hot. The well operating conditions during sampling were as follows:
183
-------
Date
4-3-75
4-9-75
Wellhead P
(psig)
220
430
Wellhead T
<°C)
210
239
Flow
6" full flow
2" bypass line
1/2" nozzle - 1°
1/2" nozzle-std.
4-14-75 440 247 2" bypass line
1/2" nozzle-std.
1/2" orifice
4-23-75 445 255 2" bypass line
1/2" nozzle-std.
3/4" orifice
During sampling, the throttle valve was adjusted to maintain a 10-Ib.
pressure drop from the pipeline to the outlet of the sample bottle. Bottles
were flushed for at least three minutes before sampling to remove residual
brine from the probe and N~ gas from the sample bottles. For samples which
were quenched, the outlet valve was closed and the bottle was flooded with
running water from a 3/4 inch hose for 30 seconds before the inlet valve
was closed. On samples taken hot, the outlet valve was closed first and
the inlet valve was closed immediately afterward.
ANALYSIS
All samples were cooled to ambient temperature and returned to the
laboratory for analysis. When the samples cooled, insoluble components
were precipitated out of the solution to give a three phase system. Each
of these phases was analyzed separately so that the composition of the
184
-------
"whole" sample could be reconstructed.
Physical and Instrumental
Instrumental measurements were made on the samples before the phases
were separated. Also, some physical measurements were made on the brine
solutions only. Results for these measurements are shown in Table 1.
The volume of the liquid phase was obtained by dividing the weight of
the liquid by the density. The weight of the liquid was taken as the
difference between the weights of the sample bottle full and empty. The
estimated accuracy of this measurement is ±10 cm .
The volume of the gas phase is the difference between the volume of the
liquid phase and the volume of the sample bottle. The estimated accuracy for
3
the volume of the gas phase is ±15 cm .
The measurements of pressure, pH, and E (oxidation potential) were
n
obtained using a pressure cell attached to the sample bottle. This cell was
a stainless steel unit fitted with a 0-200 psia pressure gauge and with
ports to admit high-pressure pH and E electrodes. The cell was first
attached to the sample bottle and flushed with argon. The sample solution
was then admitted into the cell, and the pressure, pH, and E, were measured
at ambient temperature (22°C) . The pressures given in Table 1 have been
corrected for the volume of the pressure cell and its argon content. The
estimated accuracy of these results is ±0.5 psi.
The pH and E, measurements were obtained with a Beckman pH meter and
with high-pressure (150 psi) electrodes. The pH electrodes were calibrated
in standard buffer solutions prior to each measurement.
+2 +3
The E electrodes were calibrated with a Fe /Fe solution with a
h
2
known oxidation potential. The E valves given in Table 1 were measured
with a saturated calomel electrode and corrected to the standard hydrogen
185
-------
potential. The negative voltage indicates a reducing condition in the
solution. The spread in the values is probably caused, at least partially,
by reactions between the brine and the sample bottle at points where the
Teflon lining failed.
The density of the liquid was obtained by weighing an aliquot. The
accuracy of this determination is ±1% of the value.
The total solids content of the samples was determined by evaporating
an aliquot of the liquid to dryness at 110°C in ambient air. The
reproducibility of this determination is ±5% of the value or better.
The results for the volume of solution in the top samples taken hot
indicate that water vapor condensed in these samples while the bottles were
being flushed with brine. As calculated from data given in the steam
tables, the maximum amount of water which could exist as vapor in a one-liter
bottle at 240°C and 440 psi is less than 20 cm (liquid). The maximum
3
amount of "liquid" phase brine present is less than 40 cm , as calculated
from the total solids content of sample 1THS. (For this calculation, it
is assumed that the total solids content of the bottom samples represents
the total solids content of the "liquid" phase.) Thus, the maximum volume
3
of solution in the hot top samples would be less than 60 cm if vapor did
not condense when the bottles were flushed. Therefore, the minimum amount
3
of vapor condensed when the bottles were flushed ranges from 10 cm on sample
1THU to 190 cm on sample 1THS. Similar quantities of vapor probably
condensed in the quenched top samples when they were flushed. However, for
these samples, there is no way to distinguish between vapor which condensed
when the bottles were flushed, and vapor subsequently condensed by quenching.
During bottom sampling, the concentration of total solids could be
increased by flashing of brine or decreased by condensation of vapor in the
apparatus. Because the pressure drop through the apparatus was held to 10 psi
186
-------
for each sample, flashing was minimized. Since it has been shown above
that vapor did condense in the top samples, condensation probably
predominated over flashing in the bottom samples. The amount of condensation
would depend on the liquid/vapor ratio at the tip of the probe during
sampling. Thus, much of the spread in the results obtained for total solids
on these samples is probably caused by condensation. The average solids
content for 12 bottom samples taken under restricted flow conditions is
289 g/1 with a range of 40 g/1. Or, if the value for sample 3BHS is omitted,
the average is 291 g/1 with a range of 20 g/1. Considering that four
different sampling techniques are involved, this is very good reproducibility.
Also, the good agreement between the solids content of the hot and quenched
samples indicates that there was not much vapor entrained in the "liquid" phase
of the brine. Therefore, the higher values shown for total solids content
in the bottom samples are probably very close to the true values for the
"liquid" phase of the brine.
The data shown for the density and total solids content of the 1,2,3
series of samples indicates that the top samples taken with the flow
restricted are primarily "vapor" phase samples while the bottom samples are
primarily "liquid" phase samples. The corresponding data shown for the two
samples (6"THS and 6"BHS) taken from the line under full flow conditions
indicates either that the two phases were well mixed or that the top sample
was taken when a slug of "liquid" phase passed the sample probe. The total
solids content shown for the 6"THS and 6"BHS samples is marginally higher
than corresponding values for the bottom samples from the other three sets.
This could indicate a small change in brine composition as the well
stabilized and flow conditions changed. However, it is more probable that
the higher total solids content in 6"THS and 6"BHS occurred because the
vapor/liquid ratio in the brine was higher under full flow conditions than
187
-------
under restricted flow conditions. The 6"THS and 6"BHS samples (full flow)
were taken at 220 psig and 210°C while the other three sets of samples
(restricted flow) were taken at about 440 psig and 245°C. More of the brine
would tend to vaporize at the lower pressure, causing enrichment of solids
in the liquid phase. If this is the case, the liquid phase nature of sample
6"THS indicates that there was slug flow in the pipeline under full flow
conditions.
Carbon Dioxide and Sulfur
After the instrumental measurements were made, samples of the gas
phase were taken for mass spectrometric analysis. Analyses for three
samples are given in Table 2.
The C0? and lUS in the gas phase were determined by sweeping the gases
from the sample bottles into an analytical train with argon. The C0_ was
collected on ascarite and weighed as CO . The H S was reacted with
Pb(C_H_0«)0» solution and weighed as PbS. Only five samples contained
detectable amounts of S as H S in the gas phase as shown in Table 3.
Aliquots of the solution were acidified to release CO and H?S by the
following reactions.
C03 + H+ = HCO~
HCO~ + H+ = H2C03
= HS
HS~ + H+
The C02 and H?S were swept into the analytical train and analyzed as above.
The H?S content for each liquid-phaj
limit (0.13 mg S/ml) for the method.
The H?S content for each liquid-phase sample was less than the detection
-------
Precipitates from the samples were analyzed for total S by a combustion
technique.
The total C02 and total S assays for each sample were calculated by
combining the values obtained for the gas, liquid, and solid phases. Results
are shown in Table 3.
