United States
Environmental Protection
Agency
Environmental Monitoring
and Support Laboratory
PO Box 15027
Las Vegas NV 89114
EPA-600 7-78-121
June 1978
&EPA
Research and Development
Proceedings of the
Second Workshop
on Sampling
Geothermal Effluents
February 15-17, 1977
Las Vegas, Nevada
Interagency
Energy-Environment
Research
and Development
Program Report
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad categories
were established to facilitate further development and application of environmental
technology. Elimination of traditional grouping was consciously planned to foster
technology transfer and a maximum interface in related fields. The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the INTERAGENCY ENERGY—ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports in this series result from the effort
funded under the 17-agency Federal Energy/Environment Research and Development
Program. These studies relate to EPA'S mission to protect the public health and welfare
from adverse effects of pollutants associated with energy systems. The goal of the Pro-
gram is to assure the rapid development of domestic energy supplies in an environ-
mentally-compatible manner by providing the necessary environmental data and
control technology. Investigations include analyses of the transport of energy-related
pollutants and their health and ecological effects; assessments of, and development of,
control technologies for energy systems; and integrated assessments of a wide range
of energy-related environmental issues.
This document is available to the public through the National Technical Information
Service, Springfield, Virginia 22161
-------
EPA-600/7-78-121
June 1978
PROCEED I NGS
OF THE
SECOND WORKSHOP ON SAMPLING GEOTHERMAL EFFLUENTS
February 15-17, 1977
Las Vegas, Nevada
Sponsored by
Environmental Monitoring and Support Laboratory
Las Vegas, Nevada
Industrial Environmental Research Laboratory
Cincinnati, Ohio
Office of Energy, Minerals and Industry
Washington, D.C.
of the
U.S. Environmental Protection Agency
prepared by
Geonomics, Inc.
3165 Adeline Street
Berkeley, California 94703
Contract No. 68-03-2468
Project Officer
Donald B. Gilmore
Environmental Monitoring and Support Laboratory
Las Vegas, Nevada 89114
ENVIRONMENTAL MONITORING AND SUPPORT LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
LAS VEGAS, NEVADA 89114
-------
DISCLAIMER
This report has been reviewed by the Environmental Monitoring
and Support Laboratory-Las Vegas, U.S. Environmental Protection
Agency, and approved for publication. Approval does not signify
that the contents necessarily reflect the views and policies of
the U.S. Environmental Protection Agency, nor does mention of
trade names or commercial products constitute endorsement or
recommendation for use.
11
-------
FOREWORD
Protection of the environment requires effective regulatory
actions which are based on sound technical and scientific infor-
mation. This information must include the quantitative descrip-
tion and linking of pollutant sources, transport mechanisms, in-
teractions, and resulting effects on man and his environment.
Because of the complexities involved, assessment of specific
pollutants in the environment requires a total systems approach
which transcends the media of air, water, and land. The
Environmental Monitoring and Support Laboratory-Las Vegas con-
tributes to the formation and enhancement of a sound monitoring
data base for exposure assessment through programs designed to:
develop and optimize systems and strategies
for monitoring pollutants and their impact
on the environment
demonstrate new monitoring systems and tech-
nologies by applying them to fulfill special
monitoring needs of the Agency's operating
programs
This report contains the Proceedings of the Second Workshop
on Sampling Geothermal Effluents. Collectively, the papers pre-
sented here provide much insight into the complex subject of geo-
thermal effluent sampling. It is hoped that these Proceedings
prove of value in helping to continue an exchange of ideas among
those involved in the development of geothermal resources. For
further information on this and future workshops on this sub-
ject, contact the Monitoring Systems Design and Analysis Staff
(MSA) of the Environmental Monitoring and Support Laboratory,
Las Vegas, Nevada.
The Environmental Monitoring and Support Laboratory-Las
Vegas wishes to take this opportunity to express its apprecia-
tion and gratitude to those organizations and persons who gave
so freely of their time and resources to make this workshop a
success.
George &. Mor
Director
Environmental Monitoring and Support Laboratory
Las Vegas
iii
-------
CONTENTS
Not all the papers presented at the Workshop were received
in time for publication in these proceedings. They are listed
in the Program, but not in the Contents. Those papers for
which only an abstract was received are marked with an asterisk
(*) in the Contents. These abstracts are included in this pro-
ceedings document. --
JT acre
Foreword
Program
List of Registered Participants
List Of Abbreviations and Symbols
Metric Conversion Table
Acknowledgement
*The Use of Gas Sampling Bags for the Collection and
Storage of Hydrothermal Gases 1
A. J. Soinski, LFE Corporation, Richmond, California
*Heavy Metal Emissions from Geothermal Power Plants 2
D. E. Robertson, E. A. Crecelius, J. S. Fruchter
and J. D. Ludwick, Battelle-Northwest Laboratories,
Richland, Washington
The Dynamic Measurement of Ambient Airborne Gases
Near Geothermal Areas 3
A. F. Jepsen, Eureka Resource Associates, Incorporated,
Berkeley, California,and L. Langan, Environmental
Measurements, Incorporated, San Francisco, California
*Analysis of Radon in Geothermal Effluents 14
P. Kruger, Civil Engineering Department, Stanford
University, Stanford, California
*Noble Gas Sampling of the Geothermal Heat Loop at Niland
Field, California 15
M. O'Connell, ORP, U.S. EPA, Las Vegas, Nevada
A Review of the Determination of Hydrogen Sulfide in
Air by the Cadmium Hydroxide-Stractan Colorimetric Method:
Current Practices and Modifications 16
R. A. McCurdy and S. L. Altshuler, Pacific Gas and
Electric Company, San Ramon, California
IV
-------
*The Measurement and Distribution of Geothermal Sulfur
Pollutants in The Geysers Area 42
L. A. Cavanagh, Stanford Research Institute, Menlo Park,
California
*Field Determination of Hydrogen Sulfide 42
G.A. Frye and D. W. Wheeler, Aminoil USA, Incorporated,
Santa Rosa, California
*Chevron Plans for On-Site Sampling and Analyzing of
Geothermal Waters 43
W. J. Subcasky, Chevron Oil Field Research Company,
La Habra, California
*Development of a Standard Methods Manual for Sampling
and Analysis of Geothermal Fluids and Gases 44
E. M. Woodruff, Battelle-Northwest Laboratories,
Richland, Washington
Sampling and Characterization of Suspended Solids in
Brine from Magmamax #1 Well 46
J. H. Hill and C. H. Otto, Jr., Lawrence Livermore
Laboratory, Livermore, California
Sampling and Analysis of Hot and Cold Spring Waters
and Associated Rock and Soil Samples from Potential
Geothermal Resources Areas in North Central Nevada 55
H. R. Bowman, H. Wollenberg, F. Asaro and A. Hebert,
Lawrence Berkeley Laboratory, Berkeley, California
*Methodology for Sampling and Analysis of High Pressure
Steam Lines 59
W. Hamersma, TRW Incorporated, Redondo Beach, California
How Can Standard Methods for Sampling and Analysis
Support the Motives Behind the Regulations? 70
D. E. Michels, EG&G Idaho, Incorporated, Idaho Falls,
Idaho
A Review of the Chemical Composition of Geothermal
Effluents 84
F. Tsai, S. Juprasert and S. K. Sanyal,
Geonomics, Incorporated, Berkeley, California
Development of Geothermal Gas Sampling Equipment 97
F. B. Tonani, Geochemex, Berkeley, California.
*Sampling for Reservoir Fluid Reconstruction 113
M. J. Reed, U. S. Geological Survey, Menlo Park,
California
-------
Page
Techniques of Pressure and Temperature
Measurement and Sampling in Geothermal Wells 114
L. W. Ross, L. L. Brown and R. E. Williams/
Denver Research Institute,
Denver, Colorado
Collection of Chemical, Isotope and Gas Samples
from Geothermal Wells
N. L. Nehring and A. H. Truesdell,
U. S. Geological Survey, Menlo Park, California
Geothermal Downhole Sampling Instrumentation 141
R. Fournier and J. M. Thompson, U. S.
Geological Survey, Menlo Park, California
*Case History of Geothermal Effluents Sampling
for Four Geothermal Wells 143
G. R. Conner and P. B. Needham, Jr.,
College Park Metallurgy Research Center,
Bureau of Mines, U.S. Department of the
Interior, College Park, Maryland
Estimation of Pollutant Characteristics from
Geochemical Surface Investigations 145
F. B. Tonani, Geochemex,and H. T. Meidav, Geonomics,
Incorporated, Berkeley, California
Sampling and Analysis of Geothermal Brines from
Niland Field, California 160
H. K. Bishop, San Diego Gas & Electric Company,
San Diego, California
Interpretation of Analytical Results, Thermal and
Nonthe.rmal Waters, Lava Plateaus Region of Northeastern
California and Southern Oregon 166
C. W. Klein and J. B. Koenig, GeothermEx,
Incorporated, Berkeley, California
*Geochemical Analysis of Fluids Circulated Through a
Granitic Hot Dry Rock Geothermal System 174
J. W. Tester, C. Grigsby, C. Holley and L. Blatz,
Los Alamos Scientific Laboratory of the University
of California, Los Alamos, New Mexico
Chemical Profile of the East Mesa Geothermal Field ,
Imperial Valley, California 175
R. T. Littleton and E. Burnett, U-S. Bureau of
Reclamation, Boulder City, Nevada
VI
-------
Page
Status Report, Raft River Project, Sampling,
Analysis, and Environmental Effects Studies 190
A. C. Allen, J. M. Baldwin and R. E. McAtee^
Allied Chemical Corporation, Idaho Falls,
Idaho
Borehole Geophysical Logging as Complement to
Well Effluent Sampling 211
S. K. Sanyal and R. B. Weiss, Geonomics,
Incorporated, Berkeley, California
*Use of Radioactive Tracers in Geothermal Operations 217
0. Vetter, Vetter Associates, Laguna Beach,
California
The LBL Geothermal Brine Data Compilation Project 218
S. R. Cosner and J. A. Apps
Lawrence Berkeley Laboratory,
Berkeley, California
* Analysis of Trace Contaminants in Low to Medium
Salinities Geothermal Fluids
K. Y. Chen, University of Southern California/
Los Angeles, California
Author Index 225
VII
-------
Second Workshop on
SAMPLING AND ANALYSIS OF GEOTHERMAL EFFLUENTS
February 15-17, 1977
Las Vegas, Nevada
Sponsored by
Environmental Monitoring and Support Laboratory
Las Vegas, Nevada
Industrial Environmental Research Laboratory
Cincinnati, Ohio
Office of Energy, Minerals, and Industry
Washington, D. C.
of the
U. S. Environmental Protection Agency
PROGRAM
Tuesday, 15 February
Introductory Session
9:00 - 9:10 a.m. WELCOME ADDRESS
G. B. Morgan, EMSL, U.S. EPA, Las Vegas,
Nevada
9:10 - 9:25 a.m. NEED FOR STANDARD METHODOLOGIES IN GEO-
THERMAL EFFLUENT SAMPLING AND ANALYSIS
A. C. Trakowski, Jr., Office of Monitor-
ing and Technical Support, U.S. EPA,
Washington, D. C.
9:25 - 9:40 a.m. THE ROLE OF EPA'S INDUSTRIAL ENVIRON-
MENTAL RESEARCH LABORATORY IN GEO-
THERMAL DEVELOPMENT
R. P. Hartley, IERL, U.S. EPA,
Cincinnatif Ohio
9:40 - 9:55 a.m. ERDA'S ROLE IN GEOTHERMAL ENVIRON-
MENTAL IMPACT RESEARCH
A. Jelacic, Division of Geothermal
Energy, U.S. ERDA, Washington, D. C.
9:55 - 10:10 a.m. ELECTRICAL POWER RESEARCH INSTITUTE
ROLE IN GEOTHERMAL ENVIRONMENTAL
IMPACT RESEARCH
V. W. Roberts, EPRI, Palo Alto,
California
viii
-------
10:10 - 10:30 a.m.
10:30 - 10:55 a.m.
10:55 - 11:20 a.m.
(paper withdrawn)
11:20 - 11:45 a.m.
11:45 - 12:10 p.m.
12:10 - 1:40 p.m.
1:40 - 2:05 p.m.
2:05 - 2:30 p.m.
(paper withdrawn)
2:30 - 2:55 p.m.
COFFEE
Gas Sampling and Analysis Session
Session Chairman: P. Kruger, Stanford
University, Stanford, California
THE USE OF GAS SAMPLING BAGS FOR THE
COLLECTION AND STORAGE OF HYDROTHERMAL
GASES
A. J. Soinski, LFE Corporation,
Richmond, California
SOURCE TERM ANALYSIS OF NON-CONDENSIBLE
GASES IN THE IMPERIAL VALLEY —
MEASUREMENT METHODOLOGY AND RESULTS
P. Phelps, Lawrence Livermore Laboratory,
Livermore, California
HEAVY METAL EMISSIONS FROM GEOTHERMAL
POWER PLANTS
D. E. Robertson, Battelle-Northwest
Laboratories, Richland, Washington
THE DYNAMIC MEASUREMENT OF AMBIENT AIR-
BORNE GASES NEAR GEOTHERMAL AREAS
A. F. Jepsen, Eureka Resource Associates,
Incorporated, Berkeley, California, and
L. Langan, Environmental Measurements,
Incorporated, San. Francisco, California
LUNCH
Noble Gases Session
Session Chairman: R. P. Hartley, IERL,
U.S. EPA, Cincinnati, Ohio
ANALYSIS OF RADON IN GEOTHERMAL EFFLU-
ENTS
P. Kruger, Civil Engineering Department,
Stanford University, Stanford, California
RADON MEASUREMENTS AT THE GEYSERS
P. Phelps, Lawrence Livermore Laboratory,
Livermore, California
NOBLE GAS SAMPLING OF THE GEOTHERMAL
HEAT LOOP AT NILAND FIELD, CALIFORNIA
M. O'Connell, ORP, U.S. EPA,
Las Vegas, Nevada
IX
-------
2:55 - 3:20 p.m.
3:20 - 3:40 p.m.
3:40 - 4:05 p.m.
4:05 - 4:20 p.m.
4:20 - 4:35 p.m.
(paper withdrawn)
4:35 - 5:35 p.m.
5:35 - 6:35 P.M.
Wednesday, 16 February
9:00 - 9:25 a.m.
9:25 - 9:50 a.m.
Sulfide Session
Session Chairman: D. B. Gilmore, EMSL,
U.S. EPA, Las Vegas, Nevada
A REVIEW OF THE DETERMINATION OF HYDROGEN
SULFIDE IN AIR BY THE CADMIUM HYDROXIDE-
STRACTAN COLORIMETRIC METHOD: CURRENT
PRACTICES AND MODIFICATIONS
R. A. McCurdy and S. L. Altshuler,
Pacific Gas & Electric,
San Ramon, California
COFFEE
THE MEASUREMENT AND DISTRIBUTION OF GEO-
THERMAL SULFUR POLLUTANTS IN THE GEYSERS
AREA
L. A. Cavanagh, Stanford Research Insti-
tute, Menlo Park, California
FIELD DETERMINATION OF TOTAL SULFIDES
G. Frye and D. w. Wheeler, Aminoil_USA ,
Incorporated, Santa Rosa, California
AMBIENT H S MEASUREMENTS IN THE IMPERIAL
VALLEY
P. Phelps or alternate, Lawrence Liver-
more Laboratory, Livermore, California
PANEL DISCUSSION
Chairman: V. W. Roberts, EPRI,
Palo Alto, California
Topic: "Gas Sampling and Analysis"
NO HOST COCKTAIL
General Sampling & Analysis Session
Session Chairman: B. Wiersma, EMSL,
U.S. EPA, Las Vegas, Nevada
CHEVRON PLANS FOR ON-SITE SAMPLING AND
ANALYSING OF GEOTHERMAL WATERS
W. J. Subcasky, Chevron Oil Field Re-
search Company, La Habra, California
DEVELOPMENT OF A £fANDARD METHODS MANUAL
FOR SAMPLING AND ANALYSIS OF GEOTHERMAL
FLUIDS AND GASES
E. M. Woodruff, Battelle-Northwest
Laboratories, Richland, Washington
-------
9:50 - 10:15 a.m.
10:15 - 10:35 a.m.
10:35 - 11:00 a.m.
11:00 - 11:25 a.m.
11:25 - 11:45 a.m.
11:45 - 12:00 p.m.
12:00 - 12:15 p.m.
(paper withdrawn)
12:15 - 1:45 p.m.
1:45 - 2:10 p.m.
(paper withdrawn)
2:10 - 2:35 p.m.