It is possible for some free S to be present in these samples. However,
the reducing nature of the samples indicates that most of the S should be in
the sulfide form (H2S, HS~, S~, or metal sulfide). Thus, the results for S
shown in Table 3 probably indicate sulfide sulfur. Unfortunately, portions
of the Teflon coating peeled loose from the inside of the sample bottles,
allowing sulfide sulfur to react with the stainless steel. Therefore, the
values shown in Table 3 are probably low. The highest values should be the
most nearly correct, but they could easily be in error by a factor of 2 or 3.
A comparison of results for the top samples with results for the bottom
samples indicates that sulfur is present in both the "vapor" and liquid
phases of the brine.
The data for CO- indicates that most of the CO- is in the vapor phase
of the brine as expected. The difference in concentration of CO- shown for
the hot top samples and the quenched top samples is caused by fractionation
of the CO^, which occurred when vapor condensed during the flushing
operation. In the top samples, some vapor condensed when the bottles were
flushed with brine before the outlet valve was closed. Most of the CO-
corresponding to this condensed vapor was swept out of the bottle. Thus,
the CO- in the hot top samples includes gaseous CO in the uncondensed
vapor and CO dissolved in the solution present when the outlet valve of the
bottle was closed. In the quenched top samples, the same conditions applied
until the outlet valve on the bottle was closed. Then, as the bottles were
cooled by quenching, additional vapor was drawn into the sample bottle and
189
-------
condensed. Thus, the hot quenched samples contained additional CCL
corresponding to the amount of vapor condensed by quenching.
Except for samples 2BHS and 2BCS, the results for CO- in the bottom
samples are consistent. The range of these results together with the
outlying values for 2BHS and 2B.CS is probably caused by differences in the
liquid/vapor mix encountered by the probe during sampling.
These data indicate that the concentration of gaseous constituents in
either liquid- or vapor-phase samples taken from a pipeline depends on the
sampling technique used and also on the extent of phase separation. It
may be possible to avoid fractionation of the gases from condensed vapor
by using flow-through sample bottles heated to brine temperature so that
vapor does not condense when the bottles are flushed. Sampling with
evacuated bottles may also prevent fractionation. Results from replicate
samples will be needed to evaluate the effect of phase separation.
Cations and Anions
The salts were evaporated from two samples of brine and analyzed by
spark source mass spectrometry. This method provides a qualitative
multi-component analysis with approximate concentration levels. Its main
purpose is to provide information concerning potential scale forming elements
which may be missed otherwise. Results are shown in Tables 4a and 4b. The
accuracy of the method used to analyze these samples is ± a factor of 2 times
the concentration for homogenous samples. However, fractional crystallization
of the salts during evaporation made these samples very heterogeneous.
Therefore, these results should be used only for qualitative purposes. These
results indicate that several elements such as Ni, Zn, As, Rb, Sr, and Ba
should be analyzed in more detail as possible scale components. The results
indicate that the concentration of most rare earths is too low for
190
-------
determination directly by emission spectroscopy or x-ray fluorescence
without preconcentration. They can probably be determined either by one
of the above techniques (if a preconcentration step is added) or directly
by neutron activation.
After the phases were separated, precipitates and solutions from five
samples were analyzed separately. Results are shown in Table 5.
These results indicate that the precipitates consist primarily of
silica and heavy metal sulfides. The values for Ca indicate that there may
also be an appreciable amount of CaCO. in the precipitates from 6" THS and
6" BHS. The accuracy of these results is ±5% of the concentration or
better except for the following:
In precipitates Na, Al, Ag ±10%
S (as previously discussed)
In solution Ag ±10%
Al ±1.0 ppm
The results from Table 5 were used to calculate the composition of the
"whole" liquid phase samples. Calculated compositions for the "whole"
samples are shown in Table 6.
A comparison of the results shown for total solids by summation and by
evaporation indicates that the values obtained by evaporation are biased
high. This is as expected because water of hydration associated with
chloride salts would not be completely removed at 110°C. However, the
values obtained by evaporation are within 7% of the values obtained by
summation of the elemental analyses. Therefore, the evaporative technique
can be used to give a quick determination of the approximate total solids
content for Sinclair #4 brine.
191
-------
DISCUSSION
Under ideal sampling conditions, the liquid and vapor phases of the
flowing brine would be either well mixed so that a sample would be truly
representative of the whole brine or well separated so that representative
samples could be taken from both the liquid and vapor phases. Our
observations through a sight glass indicate that this is not the case in a
pipeline where the brine is produced by a flashing mechanism. A comparison
of the data for two samples taken from the 6-inch line under full flow
conditions with data for 24 samples taken under restricted flow conditions
indicates that the extent of phase separation is dependent on flow conditions.
Results for the total solids content of 12 samples taken from the
liquid phase of the brine showed good reproducibility. The solids content
of these samples seems to be independent of the sampling techniques used.
These data indicate that samples taken from the liquid phase of the brine
under restricted flow conditions are adequate for the determination of
solid constituents.
Results for gaseous constituents indicate that the gaseous content of
the samples is dependent both on the technique used to collect the samples
and on the extent of phase separation. Adequate results can probably be
obtained from samples taken under restricted flow conditions, if fractionation
of the gaseous constituents is avoided and if replicate samples are taken
so that the effect of phase separation can be evaluated. In addition,
reliable results for H S or total S will require the development of a sample
bottle that will not react with H-S and that will serve at temperatures
above 250°C and pressures above 450 psi.
Nonreactive sample bottles are also needed to take samples for the
determination of E, and pH. However, reasonably reliable results for these
192
-------
values can probably be obtained using stainless steel sample bottles, if the
samples are analyzed immediately after they are taken.
The results of this investigation indicate that samples which are
adequate for most purposes can probably be taken from pipelines where there
is a definite boundary between the two phases of brine, even though the
phases are not completely separated. However, the extent of phase separation
is probably dependent on operating conditions. Also, the extent of phase
separation has a significant effect on sample reproducibility. It would
therefore be advantageous to take samples from a steam separator whenever
possible.
Acknowledgments
This report is a result of the efforts of many participants. In particular,
we wish to acknowledge the contributions of H. DeCoursey, R. G. Grogan, J. P.
Mahler, E. S. Peck, W. E. Sunderland and M. C. Waggoner.
References
1. Fulk, M. M., et al, General Chemistry Division Quarterly Report,
UCID-15644-75-1, p. 30 (May 1975).
2. Light, Truman S. , Anal. Chem., 44, pp. 1038-39, (1972).
NOF1CK
"I ins report was prepared as an account of work
sponsored by the United States Government. Neither
the United States nor the United States hnergy
Research & Development Administration, nor any
>l their employees, nor any of their contractors,
.ubcontractors, or their employees, makes any
warranty, express or implied, or assumes any legal
liability or responsibility for the accuracy,
Heteness or Usefulness of any information.
ippa
"epre
'atus,
product or process disclosed or
that its ^ use would not infringe
"Reference to a company or product
name does not imply approval or
recommendation of the product by
the University of California or the U.S.
Energy Research & Development
Administration to the exclusion of
others that may be suitable."
tely-owned rights.
193
-------
Table 1. Physical and instrumental analyses.
Sample
number
6"THS
6"BHS
1THS
1THU
1TCS
1TCU
2THS
2THU
2TCS
2TCU
3THS
3THU
3TCS
3TCU
1BHS
1BHU
1BCS
1BCU
So In.
vol.
(cm )
600
830
250
70
430
420
200
250
390
370
170
170
350
480
860 '
870
930
910
Gas
vol.
(cm3)
380
150
730
910
550
560
780
730
590
610
810
810
630
500
120
110
50
70
Pressure
(psia)
6.8
6.4
14
14
112
198
10.5
11.0
91
85
11.0
11.5
86
152
6.4
2.1
16
1.3
H
P
5.1
5.3
5.3
5.6
5.3
5.0
5.4
5.3
5.1
5.7
5.7
5.3
5.9
5.6
5.2
5.5
5.5
Eh
(volts vs.
Std. H)
+0.18
4O.18
+0.19
+0.16
+0.24
+0.07
+0.19
+0.28
+0.17
+0.17
+0.16
+0.08
-0.08
-0.03
+0.13
+0.17
+0.15
+0.13
Total solids
by evapor-
ation - 110°C
(g/D
300
325
11
11
2.9
3.0
3.3
1.2
0.9
0.7
3.6
4.0
6.2
0.8
290
290
300
290
Density
(g/cm3)
1.19
1.20
1.01
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.18
1.18
1.18
1.18
-------
Table 1. Physical and instrumental analyses (continued)
Sample
number
2BHS
2BHU
2BCS
2BCU
3BHS
3BHU
3BCS
3BCU
So In. Gas E,
n
Total solids
by evapor-
vol. vol. Pressure (volts vs. ation - 110°C Density
33 H
(cm ) (cm ) (psia) p Std. H)
490 490 20 4.8 +0.26
870 110 2.1 5.2 +0.18
860 120 36 4.5 +0.28
920 60 2.6 5.6 +0.16
850 130 3.2 5.1 +0.24
850 130 4.4 5.2 +0.16
910 70 22 5.3 +0.35
900 80 7.8 5.2 +0.16
Table 2. Mass spectrometric gas
(Volume percent)
Sample 1TCU 2BCS
C02 97.2 94.3
CH4 1.7 3.9
N2 0.5 0.5
H2 0.5 0.9
(g/1) (g/cm3)
284 1.18
280 1.18
300 1.18
300 1.18
260 1.18
290 1.18
290 1.18
290 1.18
analysis.3
3TCU
97.0
1.9
0.6
0.4
Excludes HO vapor
195
-------
Table 3. Carbon dioxide and sulfur analysis.
Sample
number
6"THS
6"BHS
1THS
1THU
1TCS
1TCU
2THS
2THU
2TCS
2TCU
3THS
3THU
3TCS
3TCU
1BHS
1BHU
1BCS
1BCU
Soln.
wt.
(g)
712
988
252
70
427
418
200
246
392
374
168
166
352
478
1020
1028
1096
1075
Soln.
vol.
(cm )
600
830
250
70
430
420
200
250
390
370
170
170
350
480
860
870
930
910
Vol.
gas
phase
(cm3)
380
150
730
910
550
560
780
730
590
610
810
810
630
500
120
110
50
70
Total
co2
(g, weighed)
0.32
0.18
1.60
16.4
1.30
1.37
10.6
9.2
1.42
9.8
16.8
0.18
0.23
0.23
Total
8 C02
KR liquid
0.45
0.18
6.35
39.2
6.50
5.57
27.0
24.6
8.6
27.8
35.1
0.18
0.21
0.21
S
from
H2S gas
Ong)
<0.01
<0.01
<0.01
<0.01
0.01
< .01
0.26
0.52
<0.01
3.1
1.8
<0.01
<0.01
<0.01
Total
S
mg
Kg liquid
12
9
3
3
2
2
1
0.2
5
9
10
5
5
10
7
196
-------
Table 3. Carbon dioxide and sulfur analysis, (continued)
Sample
number
2BHS
2BHU
2BCS
2BCU
3BHS
3BHU
3BCS
3BCU
So In.
wt.
(g)
578
1028
1017
1084
998
1002
1075
1064
Soln.
vol.
(cm )
490
870
860
920
850
850
910
900
Vol.
gas
phase
3
(cm )
490
110
120
60
130
130
70
80
Total
co2
(g, weighed)
1.60
0.21
1.79
0.27
0.28
0.18
0.23
0.43
Total
g co2
Kg liquid
3.27
0.20
1.76
0.25
0.28
0.18
0.21
0.40
S
from
H2S gas
(mg)
<0.01
<0.01
<0.0l
<0.01
<0.01
<0.01
<0.01
<0.01
Total
S
mg
Kg liquid
5
8
6
16
3
8
7
197
-------
Table 4a. Composition of "liquid" phase samples by spark source mass
spectography (sample no. 3BHS).a
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
H
He
Li
Be
B
C
N
0
F
Ne
Na
Mg
Al
Si
P
S
Cl
Ar
K
Ca
Sc
Ti
V
Cr
Mn
<0.05
92
0.8
Mb
39
0.5
24
<0.08
390
Mb
Mb
Mb
<0.5
<8
<0.8
<8
7500 (M)
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
Fe
Co
Ni
Cu
Zn
Ga
Ge
As
Se
Br
Kr
Rb
Sr
Y
Zr
Nb
Mo
Tc
Ru
Rh
Pd
Ag
Cd
In
Sn
4100 (M)b
<0.8
'1200
130
6100 (M)b
<2
5
100
£20
<2
5300 (M)b
4800 (M)b
<2
24
<2
<8
<5
<2
£5
<3
<40
<2
<20
51.
52.
53.
54.
55.
56.
57.
58.
59.
60.
61.
62.
63.
64.
65.
66.
67.
68.
69.
70.
71.
72.
73.
74.
75.
Sb
Te
I
Xe
Cs
Ba
La
Ce
Pr
Nd
Pm
Sm
Eu
Gd
Tb
Dy
Ho
Er
Tm
Yb
Lu
Hf
Ta
W
Re
<3
<5
<2
52
2600 (M)b
20
<1
<2
<3
<3
<5
<3
<1
<3
<1
<3
<1
<3
<1
<3
<8
<3
<2
76.
77.
78.
79.
80.
81.
82.
83.
84.
85.
86.
87.
88.
89.
90.
91.
92.
93.
94.
95.
96.
97.
98.
99.
100.
Os
Ir
Pt
Au
Hg
Tl
Pb
Bi
Po
At
Rn
Fr
Ra
Ac
Th
Pa
U
Np
Pu
Am
Cm
Bk
Cf
Es
Fm
<2
<2
<3
<3
<3
<2
500
<3
<1
<1
Results expressed in mg/1.
Major constituent.
198
-------
Table 4b. Composition of "Liquid" phase samples by spark source mass
spectrography (sample no. 4BHU).a
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
H
He
Li
Be
B
C
N
0
F
Ne
Na
Mg
Al
Si
P
S
Cl
Ar
K
Ca
Sc
Ti
V
Cr
Mn
0.15
140
2
Mb
27
1
9
0.6
270
Mb
Mb
Mb
<0.6
<6
<0.6
<6
1900
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
Fe
Co
Ni
Cu
Zn
Ga
Ge
As
Se
Br
Kr
Rb
Sr
Y
Zr
Nb
Mo
Tc
Ru
Rh
Pd
Ag
Cd
In
Sn
1000
<0.6
450
45
2200
<1
3
2700
6
<2
120
3700 (M)b
<1
6
<1
<6
<1
<1
<2
3
9
<1
<15
51.
52.
53.
54.
55.
56.
57.
58.
59.
60.
61.
62.
63.
64.
65.
66.
67.
68.
69.
70.
71.
72.
73.
74.
75.
Sb
Te
I
Xe
Cs
Ba
La
Ce
Pr
Nd
Pm
Sm
Eu
Gd
Tb
Dy
Ho
Er
Tm
Yb
Lu
Hf
Ta
W
Re
<3
<1
<2
20
3000 (M)b
15
<1
<2
<2
<2
<3
<3
<1
<2
<1
<2
<1
<2
<1
6
£6
<2
<1
76.
77.
78.
79.
80.
81.
82.
83.
84.
85.
86.
87.
88.
89.
90.
91.
92.
93.
94.
95.
96.
97.
98.
99.
100.
Os
Ir
Pt
Au
Hg
Tl
Pb
Bi
Po
At
Rn
Fr
Ra
Ac
Th
Pa
U
Np
Pu
Am
Cm
Bk
Cf
Es
Fm
<2
<1
<2
<2
<3
<1
450
<2
<1
<1
1Results expressed in mg/1.