SAMPLING AND CHARACTERIZATION OF SUS-
PENDED SOLIDS IN BRINE FROM MAGMAMAX #1
WELL
J. H. Hill, Lawrence Livermore Labora-
tory, Livermore, California
COFFEE
SAMPLING AND ANALYSIS OF HOT AND COLD
SPRING WATERS AND ASSOCIATED ROCK AND
SOIL SAMPLES FROM POTENTIAL GEOTHERMAL
RESOURCES AREAS IN NORTH CENTRAL NEVADA
H. R. Bowman, H. Wollenberg, F. Asaro
and A. Hebert, Lawrence Berkeley Labora-
tory, Berkeley, California
METHODOLOGY FOR SAMPLING AND ANALYSIS OF
HIGH PRESSURE STEAM LINES
W. Hamersma, TRW Incorporated, Redondo
Beach, California
HOW CAN STANDARD METHODS FOR SAMPLING
AND ANALYSIS SUPPORT THE MOTIVES BEHIND
THE REGULATIONS?
D. E. Michels, EG&G Idaho, Incorporated,
Idaho Falls, Idaho
A REVIEW OF THE CHEMICAL COMPOSITION OF
GEOTHERMAL EFFLUENTS
F. Tsai, S. Juprasert and S. K. Sanyal,
Geonomics, Incorporated, Berkeley,
California
GEOTHERMAL SAMPLING AND ANALYSIS EXPE-
RIENCE IN EL SALVADOR
G. Cue'llar and J. Molina, Comision Elec-
trica Rio Lempa, Soyapango, El Salvador
LUNCH
Sampling Methodology and Instrumentation
Session
Session Chairman: E. Bretthauer, EMSL,
U.S. EPA, Las Vegas, Nevada
GEOTHERMAL WATER SAMPLING IN THE IMPERIAL
VALLEY —WHAT, WHERE, AND HOW TO MEASURE
P. Phelps, Lawrence Livermore Laboratory,
Livermore, California
DEVELOPMENT OF GEOTHERMAL GAS SAMPLING
EQUIPMENT
F. B. Tonani, Geochemex, Incorporated,
Berkeley, California
XI
-------
2:35 - 3:00 p.m.
3:00 - 3:25 p.m.
3:25 - 3:45 p.m.
3:45 - 4:10 p.m.
4:10 - 4:35 p.m.
4:35 - 5:35 p.m.
SAMPLING FOR RESERVOIR FLUID RECONSTRUC-
TION
M. J. Reed, U. S. Geological Survey,
Menlo Park, California
TECHNIQUES OF PRESSURE AND TEMPERATURE
MEASUREMENT AND SAMPLING IN GEOTHERMAL
WELLS
L. W. Ross, L. L. Brown and
R. E. Williams,
Denver Research Institute,
Denver, Colorado
COFFEE
COLLECTION AND ANALYSES OF GAS AND ISO-
TOPE SAMPLES FROM GEOTHERMAL WELLS
N. L. Nehring and A. H. Truesdell, U. S.
Geological Survey, Menlo Park, California
GEOTHERMAL DOWNHOLE SAMPLING INSTRU-
MENTATION
R. Fournier and J. M. Thompson, U. S.
Geological Survey, Menlo Park, California
PANEL DISCUSSION
Chairman: E. Bretthauer, EMSL, U.S.
EPA, Topic: "Liquid Sampling and
Analysis"
Thursday, 17 February
9:00 - 9:25 a.m.
(talk cancelled)
9:25 - 9:50 a.m.
Analytical Methods, Results, Interpre-
tation Session
Session Chairman: C. Wayman, Region
VIII, U.S. EPA, Denver, Colorado
ANALYSIS OF TRACE CONTAMINANTS IN LOW TO
MEDIUM SALINITIES GEOTHERMAL FLUIDS
K. Y. Chen, Environmental Engineering
Program, University of Southern
California, Los Angeles, California
CASE HISTORY OF GEOTHERMAL EFFLUENT
SAMPLING FOR FOUR GEOTHERMAL WELLS
G. R. Conner and P. B. Needham, Jr.,
College Park Metallurgy Research Center,
U. S. Bureau of Mines, College Park,
Maryland
Xll
-------
9:50 - 10:15 a.m. ESTIMATION OF POLLUTANT CHARACTERISTICS
'FROM GEOCHEMICAL SURFACE INVESTIGATIONS
F. B. Tonani, Geochemex, and H. T. Meidav,
Geonomics, Incorporated, Berkeley,
California
10:15 - 10:35 a.m. COFFEE
10:35 - 11:00 a.m. SAMPLING AND ANALYSIS OF GEOTHERMAL
BRINES FROM NILAND FIELD, CALIFORNIA
H. K. Bishop, San Diego Gas & Electric
Company, San Diego, California
11:00 - 11:25 a.m. INTERPRETATION OF ANALYTICAL RESULTS
THERMAL AND NON-THERMAL WATERS, LAVA
PLATEAUS REGION OF NORTHEASTERN CALI-
FORNIA AND SOUTHERN OREGON
C. W. Klein and J. B. Koenig, GeothermEx,
Incorporated, Berkeley, California
11:25 - 11:50 a.m. GEOCHEMICAL ANALYSIS OF FLUIDS CIRCU-
LATED THROUGH A GRANITIC HOT DRY ROCK
GEOTHERMAL SYSTEM
J. W. Tester, C. Grigsby, C. Holley and
L. Blatz, Los Alamos Scientific Labora-
tory of the University of California,
Los Alamos, New Mexico
11:50 - 12:15 p.m. CHEMICAL PROFILE OF THE EAST MESA GEO-
THERMAL FIELD, IMPERIAL VALLEY, CALI-
FORNIA
R. T. Littleton and E. Burnett, U. S.
Bureau of Reclamation, Boulder City,
Nevada
12:15 - 1:45 p.m. LUNCH
Related Topics Session
Session Chairman: R. Littleton, U. S.
Bureau of Reclamation, Boulder City,
Nevada
1:45 - 2:10 p.m. STATUS REPORT, RAFT RIVER PROJECT, SAM-
PLING, ANALYSIS, AND ENVIRONMENTAL
EFFECTS STUDIES
A.C. Allen, J.M. Baldwin and R.E. McAtee,
Allied Chemical Corp.,Idaho Falls,Idaho
2:10 - 2:35 p.m. BOREHOLE GEOPHYSICAL LOGGING AS COMPLE-
MENT TO GEOTHERMAL WELL EFFLUENT SAM-
PLING
S. K. Sanyal and R. B. Weiss, Geonomics,
Incorporated, Berkeley, California
xiii
-------
2:35 - 3:00 p.m. USE OF RADIOACTIVE TRACERS IN GEOTHERMAL
OPERATIONS
0. Vetter, Vetter Associates, Laguna
Beach, California
3:00 - 3:25 p.m. THE LBL GEOTHERMAL BRINE DATA COMPILA-
TION PROJECT
S. R. Cosner and J. A. Apps, Lawrence
Berkeley Laboratory, Berkeley,
California
3:25 - 3:45 p.m. COFFEE
3:45 - 5:00 P.M. WRAP-UP PANEL DISCUSSION
Chairman: D. B. Gilmore, EMSL, U.S.