""Major constituent.
199
-------
X"s
CO
e
o
H
4J
r-l
O
CO
T3
C
cd
CO
01
4J
cd
rl
p.
H
fj
,,
p.
o
CO
rl
CO
^
T~"1
C
cd
^
CO
a)
r-l
a.
I
CO
0)
CO
n)
P.
1
T)
H
3
a1
rl
hJ
1C
a)
rH
cfl
H
to
H
r
NO
to
E
-
vD
CO
C
O
H p
4-1 X
3 PP
rH CO
O
to
W
PQ
CM
to
as
1-1
01
rH
a
E
frj
to
to
X
to
K
_
NO
B)
a)
4J
rt
4->
H ;=>
A r-r"
U* ri*
O co
ai
PU
to
M
CM
to
CO
1-1
01
rH
a
E
>
-* o
rH
vO
CM
.
0
ca
in co
-------
CO
x-^v
cn
iH
en
rH
Cfl
C
cfl
0)
4J
0
3
4-1
en
C
O
o
01
M
en
cu
i
(X
e
tfl
cn
o)
cn
to
rC
CX
1
o
J
rn
3
CT
H
0)
rH
tfl
H
13
33
BQ
cn
CO
33
PQ
CN
e/3
33
PQ
rH
CO
in
PQ
'vO
to
33
H
\O
4J
a
01
6
01
rH
W
rH
60
e
O rH CN
O r~
O
O
rH
60
e
O rH CN
O r-
O
iH
vO
rH
60
e
O O CN
o ON
0
V0
iH
-*^
60
G
o o cn
o oo
0
f^
rH
60
g
O **O LO
O r-~
O
CO 60 rH
*Z S rH CN
ON
CN O m r~. CN O rH rH rH
oo rH cn
rH
cn
oo>
H cn Xi
0)
^-1
W 3
32 in
PQ en
in =cu
o i in m \o v-t
ON | in ^d" CX
CN cn CN
01
> a
01 -H
0) 1-1 rJ
x; o xi
u *g
01
0) -u
en en
3 01 m
CO rH O
O pi
6 i cfl en
CN t^ rH CN rH CO CO
cn i CN CN xi xi
cn cfl 01 ex
1 XI 0)
< O rJ T3
)-i Xi -H
a, 4-> 3
cr
en M -H
C CU rH
o x:
H 4J 0)
4-1 O X!
H 4J
13 01
e x: a
O 4J -H
CJ
O in O o )-i en
O r^ rH CN 60 O ~O
cn I CN CN C 14H -H
cn -H rH
I 4-> CO O
/-. C3 C3 C3
rH x-x -H Cfl CU
^» "^^ CJ ^"s XI S
XI 60 o 60 0) 4J XI
^ ^-^ -H cj a
cn en C )-i -H
-a C cu ex cu 01 P
HO K ^ M x: c
rH -H 3 0) 60 01
O 4-1 4J 01 UH -H
C/l Cfl CO M <4H XI 60
H rl 3 -H fi
rH O 01 cn 13 01 -H
co cx o) cx cn xi en
4J CO 4J Q O) 0) 3
O>COQ)»j XiOtO
HWQH&< H4-ICJ
^
01
, 1
e
tfl
CO
-------
Slug flow
Mixed flow
o o
o o ° ° ° ° ° 0°0
0«0"0-§° O o o°0°£
°o ° o0 o °
o°o °o °
Oo°ooo o o o
o00o0o'
0°°00o0o°o° c
o°o°oOo°o°
OOP O Q O O O O O
1 o o o o ° o
*0°o00°0°o00° c
o^o ° o o o
o°o000o «
o°o
Separate phases
Vapor
Fig. 1. Two-phase flow systems.
202
-------
00
c
I
co
00
H
203
-------
APPROACHES TO INTERPRETING ENVIRONMENTAL DATA
Donald E. Michels
Aerojet Nuclear Company
Idaho Falls, Idaho
INTRODUCTION
The sampling of geothermal effluents presents many intriguing challenges,
but in the end the question will be asked, "What does the data mean?" Does
it provide a basis for legal prosecution or defense?, does it portend detri-
ment to some biological entity?, does it provide grist for intellectual
endeavor? or is it merely another set of numbers in a report destined for
oblivion?
Oblivion is deserved when the sampling fails to correspond in a logical
way with either suspected or actual features of an environmental pattern.
The notion of pattern in space or time is crucial to the setup of any sampling
procedure. It is the meager attention sometimes given to patterns that
prompts this presentation.
The four topics discussed below (superimposed dispersion patterns,
regionalized variables, risk assessment and concepts of impact) at first
seem disconnected, but upon closer reflection can be seen to have a common
underpinning. Namely, they concern the question, "how can an obscure but
non-random dispersion pattern be coped with in the taking of samples and in
using data to convey ideas about environmental conditions? The first two of
these topics are statistical but of a viewpoint not commonly used in the
U.S.. The next concerns the mismatch between real environmental inventories
of contaminants and quantities that are advertised as dangerous. The last
show<: that the scale of pattern we are interested in will determine the very
definition of environmental impact we might construct. ,
20k
-------
These comments are more general than the geothermal context of this
meeting. They are stirred by a hope that regulatory objectives now being
formed will be both fair in the legal sense and astute with respect to
physical fact.
\
Controversy becomes very heated when environmental data is used dif-
ferently by opposing factions in an issue. Much of the heat is counterpro-
ductive; environmental data should be recognized as ambiguous, especially
in small quantities. As professionals, one goal is to minimize these ambi-
guities. It is with that motive also that the following ideas are presented.
SUPERIMPOSED DISPERSION PATTERNS
A fundamental problem in environmental monitoring is to decide whether
or how much effluent has reached a particular environmental compartment.
The two facets of this problem concern: (1) what geographical extent of
the compartment should be sampled, and (2) how does effluent material
detected by analysis relate to a postulated source? Empiricism is required
for both. Analysis of statistical distributions has been useful in some cases
to answer both facets simultaneously.
If we consider for a moment a set of samples whose contaminant content
is dominated by a single source, the distribution of values can be charac-
terized by a mean (geometric)concentration and a standard (geometric) deviation,
(The adjective "geometric" is appropriate for log-normal distributions.)
Both the mean and the standard deviation are unique attributes of the source
via the dispersion mechanism. More generally, a point source of contaminant
will decorate an environmental compartment unevenly with the result that
the contaminant contents in some places will be dominated by the point
source, and in others by factors we call background. The cases where back-
ground and effluent levels are comparable in magnitude are seldom a large
205
-------
fraction of the affected area. Thus, the statistical problem is one of
resolving superimposed distributions.
One way of attempting this resolution is diagramed in Figure 1 which is
a pair of probability plots^ '. The upper plot shows the whole data set
considered as an entity. But because the data are not distributed about
a single straight line at least one of the presumptions in making the plot
does not fit physical reality. These key presumptions are:
(1) The data must come from a homogeneous (single) distribution.
That is, the best descriptor of the set is a single mean value
combined with a single standard deviation.
(2) The distribution type, Gaussian, log-normal, Weibull, etc., has
been chosen correctly in order to properly scale the axes.
If presumption 2 were violated by the plotting, the outcome would be a
curved array of plotted data rather than the sharply jointed array in the
upper plot. Thus, we test instead for two superimposed log-normal distri-
butions. This is done by dividing the data into two groups, recomputing
percentiles, and replotting as in the lower plot. There are several ways
to judge whether the division of data, as done, leads to a good way of
describing physical reality. The smaller sub-set in this case did
conform to actual estimates of a background distribution and the excellent
linearity shows that at least the data assigned are statistically homogeneous,
Furthermore, checking the sample locations for all the samples in the higher
value sub-set showed that on a map they could be separated from the back-
ground set with a single simple line. That is, they are geographically
correlated as well as statistically correlated. One could have bypassed
the statistics and drawn a line on a map of sample locations, outlining
the area which contained the high samples and the result would have been
largely correct. The advantage of using also the statistical plot lies in
206
-------
1000
t/1
a:
O
SE
UJ
a:
I
to
a:
UJ
D-
(XI
UJ
a:
o
100
10
10 50
PERCENTIl.E
90
99
A.
THE JOINTED LINE THROUGH THE PLOTTED
DATA INDICATES TWO DISTRIBUTIONAL
MODES. THE POSSIBILITY IS REALISTIC
IN CONCEPT SINCE THE SOILS WERE
SAMPLED BECAUSE OF A SUSPECTED
CONTAMINATION ZONE, HENCE SOME
MUST REPRESENT BACKGROUND.
B.
SAME DATA AS IN (A) BUT WITH
PERCENTILES REASSIGNED ASSUMING
A MIXED DISTRIBUTION WITH LITTLE
OVERLAP
THE LOWER VALUE DISTRIBUTION CAN
BE INTERPRETED AS BACKGROUND. THE
FACT OF CONTAMINATION IS EVIDENCED
BY PRESENCE OF A DISTRIBUTION MODE
THAT IS DISTINCT FROM BACKGROUND.
GEOGRAPHICAL LIMITS OF THE
CONTAMINATION CAN BE MAPPED BY
REFERENCING THE SITES IDENTIFIED BY
THE TWO DISTRIBUTIONS.
FIGURE 1 Separation of Mixed Distributions
207
-------
reducing the number of data points that have ambiguous levels. This is
one example of how to reduce subjectivity in judging which samples do not
show influence of a suspected source.
ALEATORY VS. REGIONALIZED VARIABLES
Classical statistics concerns variables which have two properties that
are exemplified by the flipped coin. (1) the variable can be determined
an infinite number of times, at least in principle and (2) the outcome of
one determination is not functionally related to the outcome of another
determination of the same kind.
Those two properties are generally absent in environmental situations.
That is, the effluent material in a single sample is unique in the sense
that the identical amount of effluent should not be expected in a repeat
sample. Note also that the repeat sample involves a different portion of the
medium. Furthermore, the analytical results of multiple environmental samples
taken closely together would be more nearly alike than the analytical results
of similar samples taken over a broader spacing. That is, the outcomes
of the individual measures are not independent in the aleatory sense.
Because classical statistical tests are based on concepts of randomness
and independence, the results of those tests should be interpreted with
caution in regards to environmental dispersions.
Regionalized Variable
In the last two decades mining engineers and geologists largely from
(2)
France and South Africa have developed a discipline called geostatistiesv .
The focus is on ore grades and ore reserves and error estimations for those
values. There are substantial parallels between dispersions of ore components
and environmental contaminations. A key concept in geostatisties is the
regionalized variable which (1) has a definite value at each point of
208
-------
space, (2) shows a more or less continuity in its spacial variation, and
(3) may show different kinds of zonal effects. Generally concentration is
the regionalized variable which is to be studied by measuring its value
at selected points in space and time.
Variogram
A principal tool for investigating the regionalized variable across space
is the variogram. This diagram is essentially (but not exactly) a plot of
apparent standard deviation for concentration versus separation distance of
samples used to compute the standard deviation. The variogram can take several
forms but they all represent different degrees of continuity of concentration
across space, as shown in Figure 2.
For a single dispersion it sometimes happens that different variograms
apply to different directions. These situations can result either from a
depositional environment which is directionally non-uniform or from a dis-
persion mechanism that is directionally non-uniform. Environmentally, the
latter corresponds to atmospheric mixing coefficients that differ downwind,
cross-wind, and vertically.
All the variograms tend toward an asymptote-like condition at large
separations between samples. This brings in a concept important to both the
layout of sampling nets and the interpretation of data. If the sample spacing
is too small, adjacent samples are quite similar and the detail obtained
for the distribution pattern may be greater than required. This kind of
overdetermination is somewhat cost ineffective. On the other hand, if
samples are taken at too great a spacing some of the space between samples
is poorly represented by any "nearest" sample. This results in a possibility
that an important occurrence may pass unnoticed.
209
-------
Figure 2: Types of Variograms
Linear: The most common; represents a
simple ideal regionalized variable.
distance
Continuous: Small
range shows a high
The thickness of a
variability at short
continuity, for example,
sedimentary bed.
distance
Nugget effect: Distribution is discon-
tinuous in small samples, not fully
conforming to the ideal regionalized
variable.
distance ->-
Random: A limiting case corresponding
to an aleatory variable.
distance
210
Adapted from G. Matheron, 1963
-------
Early in the sampling program a variogram could be constructed from
preliminary data. Ideally, it would be used to set up a routine sample
spacing that would be sensitive to pre-selected scale of differences between
analytical values. It may turn out that sample spacing should be different
in different places as well as in different directions.
This approach can be applied to the technique for taking individual
samples as well as in the setup of an extensive net. For example a single
scoop of soil, which physically involves only a few square inches of area
may be taken to represent tens of thousands of square feet of geography.
Represent!'vity could be improved by compositing several scoops from an area
of hundreds of square feet that constitute the sample site. The question
arises, "how much separation between scoops?" In one situation, soil sampling
for Cs from fallout*- ' t several 10-cm square samples were taken and the
results used to construct a variogram. It was found that samples separated
by one meter had a relatively small standard deviation, whereas samples on
a two-meter separation showed a larger standard deviation, about the same
size as for the set as a whole. Thus, in compositing, the SCOODS should hp
separated by more than one meter, but there is not statistical advantage in
having the scoops separated by more than two meters. Notice that in compositing,
one goal is to submerge the small scale variability by swamping it with
components that are random with respect to that small scale variability
within the sample site. Contrast that goal with a main objective of a
sampling program - - the taking of (composite) samples in such a way that
non-random variability between sites becomes more apparent that it would be
by any other sample spacing.
RISK ASSESSMENT
This issue has been contaminated by "Environmental Jitters". Some
(4)
peoplev ' feel that scientists who encourage public fears on the basis of
211
-------
of incomplete or ill-digested evidence constitute a serious environmental
problem. The reference (4) concerns fluorocarbon aerosol propellents and
there are other examples from cyclamate to C^.
There is a point in the plutonium debate that seems to be seldom re-
cognized. The same point occurs in debate about other contaminants so the
intent here is not to continue the plutonium debate, but rather to show that:
(1) environmental exposures with some similarity to postulated calamities
have been underway for decades; they deserve to be assessed in terms of
transfer coefficients between media and populations, and (2) numeric
values for transfer coefficients can be estimated, at least in gross aspects
from fairly general data. These estimates should serve to limit speculation
about how severe a postulated event might be. Although this discussion is
limited to plutonium, counterpart calculations can be made for lead, cadmium,
selenium, mercury, radon etc.
Fallout from weapons testing prior to 1970, dispersed about 300,000
curies (7 tons) of Pu-239 and Pu-240^- '. This amount is large compared to
industrial losses even in serious incidents. Our expectations about what
happens to plutonium that escapes from industrial containment should be
tempered by what has happened to the plutonium from weapons fallout. The
behavior of the plutonium from the two kinds of sources will not be identical
since, for example, the particle size distributions are not alike and the
atmospheric distribution mechanisms are partly different. However, the
fallout case can serve as a reference. Whether the industrial context
suggests a more efficient or less efficient transfer of plutonium into
people can be estimated from details about the industrial context and how
they contrast with fundamentals of the fallout context.