EPA, Las Vegas, Nevada
-------
LIST OF REGISTERED PARTICIPANTS
1. William Adams
U.S. EPA, EMSL
P. 0. Box 15027
Las Vegas, Nevada 89114
2, John A. Apps
Building 90-2130
Lawrence Berkeley Laboratory
Berkeley, California 94720
3. Harry G. Arnold
P. 0. Box X
Building 4500N, D-33
Oak Ridge, Tennessee 37830
4. Dr. David J. Atkinson
Hydro-Search, Incorporated
333 Flint Street
Reno, Nevada 89501
5. L. H. Axtell
Geothermal Services, Incorporated
7860 Convoy Court
San Diego, California 92111
6. Allan R. Batterman
U.S. EPA, EMSL
P. 0. Box 15027
Las Vegas, Nevada 89114
7. Harry Bishop
San Diego Gas & Electric Company
P. O. Box 1831
San Diego, California 92112
8. Robert Boshan
Hughes Aircraft Company
2012 Midvale Avenue
Los Angeles, California 90025
9. Harry R. Bowman
Building 70-189
Lawrence Berkeley Laboratory
Berkeley, California 94720
10. Erich Bretthauer
U.S. EPA, EMSL
P. O. Box 15027
Las Vegas, Nevada 89114
xv
-------
11. Charles Brook
U.S. Geological Survey
345 Middlefield Road
Menlo Park, California 94025
12. Ken W. Brown
U.S. EPA, EMSL
5127 Blanton Drive
Las Vegas, Nevada 89122
13. Larry L. Brown
University of Denver
1442 South Vivian Way
Lakewood, Colorado 80215
14. Dwight Lee Carey
Republic Geothermal, Incorporated
11823 East Slauson, Suite 1
Santa Fe Springs, California 90670
15. Harry M. Castrantas
FMC Corporation
407 South Norwood Avenue
Newton, Pennsylvania 18940
16. Leonard A. Cavanagh
Stanford Research Institute
1550 Calaveras Avenue
San Jose, California 95126
17. Raymond Cedillo
Southern California Edison Company
9329 Glendon Way
Rosemead, California 91770
18. Donald J. Christoffersen
Union Oil Company
P. 0. Box 76
Brea, California 92621
19. Gerald Conner
U.S. Bureau of Mines
College Park Metallurgy Research Center
College Park, Maryland 20740
20. Steven R. Cosner
Building 90-2024P
Lawrence Berkeley Laboratory
Berkeley, Califonia 94720
xvi
-------
21. Glen Coury
Coury & Associates
7400 West 14th Avenue, Suite 2
Lakewood, Colorado 80214
22. Alan B. Crockett
U.S. EPA, EMSL
P. 0. Box 15027
Las Vegas, Nevada 98114
23. Gustavo Cueliar
Coraisi6n Ele'ctrica Rio Lempa
Calle Los Claveles-21
Col. La Sultana - Antiguo
Cuscatlan, El Salvador
24. Gene Culver
Geo-Heat Utilization Center
Oregon Institute of Technology
Klamath Falls, Oregon 97601
25. Lester Danterive
U.S. Geological Survey
345 Middlefield Road
Menlo Park, California 94025
26. Otis L. Day
Geonomics, Incorporated
3165 Adeline Street
Berkeley, California 94703
27. Frank Dellechaie
AMAX Exploration, Incorporated
4704 Harlan Street
Denver, Colorado 80212
28. Daniel A. Demeo
Hughes Aircraft Company
Building 6-D-133
Centinela and Teale Streets
Culver City, California 90230
29. Leo Deffending
Battelie-Northwest Laboratories
P. 0. Box 999
Richland, Washington 99352
30 •. Sam Dermengian
Geothermal Energy Magazine
318 Cherrywood Street
West Covina, California 91741
xvii
-------
31. Jay D. Dick
Denver Research Institute
University of Denver
Denver, Colorado 80208
32. Jane Dionne
Federal Energy Administration
1200 Pennsylvania Avenue, Northwest
Washington, D. C. 20461
33. Gene Elliott
Federal Bureau of Mines
1605 Evans Avenue
Reno, Nevada 89512
34. Giancarlo Facca
1023 Timothy Lane
Lafayette, California 94549
35. Wayne A. Fernelius
Bureau of Reclamation
P. 0. Box 427
Boulder City, Nevada
36. Bill Fiero
University of Nevada
4505 Maryland Parkway South
Las Vegas, Nevada 89154
37. Peter Fintschenko
FMC Corporation
2000 Market Street
Philadelphia, Pennsylvania 19103
38. Carl Flegal
TRW, Incorporated
01-2030, One Space Park
Redondo Beach, California 90278
39. Jack F. Foehr
CER Corporation
P. 0. Box 15090
Las Vegas, Nevada 89114
40. Gary W. Frey
U. S. Bureau of Reclamation
310 Wyoming Street
Boulder City, Nevada 89005
41. George A. Frye
Aminoil USA, Incorporated
P. 0. Box 11279
Santa Rosa, California 95406
xviii
-------
42, John Fruchter
Battelle-Northwest Laboratories
P. 0. Box 999
Richland, Washington 99352
43. Murray C. Gardner
GeothermEx, Incorporated
901 Mendocino Avenue
Berkeley, California 94707
44. Donald B, Gilmore
U.S. EPA, EMSL
P. 0. Box 15027
Las Vegas, Nevada 89114
45. Louis H. Goldsmith
Geonomics, Incorporated
Apartado 2475
Managua, Nicaragua
46. Paul Gudiksen
Lawrence Livermore Laboratory
P. 0. Box 808
Livermore, California 94550
47. M. S. Gulati
Union Oil Company
P. 0. Box 6854
Santa Rosa, California 95405
48. W, Roger Hail
W. A. Wahler & Associates
1023 Corporation Way
Palo Alto, California 94303
49. Beverly A. Hall
Geothermal Resources Council
P. 0. Box 1033
Davis, California 95616
50. J. Warren Hamersma
TRW Defense & Space Systems
01-2020 , One Space Park
Redondo Beach , California 90278
51. Robert P. Hartley
U.S. EPA, IERL
5555 Ridge Avenue
Cincinnati, Ohio 45268
xix
-------
52. Jonathan Herrmann
U.S. EPA, Region VIII
1860 Lincoln Street, Suite 900
Denver, Colorado 80011
53. John H. Hill
Lawrence Livermore Laboratory
P. 0. Box 808, L-404
Livermore, California 94550
54. Anders F. Jepsen
Eureka Resource Associates, Incorporated
2161 Shattuck Avenue, Suite 317
Berkeley, California 94704
55. Richard L. Jodry
Suntech, Incorporated
P. 0. Box 936
Richardson, Texas 75080
56. F. B. Johns
U.S. EPA, EMSL
P. 0. Box 15027
Las Vegas, Nevada 89114
57. Stuart D. Johnson
Phillips Petroleum Company
P. 0. Box 752
Del Mar, California 92014
58. Robert Kent
U.S. Geological Survey
345 Middlefield Road
Menlo Park, California 94025
59. Christopher W. Klein
GeothermEx, Incorporated
901 Mendocino Avenue
Berkeley, California 94707
60. William T. Kreiss
Physical Dynamics, Incorporated
5737 37th Avenue Northeast
Seattle, Washington 98105
61. Paul Kruger
Civil Engineering Department
Stanford University
Stanford, California 94305
xx
-------
62. Robert T. Littleton
U.S. Bureau of Reclamation
624 Don Vincente Drive
Boulder City, Nevada 89005
63. Carlos V. Lopez
Empresa Nacional de Luz & Fuerza
Altamira d1 Este 317
Managua, Nicaragua
64. J. D. Ludwick
Battelle-Northwest Laboratories
1616 Woodbury
Richland, Washington 99352
65. Bruce Mann
U.S. EPA, ORF
P. 0. Box 15027
Las Vegas, Nevada 89114
66. Skip Matlick
Republic Geothermal, Inc.