212
-------
Autopsies of people from the general public who were living throughout
the major period of weapons fallout, show that they carried plutonium in
amounts near an average of 5x10" curies (0.5 picocuries per person)^ '.
From that result, the total amount of Pu inside living people can be estimated
by multiplying the average burden by the world population; (0.5 picocuries/
person) times (3x10 persons/world population) = 1.5x10" curies/world
population. Thus, of the entire 300,000 curies of Pu in the world environ-
-9
ment, only about five billionths (5x10" ) is actually inside people where
it exposes them to alpha radioactivity. Thus, a transfer coefficient from
_g **
environment to human bodies can be crudely estimated at 5x10 .
In applying this result to the context of malcontainment of industrial
plutonium we should consider three aspects of the contamination. First, is
the total plutonium in an installation. Second, if a release occurs only a
part of the total plutonium will actually escape into the environment before
cleanup operations repackage what can be recovered. Third, of the plutonium
which escapes, only a fraction will actually end up in people. It is the
size of this last fraction which is of ultimate concern to human health.
Whatever detriment is calculated to accrue to humanity must be based on
this smallest fraction.
*
Probably this gives an overestimate since the 0.5 picocurie figure is an
average based mainly on persons who were alive throughout the major period
of fallout from weapons. Younger people missed the major part of the fallout
episode, however, measurements of the plutonium burdens have not been included
in proportion to their population. Including young people in the population
figures of 3xl09 is conservative in the sense that the results of the calculation
will show more plutonium in the world population than what a more refined
calculation would show.
This value lumps the oceanic and land inventories of plutonium but the
distinction is largely irrelevant if most of the population burden was
acquired by inhaling primary fallout, in contrast to inhaling fallout re-
entrained from soil or ingestion through the food chain.
213
-------
Many scare comments about industrial effluents ignore the inefficiency
by which environmental levels are carried on to people. Several statements
in print say, in essence, that one ounce of plutonium can poison all the
world population. Credibility of such statements seems strained since we
do not appear to be seriously poisoned by the 7 tons already in the environment.
Distinctions must be made, of course, between materials like DDT which
increase up the food chain and those like plutonium which decrease.
More sophisticated estimates of transfer coefficients from environment
to human populations can be made. Some already exist for specific conditions
of soil content, reetrainment, and other aspects of environmental pathways^ '.
These kinds of calculations, based on simple or complex assumptions deserve
to be recognized and used publicly in describing how socially important
decisions could be based on environmental data. One proper goal is to
counter scare comments with objective comparisons between current dispersions
of contaminants and inducible changes in the patterns.
THREE CONCEPTS OF IMPACT
Part of the difficulty in laying out environmental sampling (monitoring)
programs is due to obscurity of what an impact is and how one quantifies an
impact by looking at data. There are three distinct ways of looking at
environmental impact and they require different techniques for sampling
and data interpretation. Debates about impact commonly obscure the dis-
tinctions to the detriment of resolving issues. These approaches are:
(1) Summing emissions from sources,
(2) Measuring concentrations of contaminants in air, soil, water or
biota at distance from the source,
(3) Observing the biological effects due to increased levels of con-
taminants in parts of the ecosystem.
214
-------
The third approach is of ultimate significance, of course, but conclusions
about detriment are seldom timely and early observations are difficult to
present in a convincing way to persons who are sceptical that the cause of
an affliction has been identified. Furthermore, causes of environmentally
t
induced afflictions are seldom singular and it is difficult to obtain agree-
ment upon which of the causitive factors is most important, especially in
a legal contest.
The second approach has value to a defensive legal posture since it
can identify the geographical boundaries outside of which responsibility for
effects can be denied. The results from this approach have immense academic
interest to some scientific fields, but even the best work can get a "So
what's new?" response from a person interested narrowly in the third approach.
The first approach is the handiest for legal applications (especially
prosecution) because sampling and analytical results are the least ambiguous
of all the approaches. However, it is sometimes a long environmental path
between a source and a biological effect so that the precise data from the
first approach risk being thoroughly irrevelant. For example, the chemical
form of an effluent may have a short lifetime in the environment or the
effluent may be a trivial increment to a naturally occurring background.
I have belabored the distinctions among these three approaches because
the standards of good work in one approach are quite different from the
counterpart standards in other approaches. Furthermore, persons who narrowly
champion one approach have no logical basis for debate with a representative
of another approach. Indeed, there are no reliable logical connections
between the approaches except the trivial one that emission precedes disper-
sion which precedes biological effects. Even the principle of mass conser-
vation applies unevenly because most effluents are changed chemically within
the environment. This is to say that in only a special few cases does the
215 ]
-------
integrated effluent equal the (incremental additions to) contents in
environmental compartments.
CONCLUSIONS
Some persons at this meeting have commented about how individualized
the sampling technique must be for different geothermal sources, even of the
same type. Perhaps this is due to the obscurity of important features of
basic patterns. In order to establish standard or reference procedures aimed
at regulation of geothermal sources we would be well advised to first study
the patterns involved. In this way, the standardized parts of a procedure
could be well aimed at particular features which geothermal sources have
in common in their own patterns of behavior or patterns of dispersion of their
effluents. Comparability must be one of the motives behind setting up
standardized procedures. Not all components of individual patterns are
equally useful in comparing one geothermal source with another. We should
learn which features are important to our purposes, either academic or
regulatory, and then focus the standardization of sampling toward those
features. The interpretation of geothermal data must begin before the
samples are taken.
216
-------
REFERENCES
1. D. E. Michels, Log-normal Analysis of Data for Plutonium in the Outdoors:
in Proceedings of Environmental PlutonTum Symposium, Los Alamos, Aug 4-5,
USAEL Rpt LA-4756, 1971.
2. G. Matheron, Principles of Geostatistics, Econ. Geol., V 58,
pp 1246-1266, 1963.
3. D. E. Michels, Analysis of Paired Data, Sequential in Space or Time and
the Relationship to Sampling Continuous Cyclic Distributions, USAEC
Rpt. RFP-2165, Dow Chemical, Rocky Flats Div., 1974.
4. R. Scorer, The Danger of Environmental Jitters, New Scientist,
pp 702-703, June 26, 1975.
5. J. H. Harley, Worldwide Plutonium Fallout From Weapons Tests, in
Proceedings of Environmental Plutonium Symposium, 'Los Alamos, Aug 4-5,
1971, USAEC Rpt LA-4756.
6. E. E. Campbell, et. al., Plutonium in Autopsy Tissues, USAEC Rpt LA-4875,
1973.
7. D. H. Denham, et. al., Radiological Evaluations for Advanced Waste Manage-
ment Studies: USAEC Rpt BNWL-1764, Battelle Pacific Northwest Laboratories
1973.
217
-------
Sampling and Preservation Techniques for Waters
in Geysers and Hot Springs
by
James W. Ball, Everett A, Jenne, and J. M. Burchard
with a section on
Gas Collection
by
Alfred H,' Truesdell
U, S. Geological Survey
345 Middlefield Road
Menlo Park, California 94025 .
ABSTRACT
Sampling of geothermal fluids presents unique challenges, due to
the instability and wide concentration range of many constituents.
Unstable parameters, such as pH, Eh, temperature and dissolved oxygen.
are determined on-site. Filtration of geothermal water is done with
absolute minimum time-delay and exposure of the sample to air, by
using a portable pump and non-contaminating, all-plastic or all-metal
filter apparatus, for samples for inorganic or organic determinations,
%
respectively. Sample preservatives and their safe, efficient
transportation are discussed. Two techniques of geothermal gas
collection are presented along with details of gas sampling apparatus.
218
-------
Sampling and Preservation Techniques for Waters
in Geysers and Hot Springs
by
James W. Ball, Everett A. Jenne, and J. M. Burchard
with a section on
Gas Collection
by
Alfred H. Truesdell
U. S. Geological Survey
345 Middlefield Road
Menlo Park, California 94025
INTRODUCTION
Unique problems are encountered in the sampling, preservation
and analysis of geothermal fluids because of the instability and/or
high concentrations of some constituents. Silica may occur at
concentrations up to several hundred milligrams per litre; therefore,
silica may polymerize during sample storage and not react with
molybdate color forming reagents, or precipitate, with concurrent loss
+3 +2
of coprecipitating elements. Arsenic , iron and other variable
valence ions may be rapidly oxidized upon aeration at elevated
temperatures, precluding valence species determinations. Sulfides may
be rapidly oxidized by dissolved oxygen, and hydrogen sulfide gas is
unstable in the presence of moisture, oxygen, and to a les'ser extent,
ultraviolet light. The pH may rise rapidly (up to two orders of
magnitude in our experience) due to exsolution of dissolved carbon
219
-------
dioxide, or decrease due to carbon dioxide uptake or hydrogen sulfide
oxidation.
Due to the greater cost, and the tendency to lowered precision
and accuracy, of field analyses as compared to laboratory analyses,
analyses should be done on site for only those constituents which
cannot be preserved for laboratory analysis. Exceptions may occur
when knowledge of some parameter is needed to guide subsequent sampling,
and when adequate analytical tools are available. pH, Eh, and
temperature are, of course, determined on site. If necessary,
carbonate-bicarbonate alkalinity can be determined later as long as
the non-carbonate alkalinity is not significant and the pH is also
remeasured (Ellis, et al., 1968; A. H. Truesdell, unpub. data).
Alternatively, immediate alkalinity titration may be difficult because
at the time of sampling even gentle stirring promotes the vigorous
exsolution of supersaturated carbon dioxide resulting in pH increase
(equation 1)
H+ + HCO~ ^ H20 + C02"K 1)
even while acid is being added.
Due to the instability of the dissolved constituents of geothermal
fluids, immediate filtration of the hot water without allowing
degassing of waters supersaturated with carbon dioxide is necessary
(equation 1). The high concentration of trace elements in the sinter
and.other surficial deposits mandates stringent precautions to avoid
water sample contamination.
220-
-------
Methods of collection and analysis of geothermal waters have been
published by Ellis, et al. (1968) and by Presser and Barnes (1974).
Geothermal gas collection and analysis schemes have been described
by Ellis, et al. (1968), Akeno (1973), and Giggenbach (1976). The
entire field has been reviewed by Finlayson (1970).
221
-------
WATER COLLECTION
ON SITE ANALYSES
Determination of pH and Eh is accomplished elegantly by pumping
sample fluid (slowly for pH, rapidly for Eh) through an insulated cell
(Fig. 1) containing pH, Eh and temperature probes. The probes are of
rugged, high-impact plastic construction, and glass electrode membranes
are well protected. Liquid junctions are of a type highly resistant to
clogging and fouling, and the electrodes are usable up to and including
100 C without damage. A meter with compatible electrodes is used to
Fig. 1. Flow-through pH-Eh cell.
measure both pH and Eh. If equipped with intercept, slope, and
temperature compensation knobs, the meter may be standardized with
bracketing buffer solutions at ambient temperature prior to each
measurement (Sargent-Welch Scientific Co., undated). Otherwise
222
-------
standardization with buffers at the sample temperature is necessary.
Actual Eh values are calculated later from the EMF value, the half-cell
potential of the reference cell and the temperature. The lid of the
meter's carrying case (Fig. 2) serves as a sun shield to minimize
temperature changes in the meter.
Dissolved oxygen is determined by modified Dinkier titration;
manganous sulfate and alkali-iodide-azide contained in sealed pieces
of plastic tubing are added on site. The samples are thereafter kept
out of direct sunlight. The sulfamic acid is added just before
analysis, which is carried out as soon as possible.
223
-------
SAMPLING
Due to frequent remoteness of sampling sites, compact, lightweight
equipment is used which may be transported, along with the samples
taken, in two backpacks (weight £ 20 kg each). The portable pump
(Fig. 2) used to pump sample water through the flow-through pH-Eh cell
and the filter apparatus, is capable of delivering the sample fluid
from a depth of at least three meters to the_ apparatus with a head of
2
at least 20 psig (1.4 kg/cm ) for at least four continuous hours, the
most extreme conditions encountered thus far.
Fig. 2. Field setup, showing sampling line, portable pump,
flow-through pH-Eh cell, telethermometer and pH meter,
The polyvinyl tubing through which sample water is pumped is held
in place using a 2.5 cm o.d. sectioned aluminum pole having an
adjustable stainless steel laboratory clamp attached to one end (Fig. 3).
22k
-------
Fig. 3. Sampling line in use in hot spring overflow.
The sampling line inlet is weighted to prevent flotation of the poly-
vinyl tubing. The effectiveness of the pole for positioning the
sample tubing over a small spouting hot spring is shown in Fig. 4.
Fig. 4. Positioning the sampling line over spring.
225
-------
The all-plastic filter used (Fig. 5) holds a 0.1 ym 142 mm
membrane between two plexiglass discs sealed with a viton rubber
o-ring; the air is released at the start of filtration using an
integral valve; and the assembly is supported by three polyvinyl
chloride legs threaded into the underside of the bottom disc. The
two discs are secured together by integral, .swing-away nylon
bolts and nuts hinged in the bottom plate.
Sample fluid is pumped from the source through the filter assembly
into all-polyethylene or glass collection bottles (Fig. 5). The
membrane may be saved by placing it in an acid-washed petri dish.
Cleaning consists of wiping the inside of the filter with tissue,
rinsing with distilled water, and changing the filter membrane; sample
water from the new site is used to thoroughly flush the sampling line
during pH and Eh measurements; the sampling line is then attached to
the filter and the membrane is flushed with 250-500 ml of sample.
Fig. 5. Filtration assemblies.
226
-------
Samples are collected after on-site analyses are completed. Samples
for major cations(250 ml), major anions(500 ml), nutrients(250 ml),
sulfide(250 ml), and iron 2/3 and arsenic 3/5(250 ml) are collected
first; then samples for Hg; then samples for trace elements. Duplicate
samples may be taken for trace constituents.
Samples for mercury determination are collected in a 250, 500 or
1000 ml borosilicate glass reagent bottle, any of which is compatible
with our mercury analytical system. The mercury sample bottles are
cleaned with an overnight soaking with chromic acid, rinsed five times
with distilled water and oven dried at least two hours at >200 C
to drive off residual mercury.
Samples to be analyzed for total dissolved organic carbon are
pumped through a silver filter membrane housed in a cylindrical filter
assembly of stainless steel and Teflon construction (Fig. 6) (Malcolm
and Leenheer, 1973). The membrane is flushed with approximately 200 ml
sample water; 30-40 ml of filtrate is then collected in a glass bottle
which has been fired at 500 C for six hours and capped with fired
aluminum foil (Malcolm and Leenheer, 1'973).
An effective way to avoid contamination of the sample water by
surrounding sinter is use of a nylon-reinforced polyethylene ground
sheet. The operator and all his equipment fit on a 10-foot square
sheet (Fig. 1, 2, 5).
227
-------
Fig. 6. Organic carbon filtration unit.
228
-------
SAMPLE PRESERVATION
Samples for major cation, iron, and arsenic analysis are
acidified with 2 ml hydrochloric acid/250 ml; samples for mercury
analysis are treated with 10 ml 5% potassium permanganate/400 ml
(Avotins and Jenne, 1975); trace element samples are acidified with
5 ml redistilled nitric acid/1. The pH of acidified samples is
checked at the end of the day to assure that a pH of <1.2 has been
reached. The above three preservatives are transported to the field
in standard borosilicate glass ampules, avoiding the risk of
contamination, total loss or injury associated with single-container
storage. The nitric acid must be stored in an ampule with scored
constriction, as the color-break ring of the standard ampules contains
some trace elements, notably lead, which will contaminate a trace-element
p *
sample.