P. 0. Box 3388
Santa Fe Springs, California 90670
67. R. E. McAtee
Allied Chemical Corporation
550 2nd Street
Idaho Falls, Idaho 83401
68. Richard McCurdy
Pacific Gas & Electric
3400 Crow Canyon Road
San Ramon, California 94583
69. Katherine F. Meadows
Geothermal World Directory
P. 0. Box 997
Glendora, California 91740
70. H. Tsvi Meidav
Geonomics, Incorporated
3165 Adeline Street
Berkeley, California 94703
71. Mae Z. Meidav
Geonomics, Incorporated
3165 Adeline Street
Berkeley, California 94703
-------
72. Donald E. Michels
EG&G Idaho, Incorporated
307 2nd Street
Idaho Falls, Idaho 83401
73. Leland L. Mink
U.S. EPA, EMSL
P. O. Box 15027
Las Vegas, Nevada 89114
74. Jorge R. Molina
Comisi6n ElSctrica Rio Lempa
31 Av. Sur #209 Col Cucumacryan
San Salvador, El Salvador
75. Robert E. Moran
U.S. Geological Survey
M. S. 415, Federal Center
Denver, Colorado 80215
76. George B. Morgan
U.S. EPA, EMSL
P. 0. Box 15027
Las Vegas, Nevada 89114
77. Gi6 M. Morse
Geonomics, Incorporated
3165 Adeline Street
Berkeley, California 94703
78. John G. Moylan
Lockheed Electric Company
4220 South Maryland Parkway
Las Vegas, Nevada 89109
79. David A. Myers
Battelle-Northwest Laboratories
2533 Davison
Richland, Washington 99352
80. Nancy L. Nehring
U.S. Geological Survey
345 Middlefield Road
Menlo Park, California 94025
81. Peter 0. Nelson
Department of Civil Engineering
Oregon State University
Corvallis, Oregon 97331
xx 1.1
-------
82. Zachary Nelson
U.S. EPA, EMSL
P. 0. Box 15027
Las Vegas, Nevada 89114
83. Charles Otto
Lawrence Livermore Laboratory
Box 808, L-404
Livermore, California 94550
84. Alan 0. Ramo
Sunoco Energy Development Company
12700 Park Central Place, Suite 1500
Dallas, Texas 75251
85. Kenneth Rea
Los Alamos Scientific Laboratory
P. 0. Box 1663
Los Alamos, New Mexico 87545
86. Marshall Reed
U.S. Geological Survey, MS 92
345 Middlefield Road
Menlo Park, California 94025
87. Jerry A. Richl
Northwest Environmental Technical Laboratories,
Incorporated
5709 80th Avenue Southeast
Merrer Island, Washington 98040
88. David E. Robertson
Battelle-Northwest Laboratories
P. o. Box 999
Richland, Washington 99352
89. Stephen Ronshaugen
U.S. EPA, EMSL
P. 0. Box 15027
Las Vegas, Nevada 89114
90. L. W. Ross
Denver Research Institute
University of Denver
Denver, Colorado 80208
91. Subir K. Sanyal
Geonomics, Incorporated
3165 Adeline Street
Berkeley, California 94703
xxiii
-------
92. David Sarkaria
Aminoil USA, Incorporated
P. 0. Box 191
Huntington Beach, California
93. Lew Schalit
Acurex-Aerotherm
485 Clyde Avenue
Mountain View, California 94042
94. Arthur J. Soinski
LFE Corporation
2030 Wright Avenue
Richmond, California 94804
95. Richard E. Stanley
U.S. EPA, EMSL
P. O. Box 15027
Las Vegas, Nevada 89114
96. William P. Staub
Oak Ridge National Laboratory
P. 0. Box X, 4500N, D-33
Oak Ridge, Tennessee 37830
97. Beverly Strisower
Lawrence Berkeley Laboratory
Building 90-3081
Berkeley, California 94720
98. Wayne J. Subcasky
Chevron Oil Field Research Company
P. 0. Box 446
La Habra, California 90631
99. Glenn Suter
Oak Ridge National Laboratory
P. O. Box X
Oak Ridge, Tennessee 37830
100. William Sutton
U.S. EPA, EMSL
P. 0. Box 15027
Las Vegas, Nevada 89114
101. Chandler A. Swanberg
Department of Physics, Box 30
New Mexico State University
Las Cruces, New Mexico 88003
xxiv
-------
102. Jefferson W. Tester
Los Alamos Scientific Laboratory
Mail Stop 981
Los Alamos, New Mexico 87545
103. J. M. Thompson
U.S. Geological Survey
345 Middlefield Road
Menlo Park, California 94025
104. Franco B. Tonani
Geochemex, Incorporated
3165 Adeline Street
Berkeley, California 94703
105. Albert C. Trakowski
U.S. EPA, ORD
401 M. Street SW
Washington, DC 20460
106. Felix Tsai
Geochemex, Incorporated
3165 Adeline Street
Berkeley, California 94703
107. Otto Vetter
Vetter Associates
580 Vista Lane
Laguna Beach, California 92651
108, Jim Watson
Battelle-Northwest Laboratories
P. 0. Box 999, Area 300, Building 314
Richland, Washington 99352
109. Cooper H. Wayman
U.S. EPA
1860 Lincoln, Suite 900
Denver, Colorado 80203
110. Richard B. Weiss
Geonomics, Incorporated
3165 Adeline Street
Berkeley, California 94703
111. Herbert C. Wells
Engineering Department
University of Nevada
4505 Maryland Parkway
Las Vegas, Nevada 89154
xxv
-------
112. Robert N. Wheatley
Union Oil Company Research Center
1409 Skyline Drive
Fullerton, California 92631
113. D. W. Wheeler
Aminoil USA, Incorporated
P. 0. Box 191
Huntington Beach, California 92648
114. Tricia White
Las Vegas Sun
P. 0. Box 4275
Las Vegas, Nevada 89106
115. Barry Williams
7860 Convoy Court
San Diego, California 92111
116. Rod D. Wimer
Portland General Electric
121 Southwest Salmon
Portland, Oregon 97204
117. E. M. Woodruff
Battelle-Northwest Laboratories
P. 0. Box 999
Richland, Washington 99352
118. Richard V. Wyman
Engineering Department
University of Nevada
4505 Maryland Parkway
Las Vegas, Nevada 89154
xx vx
-------
LIST OF ABBREVIATIONS AND SYMBOLS
yg/h or yg/hr
Ug/1
Ug/m or ug/min
ymho/cm
micrograms per hour
micrograms per liter
micrograms per minute
micromhos per centimeter
AA or AAS = atomic absorption
spectrometer
atm = atmospheres
cc = cubic centimeter(s)
cm = centimeter(s)
CP-grade = chemically pure
DT = detector tube
ERDA = U.S. Energy Research and
Development Administration
EPA = U.S. Environmental
Protection Agency
e p m = equivalents per million
g or gm
g/m
g/m3
GC
h or hr(s)
id, ID or i.d.
in
kg
KGRA
km
Ib
liq
1pm, 1/m or
1/min
mM/lt
mi
gram(s)
grams per meter
grams per cubic meter
gas chromatograph
hour(s)
inside diameter
inch(es)
kilogram(s)
Known Geothermal Resource
Area
kilometer(s)
pound(s)
liquid
liters per minute
millimoles per liter
mile(s)
liiin = minute (s)
ml = milliliter
MV = megavolt
mV = millivolt
MW = megawatt
nm = nanometer(s)
od, OD or = outside diameter
o.d.
ORP = oxidation-reduction
potential
P log = pressure log
ppb = parts per billion
ppm = parts per million
ppt = parts per thousand
psi = pounds per square
inch
SP log = self potential log
SpC = specific conductance
T log = temperature log
TDS = total dissolved
solids
vap = vapor
V/V = volume per volume
wt = weight
The following items appear-
ing in this document are
Registered Trademark items
(See DISCLAIMER, page ii):
Pyrex
Tygon
Teflon
STRactan-10
x-xvii
-------
METRIC CONVERSION TABLE
Non-metric unit Multiply by Metric unit
Fahrenheit 5/9(°F - 32) Celsius
foot(feet) 0.3048 meter(s)
gallon(s) 3.7854 liter(s)
inch(es) 25.4 nrinimeter(s)
mile(s) 1.609 kilometer(s)
ounce 28.35 gram(s)
pound (s) 0.454 kilogram(s)
psi (pounds per 6894.7 pascal(s)
square inch) 0.06805 atmosphere(s)
ton(s) 0.9072 metric ton(s)
British units were used interchangeably in this document because
much of the source data were not available in metric units.
xxvi11
-------
ACKNOWLEDGEMENT
Geonomics, Inc., would like to express its appreciation to those
who presented papers at the workshop and especially to those who
also submitted papers for inclusion in this volume. The work-
shop would not have been the success it was without those who
participated in the panel discussions, those who helped us
contact prospective speakers and participants, those who made
valuable suggestions on the format and content of the workshop
and, finally, those who attended the workshop. We would also
like to thank Mr. Donald B. Gilmore, U.S. Environmental
Protection Agency, for his personal participation and support
in the organization of this workshop.
XXIX
-------
ABSTRACT
THE USE OF GAS SAMPLING BAGS FOR THE COLLECTION AND
STORAGE OF HYDROTHERMAL GASES
A. J. Soinski
LFE Corporation
Environmental Analysis Laboratories
Richmond, California
The analysis of geothermal gas samples by means of gas chroma-
tography or potentiometry is now routine and reliable. However,
difficulties in the collection and storage of hydrothermal gases
remain. The collected gas must be representative of the effluent,
and sample modification by processes such as the reaction of
component gases, contamination with ambient air, diffusional loss,
or wall loss should not occur. In addition, it is desirable that
the collection system be rugged, inexpensive, and easy to use.
Gas sampling and storage bags, made from Tedlar, Teflon, or poly-
ethylene film, have these latter convenience characteristics, but
they may fail to meet the fundamental requirements for sample
preservation.
We have performed some preliminary laboratory measurements of the
stability of reactive and residual gases during storage in plastic
bags. Sulfur-containing species are unstable, and they should be
stabilized in aqueous solutions or on a solid matrix and/or ana-
lyzed on site. Diffusional losses of hydrogen are appreciable
also. We conclude that the utility of plastic bags for the col-
lection and storage of geothermal gases is limited to chemically
stable and higher molecular weight species.
Solid adsorbents, such as silica gel and activated charcoal, are
gaining favor over sampling bags for the collection of odorous
gases in the workplace environment. The general requirements of
component stability and freedom from contamination are fulfilled.
We suggest that silica gel-filled adsorption tubes, similar to the
one used for volcanic gases by Naughton, are preferable to plastic
bags for geothermal gas collection and storage.
-------
ABSTRACT
HEAVY METAL EMISSIONS FROM GEOTHERMAL POWER PLANTS
D. E. Robertson, E. A. Crecelius, J. S. Fruchter, J. D. Ludwick
Battelle-Northwest Laboratories, Richland, Washington
The high temperature magmatic processes which create the hydro-
thermal provinces of high geothermal energy potential also result
in the accumulation of undesirable gases and liquids. During the
tapping and utilization of geothermal fluids for power production,
elevated levels of these naturally produced contaminants can be
released to the surrounding environment. We have initiated a study
to identify the quantities and chemical forms of volatile and water
soluble heavy metals released in geothermal effluents. To date,
our studies have been conducted at The Geysers, Cerro Prieto and
Raft River geothermal sites.
Of the heavy metals studied, mercury is the most volatile, and gms/
hour quantities are released from the generating units at The Gey-
sers and Cerro Prieto. Mercury concentrations in Raft River steam
and water are one to two orders of magnitude lower than at The Gey-
sers and Cerro Prieto. The volatile mercury at all locations is
predominantly elemental Hg° vapor, which was surprising because of
the high concentrations of HjS present.
Arsenic primarily follows the liquid phase at the hot water domi-
nated sites and concentrations in the geothermal waters ranged
from 28 yg/1 in Raft River hot water to 2250 ug/1 in the Cerro
Prieto brine. At Cerro Prieto 2480 gms/hr of arsenic is releas-
ed to an evaporation pond. At Cerro Prieto the arsenic in the
fresh brine is primarily in the +3 oxidation state, but appears
to oxidize rapidly to the +5 form when exposed to the atmosphere.
The arsenic in fresh hot water from Raft River is initially pre-
sent mainly in the +5 form.
-------
THE DYNAMIC MEASUREMENT OF AMBIENT AIRBORNE GASES
NEAR GEOTHERMAL AREAS
Anders F. Jepsen
Eureka Resource Associates, Inc.
Berkeley, California
and
Lee Langan
Environmental Measurements, Inc.
San Francisco, California
INTRODUCTION
In geothermal areas, as elsewhere, there are both man-made and
natural sources of atmospheric contamination which are of interest
from the point of view of either environmental impact or exploration.
The most obvious are the clouds of smoke and steam from active vol-
canoes and the steam exhausts from operating geothermal power units,
which can be seen for many miles. Less apparent but still of great
human concern are the sulfide odors and mists from fumaroles and hot
springs. Least obvious, in fact usually undetectable to the unequipped
human (except for hydrogen sulfide) are the clouds of vapor emitted
naturally in small but clearly measurable amounts by certain geological
conditions as well as by development processes.
The assessment of the environmental impact caused by development
of geothermal resources requires a quantitative distinction between
the pre- and the post-development condition. In the past such a dis-
tinction has been difficult because of the lack of hard data, al-
though that did not hinder some workers from assuming a zero level of
contamination prior to development. Now, background data is a
fundamental requirement in development programs in geothermal areas
as well as elsewhere. Such data may be gathered by conventional means
through pre-development stationary monitoring. They also may be
obtained by such innovative methods as the dynamic measurement tech-
nique, in which clouds of air contaminant are mapped to define their
measurable limits in space, that is, to background.
This paper will briefly describe the dynamic measurement tech-
nique and present some data obtained in this way from geothermal areas.
Applications include not only environmental assessment but also geo-
logical reconnaissance.
-------
THE DYNAMIC MEASUREMENT TECHNIQUE
The technique of dynamic measurement of air pollution has been
applied on and off for many years, particularly to the study of
specific source-oriented problems. Airborne measurements of plume gas
concentrations have been made to try to verify emission rates and
predictive model reliability. Ground-based dynamic measurements have
been employed to trace suspected clouds of gas emitted by stationary
sources, in response to complaints or to special scientific interest.
All these efforts have been limited by the difficulty inherent in the
slow-responding, delicate, chemical analyzers available for gas analy-
sis. However, now that rugged, fast response air monitors which operate
mostly on physical rather than chemical principles are available,the
dynamic measurement technique is more widely used.
Dynamic air quality measurements are those which are made while
the monitor is in motion, in contrast to static measurements, those
traditionally made by a stationary monitor operating continuously at
some location. The difference between static and dynamic measurement
is that the stationary monitor waits for the air pollution to come to
it while the moving monitor moves to the contaminating plume and
measures it in situ. Because the dynamic measurements are made while
the monitor is in motion, it is possible to map the distribution of the
air pollution in both space and time. The data are both spatial and
temporal in character. In contrast, the static measurement data are
temporal only and contain no information as to location relative to the
maximum concentration in the plume.
Meteorologists have been making dynamic measurements of the atmos-
phere for many years. The most familiar technique has been to measure
the vertical wind profile, that is, the horizontal speed of the wind
measured from the surface up to some elevation above the earth. This
is accomplished by releasing a pilot balloon and tracking its movement
through the atmosphere by means of a theodolite. Ultimately, from
this set of measurements, the position of the balloon in space may be
calculated as a function of time, and hence the horizontal wind speed
determined as a function of altitude. Similar dynamic techniques can
be used to measure pollutant profiles through atmospheric plumes to
learn more about how they are dispersed and what levels of concentration
actually prevail in them.
MOVING POINT ANALYZERS
To provide meaningful data in the moving mode, point analyzers
must be able to meet strict specifications. They must be able to
operate under varying conditions of temperature and humidity while
experiencing the acceleration and shock common to vehicular operation.
-------
Power demands should be low so several instruments can be used in the
same system. The time response should be fast, on the order of a few
seconds if possible, to maximize spatial resolution.
The time response is particularly important. The dynamic measure-
ment procedure is specifically aimed at defining plume geometry and
hence requires a measurement system which responds fast enough to sense
and record gas levels at the plume edges, as well as at the point of
maximum concentration.
Recent advances in air quality monitors make possible the dynamic
measurement of several pollutants. The flame photometric detector and
and pulsed fluorescent monitor both permit the accurate measurement of
total sulfur of SO^. The chemi1uminescent method permits the accurate
measurement of NO, NOX, and 0^. A later section will discuss measure-
ments made in a geothermal area using a Hg-sensitive spectrometer.
REMOTE SENSORS
Remote sensors provide a separate, unique approach in dynamic
measurements. They are able to detect overhead clouds of gas without
having to be in them,and their mode of operation permits moving
measurements while they are themselves stationary. Passive remote
sensors, such as the correlation spectrometer (COSPEC) , rely on natural
light from the sky as the source of energy and monitor the energy
absorption by molecules of specific gases along the line of sight.
Active remote sensors include the COSPEC, when aimed at an artificial
light, and the laser radar (LIDAR) which measures reflections of its own
signals from aerosols and particles. These are operational systems.
Still others (e.g./gas filter correlation instruments) are in the devel-
opmental stage.
Remote sensors can be used from a fixed point to monitor stack
emissions at the exit or to map the pollutant dispersal pattern down-
wind. If viewing the stack exit, a tripod-mounted remote sensor is
panned horizontally across the stem of the plume. This can yield the
true gas concentrations, provided the stack diameter and gas exit
temperatures are known. Further downwind the stationary remote sensor
can measure in a vertical plane through the near-horizontal plume. This
can yield the effective stack height and vertical mixing coefficient
of the plume for modeling purposes. From one site a number of separate
profiles may be obtained for increasing distances downwind simply by
changing the horizontal angle between vertical profiles. If the wind
speed is known, the net flow of gas across the plane may also be cal-
culated.
Finally, remote sensors can be installed in moving laboratories,
along with ground level monitors, to view upward while traversing.