Sulfide samples are fixed by sequentially adding 2 ml of 1 M
zinc acetate and 2 ml of 1 M sodium hydroxide/250 ml (Amer. Public Hlth. Assoc.,
1970). Since the former solution contains a large amount of zinc, a
frequently determined metal, it is stored in a septum-stoppered bottle
and dispensed with a disposable syringe, both of which are stored in
a zip-lock bag. Sulfide and dissolved organic carbon samples are
chilled as soon as possible after collection, and held at 4 C until
analysis. Aliquots for silica analyses are normally diluted 10-fold
at the end of the day of collection, but samples containing >700 mg/1
silica must be diluted immediately (M. Thompson, oral cornrn., 1976). Nutrient
samples are frozen as soon as possible with dry ice» and kept frozen
until analysis.
229
-------
G/S COLLECTION
Geothermal gases (other than steam) generally consist of carbon
dioxide with lesser quantities of nitrogen, methane, hydrogen sulfide,
hydrogen, oxygen, C--C, hydrocarbons, and inert gases in approximate
order of decreasing abundance.
Two methods of gecthermal gas collection have been tested. One
method, described in more detail elsewhere (Truesdell and Pering, 1974),
utilizes syringes for both volume measurement and gas absorption in
sodium hydroxide, and is particularly suited for partial field analysis.
The other method, in which absorption in sodium hydroxide takes place
in an evacuated bottle, is suited for isotope analysis wherein larger
quantities of gas are required. In the syringe method, hydrogen
sulfide is fixed by absorption in sodium hydroxide solution followed
by immediate precipitation using cadmium acetate or by eventual total
oxidation to sulfate. The sodium hydroxide also absorbs carbon dioxide,
greatly reducing the bulk of the gas sample, and the reduction in
volume of gas accurately indicates the ratio of carbon dioxide plus
hydrogen sulfide to other gases. If cadmium sulfide is precipitated,
the intensity of the yellow color gives a rough indication of the
relative quantity of hydrogen sulfide.
The syringe apparatus (Fig. 7) is flushed with* spring water and
10 to 20 ml of 3 M sodium hydroxide solution is introduced- into
syringe B. Gas is drawn into syringe A, its volume measured by noting
the change of position of the plunger and it is pumped into syringe B.
230-
-------
With vigorous shaking of this syringe, carbon dioxide and hydrogen
sulfide are absorbed and the decrease in volume gives a field analysis
for these gases which has been found to be within ± 1% of the amount
found absorbed in the sodium hydroxide. When 25 ml of residual gas
have accumulated, these are transferred into the evacuated gas bottle.
The procedure is repeated until the gas bottle is full. Detailed
descriptions of the construction of the apparatus and of its manipula-
tion during sampling are given by Truesdell and Pering (1974). After
sample collection, half of the sodium hydroxide solution is saved for
carbon dioxide analysis and half is treated with cadmium acetate to
preserve hydrogen sulfide.
The second method used for gas collection is similar in principle.
50 ml of sodium hydroxide solution is contained in an evacuated gas
bottle. When the tubing and funnel are flushed with water and filled
with gas, the bottle is attached and opened so that the gas bubbles
through the caustic solution. The carbon dioxide and hydrogen sulfide
are absorbed rapidly during shaking. The collection is complete when
no more gas is absorbed in a reasonable time. This method is a
simplification of that proposed by Giggenbach (1976) for volcanic
gases.
231
-------
Syringe Syringe
A B
(Pump) (Caustic)
3M
NaOH
Handle indicates
off direction
50ml gas
sample bottle
Detail of
Stopcock
Fig, 7. Syringe apparatus for gas collection.
232
-------
9. 1267
10-
11
12
13
14
15-
16
17
18
19
20-
21
22
REFERENCES
Akeno, T., 1973, Rapid chemical analysis of volcanic gases in geo-
thernal fields (in Japanese): Jour. Japan Geotherraal Energy
Assoc., v. 10, no. 1, p. 13-21.
American Public Health Association, 1970, Standard Methods for the
Examination of Water and Wastewater, 13th ed.: Araer. Public Hlth
Assoc., Washington, D. C., p. 336.
Avotins, P. and Jenne, E. A., 1975, The time stability of dissolved
mercury in water samples II. Chemical stability: J. Environ.
Qual., v. 4., no. 4, p. 515.
Ellis, A. J., Mahon, W. A. J., and Ritchie, J. A., 1968, Methods of
collection and analysis of geothermal fluids, 2nd edition:
Chemistry Division, N. Z. Dept. Sci., Ind. Res. Report CD 2103,
51 p.
Finlayson, J. B., 1970, The collection and analysis of volcanic and
hydrotherraal gases: Geotherraics, Special Issue 2, v. 2, pt. 2,
p. 1344-1354.
Giggenbach, W. F., 1976, A simple method for the collection and
analysis of volcanic gas samples: Bull. Volcanol. (in press).
Malcolm, R. L., and Leenheer, J. A., 1973, The usefulness of organic
carbon parameters in water quality .investigations: Proc.
Inst. Environ. Sci. 1973 Ann. Meet., Anaheim, CA, p. 336-340.
Presser, T. S., and Barnes, Ivan, 1974, Special techniques for deter-
mining chemical properties of peothermal water: U.S. Geol. Survey,
233
U. S GOVERNMENT PIIINTING OFFICE : 1972 O - «7-084
66T-IC,
-------
Sargent-Welch Scientific Company, undated, Instruction manual for
Sargent-Welch pH meters: Sargent-Welch Scientific Company, 23 p.
Truesdell, A. H,, and Pering, K. L., 1974, Geothermal gas sampling
methods: U.S. Geol. Survey Open-File Report 74-361, 6 p.
234
-------
TECHNICAL REPORT DATA
(Please read Inductions on the reverse before completing)
1 REPORT NO.
EPA-600/9-76-011
4. TITLE ANDSU8TITLE
PROCEEDINGS OF THE FIRST
GEOTHERMAL EFFLUENTS
WORKSHOP ON SAMPLING
3. RECIPIENT'S ACCESSION NO.
5. REPORT DATE
1Q76
6. PERFORMING ORGANIZATION CODE
7 AUTHOR(S)
(Workshop participants)
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Environmental Monitoring and Support Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Las Vegas, NV 89111*
10. PROGRAM ELEMENT NO.
EHE624
11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
13. TYPE OF REPORT AND PERIOD COVERED
Proceedings Oct. 1975
Same as above
14. SPONSORING AGENCY CODE
EPA-ORD Office of Energy,
Minerals and Industry
15. SUPPLEMENTARY NOTES
16. ABSTRACT
This is a compilation of papers presented at the first in a series of workshops
on environmental monitoring of geothermal energy development held on October 20 and
21, 1975 at the U.S. Environmental Protection Agency's Environmental Monitoring and
Support Laboratory in Las Vegas, Nevada. The purpose of this workshop was to gen-
erate the exchange of ideas and knowledge needed to develop a set of standard geo-
thermal sampling methods with assurance of quality in those methods. Representatives
of industry, universities, and government presented 19 technical papers, 12 of which
are published in this document. Their content and the discussions which followed the
presentations provided guidance for developing a recognized Referenced Sampling
Method Handbook.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
moni tors
pol1ut ion
geothermal energy
methods development
13B
UA
14B
11»C
IAD
8. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (ThisReport)
UNCLASSIFIED
21. NO. OF PAGES
20. SECURITY CLASS (Thispage)
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (9-73)
-------
.
s
------- |