-------
This is the most efficient use of these instruments. This technique
permits the integrated measurement of burden of gas overhead and hence
the mapping in horizontal coordinates of the vertically integrated
total burden of contaminant in the atmosphere. A remote sensor, when
operated in a moving laboratory together with ground level monitors, is
a powerful device for quantifying the contribution of individual sources
to the total air contaminant impact of an airshed.
EXPERIMENTAL DESIGN
Dynamic measurements are especially suited to the study of com-
plicated air monitoring problems, including multiple source pollutants
and plume dispersal in rough terrain.
Consider the source and plume illustrated in Figure 1. The con-
taminant is emitted from the source into the atmosphere and dispersed.
The dispersal is controlled by the wind speed and atmospheric stability.
The ground level concentrations downwind depend on the amount of
contaminant emitted, the atmospheric dispersal parameters and position.
However, the contaminant may be emitted into an atmosphere already
bearing some other source's effluent. The air flow may be distorted by
hills near the plant.
The moving laboratory with its remote sensor and point monitors is
able to separate the contaminant plume from other such plumes which may
be in the area. By traversing upwind as well as downwind, it estab-
lishes the background into which the gas is dispersed. It can then map
the downwind dispersal patterns to locate the point of maximum ground
level concentration and to measure the horizontal crosswind dimension
of the plume.
A remote sensor can be set up at the site to make lateral obser-
vations of the plume from the site. It first is scanned upwind to
define background conditions. Then it is used to measure the vertical
profile of the plume to determine its vertical dimension and the effec-
tive stack height.
Simultaneously, dynamic meteorological measurements can be made to
determine the character of the dispersing atmosphere. Double theodolite
tracking of pilot balloons permits the accurate measurement of the
vertical wind profile and hence the wind speed at the elevation of the
plume. The vertical temperature profile is obtained either from a sen-
sor attached to the balloon or with a light aircraft. Data from a met
tower provides continuous surface conditions with which to correlate
the upper atmospheric data.
-------
WIND, STABILITY
CONTINUOUS
WIND
DYNAMIC
METEOROLOGY
CONCEN-
TRATION
REMOTE
SENSOR
GROUND LEVEL
THRESHOLD OF
DETECTABILITY
Figure 1. Complete plume characterization by dynamic
measurements
WIND
S02 PROFILE
Figure 2. Sketch showing configuration of volcanic emissions
and remote sensor. Rate of S02 emission was
calculated by multiplying the wind velocity by
the measured S02 load in the plume
-------
Such a dynamic measurement program produces a broad, detailed array
of data. It is designed to provide data for both model tuning and the
quantitative separation of the contributions of different sources. It
can also be used for reconnaissance mapping of gas clouds for both envi-
ronmental and exploration purposes.
APPLICATION IN GEOTHERMAL AREAS
The limited access of a typical geothermal area in the early stages
of development places a significant constraint on the design of a
measurement experiment. If the observations are to be made from an
aircraft, the problems are minimal. However, ground based measurements
with point monitors are limited to the available roads. Remote sensor
locations are similarly limited to sites within reach of the field team.
The ability of the field team becomes an important factor in measurement
planning.
The specific targets of interest must be defined beforehand.
Whether the source is a venting well, a natural geyser, or an operating
power unit, its location relative to roads, topography and the dispers-
ing wind, and the possibility of cyclical emissions will, of course,
control what is measured and where.
In the early stages of a field experiment dynamic measurements can
also serve as a reconnaissance tool to identify possible sources of
airborne molecules whose presence was previously unsuspected. Those
targets are usually natural sources such as fault zones or old mine
workings.
Example: SO? Emissions from Central American Volcanoes
Near active volcanoes, access is usually so limited that a remote
sensor is the only practical way to obtain data. A study by Stoiber
and Jepsen (1973) measured the rate of emission of S02 from several
active volcanoes in Central America, using a correlation spectrometer
and dispersing wind speed observations (see Figure 2). This study con-
cluded that the total emissions from Central American volcanoes were at
the rate of 1300 tons per day. Table 1 identifies the measured emission
rate of each of the volcanoes studied. No other technology was capable
of making these observations.
-------
TABLE 1
AMOUNTS OF S02 GAS ISSUING FROM VOLCANOES
IN GUATEMALA, EL SALVADOR, AND NICARAGUA.
S02
Volcano (metric tons per day)
Guatemala
Santiaguito 420
Fuego 40*
Pacaya 260
Nicaragua
San Cristobal 360
Telica 20*
Momotombo 50*
Masaya 180
TOTAL 1330t
'•Data less reliable. tVery small plumes issued from
fumaroles in the craters of the volcanoes Santa Ana
and Izalco, El Salvador. Very approximate estimates
from field data suggest that these plumes account for
no more than a few tons of S02 per day.
(Reproduced from Stoiber and Jepsen, 1973)
Example: Mercury at The Geysers, California
In a program carried out in 1971 to map metallic mercury vapor
levels near areas of natural and man-made mercury concentration
(Jepsen, 1973), dynamic ground level observations were made at a number
of locations, including The Geysers area of Northern California.
The measurement system included a fast response metallic mercury
vapor detector and strip chart recorder mounted in a generator-equipped
vehicle. The sensor was an atomic absorption spectrophotometer sensi-
tive to 5 x 10~^gm/m3 with a time response of a few seconds. The
instrument was a continuous monitor originally developed by Barringer
Research Ltd. for prospecting and was rugged enough to withstand the
rigors of our field work. Calibration was carried out periodically in
the field by injecting an aliquot of mercury into the sample line. The
system was also zeroed periodically in the field by passing the sample
air flow through a mercury scrubber (>36% efficiency) upstream from the
analysis chamber. Figure 3 illustrates the complete measurement system.
-------
-AIR BIVALVE ASSEMBLY
IN
L_
TAKtl UCDTIID
k MtKLUK
V A
DbUKBtK M
( FREE FLOW (JJ.
I
DIGITAL
VOLTMETER
LAMP [ '
ce &
f i ter
moa
oven
PREAMP
Dhoto-
tube
.2|
OPTICAL ASSEMBLY
FOLDED LIGHT PATH
cables |
CVLJ/M ICT
STRIP CHA
&
RANGE
SELECTIOf
U
\
ELECTRONICS RACK
i .
cell filter
amplifier amp fier 2 5~ c -a
1 . ft- -s -os-
DIFFERENTIAL oa^ -•->
AMPL FIER ">
Figure 3. Simplified block diagram of mercury vapor
measurement system.
MERCURYVILLE
51 X 10 9 gm/m3
MIN
Figure 4.
Portion of strip chart record showing mercury
background and low peak.
10
-------
The data obtained on February 15, 1971 show the character of data
possible from such a survey. The measurement system was operated
throughout a traverse from Geyserville northward into The Geysers Resort.
Figure k shows the nature of the signal as recorded at a high sensitivity
setting on the strip chart. Key locations were noted on the strip chart
to facilitate later correlation of the data with position. Whenever
signals greater than the dynamic range of the recorder were anticipated,
the recorder scale was reduced to permit accurate measurement of the
peak values. (Today on-board microcomputers and inertia! guidance
systems do all the work of spatial correlation and scale adjustment,
and even draw maps and profiles in real time as the traverse is carried
out.)
The presence of apparently naturally occurring mercury vapor in the
atmosphere was detected at two separate areas along the traverse (Fig-
ure 5). The first was about 800 meters northeast of Mercuryville where
a peak of 51 x 10 gm/m3 was observed. The second area contained four
peaks, located along the road beside Big Sulfur Creek from Adit #2
of Buckman Mines (300 x 10~9gm/m3) eastward to the power plant at The
Geysers (900 x 10~9gm/m3). Along the way a peak of about 800 x 10~9
gm/m3 was observed near The Geysers Resort.
Past the power plant the traverse led up the side of the hill be-
tween venting wells. Levels there ranged from 3500 x 10~9 gm/m3 along
the road to a single observation of 28,000 x 10~9gm/m3 taken while
directly sampling a venting well.
The meteorology was probably very important' in controlling the
levels observed that day. Winds were light from the west, tending to
move the gaseous material gently eastward from its source. As a re-
sult there may have been a general funnel ing and recirculation of the
air in the valley, explaining the high ambient levels measured at The
Geysers and the Resort.
The data near Mercuryville, however, is less easy to explain, since
the mine development is some distance from the road. There is some
topography which might interfere with atmospheric dispersion. However,
another explanation is possible.
A geological map of the area shows a zone of alteration, reflect-
ing a permeable fracture zone, crossing right at the location of the
peak (Erskine, 1977). This coincidence of fracture zone and airborne
mercury vapor suggests that the mercury may have been naturally emitted
from the fracture. If this were the case, then in other areas the
presence of mercury vapor may also indicate the location of permeable
fracture zones. In particular, the presence of that zone in the Big
Sulfur Creek area suggests that the mercury levels there may be due
11
-------
Fracture Zones
Figure 5. Profile showing mercury concentrations measured
along traverse route and mapped zones of permeable
fracture
12
-------
not only to human activity in both geothermal development and mercury
mining, but also in significant part to emissions from the local
geology. Pre-development measurements may have supported this idea.
If the presence of mercury vapor is a reliable guide to zones of
fracture and permeability in the local geology, then the dynamic mea-
surement of that gas in the atmosphere could also serve as a tool to
map heat conducting zones. In that case, the usefulness of the dynamic
measurement of mercury vapor in geothermal areas would have come full
circle, back to its original intent of geological exploration.
REFERENCES
Stoiber, Richard E., and Anders F. Jepsen, Sulfur Dioxide Contribu-
tions to the Atmosphere by Volcanoes, Science, v. 182, pp. 577~
578, 9 November 1973-
Jepsen, Anders F. , Measurements of Mercury Vapor in the Atmosphere,
IN: Trace Elements in the Environment, E. L. Kothny, Editor.
Advances in Chemistry Series 123, American Chemical Society,
Washington, D. C. 1973-
Erskine, Melville C., Jr., 1977, personal communication.
13
-------
ABSTRACT
ANALYSIS OF RADON IN GEOTHERMAL EFFLUENTS
P. Kruger
Stanford University, Stanford, California
Sampling and analysis of geothermal effluents is an important
problem in the effort to ensure the development of geothermal
energy in an environmentally responsible manner. Yet the need
for good sampling and analytical methods is common to many
operational aspects of geothermal energy across the full cycle
of exploration, assessment, production, and utilization. An
example is the study of radon emanation in geothermal tiuids.
Radon sampling, detection and measurement have had extensive
study because radon is a radioactive gas. Yet as a chemical
constituent of geothermal fluids it can act as an internal
tracer for studies of reservoir physics and engineering. The
concentration of radon in produced geofluids depends on several
independent factors, among them the distribution of radium in
the formation, the emanating power of the produced radon, and
the transport time of the radon from emanation to sampling sites.
The distribution of radium in a geothermal system is a function
of the hydro chemical and thermochemical history of the forma-
tion and its infiltrating geofluids. The emanating power is a
function of the physical state of the formation. And the trans-
port time of the radon is a function of the hydrodynamic pro-
perties of the reservoir. Thus, studies of radon concentration
in produced geofluids, as a function of time in a single well,
in wells distributed over a geothermal area, and in wells in
different geothermal areas, each with flow conditions as a major
parameter, may afford a detailed understanding not only of the
role of radon as a component of geothermal fluids, but also
of reservoir transport phenomena both under steady-flow and
transient-flow conditions.
14
-------
ABSTRACT
NOBLE GAS SAMPLING OF THE GEOTHERMAL
HEAT LOOP AT NILAND FIELD, CALIFORNIA
M. O'Connell
U.S. Environmental Protection Agency
Office of Radiation Programs, Las Vegas, Nevada
This paper will describe the sampling method and present the an-
alytical results of a recently completed study at the San Diego
Gas and Electric Geothermal Facility in the Imperial Valley of
California. A program to sample Radon-222, Argon, and Helium at
the 3 stages of steam-water separation, the geothermal brine sup-
ply well and at the non-condensible exhaust stack is described in
detail. There was a hope to describe the technology on a mass
balance scale but because of uncertain flow conditions, the re-
sults were expressed in a correlated manner.
This study was performed by personnel from the U.S. EPA Office
of Radiation Programs, Lawrence Livermore Laboratory, and the
U.S. Geological Survey.
15
-------
A REVIEW OF THE DETERMINATION OF HYDROGEN SULFIDE IN AIR BY THE
CADMIUM HYDROXIDE-STRACTAN COLORIMETRIC METHOD:
Current Practices and Modifications
R. A. McCurdy and S. L. Altshuler
Pacific Gas and Electric Company
Department of Engineering Research
San Ramon, California
SUMMARY
An in-depth analysis has been performed on several calibration procedures
for the cadmium hydroxide-STRactan method for one-hour sampling of ambient
hydrogen sulfide (H2S) concentrations. This is the method recommended by the
California Air Resources Board. Results show the calibration procedures to
be susceptible to varying sulfide losses during the development of the color
depending on the glassware and volumes of reagents used. Agreement can be
achieved between the use of aqueous sulfide solutions standardized by an iodo-
metric titration and the use of gravimetrically calibrated H2S permeation
tubes when similar glassware and reagent volumes are used.
In addition, the use of the cadmium hydroxide-STRactan method for 24-hour
sampling was evaluated; a 46 percent loss was observed.
INTRODUCTION
The cadmium hydroxide-STRactan method is the referee method for the deter-
mination of hourly average ambient HoS concentrations in the State of Cali-
fornia. Recent experience has shown differences of 20 to 30 percent between
various calibration procedures for this method. The more recent, popular use
of calibrated permeation tubes, whereby the concentration of a calibration
gas is determined gravimetrically, has disagreed with the results when the
method is calibrated using aqueous sulfide solutions standardized by iodo-
metric titrations. As a result of these discrepancies, considerable effort has
been devoted by Department of Engineering Research personnel to investigate the
procedures for the preparation of calibration curves in the laboratory and for
the samples in the field. The results of these investigations are reported
herein.
ANALYTICAL PROCEDURE USING THE CADMIUM HYDROXIDE-STRACTAN METHOD FOR THE
MEASUREMENT OF H2S IN AMBIENT AIR, 100 ppb
The cadmium hydroxide-STRactan method for the determination of H2S in am-
ibent_air has been described by several authors. 1,2,3 The chemical reagents
used in the method are all analytical grade, ACS. The accepted procedure is
described in Appendix A. Work described herein involved some modifications of
this procedure. In working with aqueous calibration procedures, extreme care
must be exercised to insure the use of oxygen-free water or some of the dis-
solved sulfide can be lost by oxidation. This water is prepared by boiling
deionized-distilled water to expel dissolved oxygen, and then cooling while
bubbling ultrapure nitrogen through the water to prevent oxygen from becoming
redissolved.
16
-------
LABORATORY CALIBRATION OF THE CADMIUM HYDROXIDE-STRACTAN METHOD
Three different procedures to prepare a calibration curve for the
cadmium hydroxide-STRactan method have been compared: (1) sodium sulfide,
Na2S-9H20, has been used to prepare an aqueous sulfide solution which is
then standardized by iodometric titration, (2) pure CP grade H?S gas has
been used to generate a standard aqueous sulfide solution, and (3) a
gravimetrically calibrated H2S permeation tube has been used to generate
an H2S stream of known concentration. In addition, different types of
volumetric glassware have been investigated to determine if they could
affect the calibration curve.
1. Use of Sodium Sulfide in the Preparation of a Standard Sulfide Solution
The sodium sulfide stock solution was prepared by first washing a
crystal of Na2S-9H20, weighing about 1.5 grams, in deionized water while
being contained in a beaker. In this manner, most thiosul fates and other
polysulfides which may interfere positively in the later iodometric stan-
dardization of the sulfide stock solution were eliminated. The crystal
was washed down to approximately 0.2 gm in weight and quickly placed in
a nitrogen purged 1000-ml volumetric flask. The flask was quickly filled
to volume with oxygen-free water and mixed thoroughly. This stock stan-
dard contained the equivalent of approximately 28.3 mg
Standardization was then performed on the stock solution. 25 ml of
standard 0.01N iodine were pipetted into a 250-ml erlenmeyer flask con-
taining 50 ml of deionized water. 10 ml of 50% V/V HC1 were then added
to the flask. 50 ml of the stock sulfide solution were accurately pi-
petted into the flask, and the mixture was immediately back-titrated
with 0.01N sodium thiosulfate to the starch endpoint. The starch indi-
cator was not added to the flask until the iodine had turned to a pale
yellow color. The entire titration was done in less than two minutes.
A blank determination using 25 ml of iodine was run prior to the back-
titration. Duplicate runs were made which were within 5 ppt* in precis-
ion. Immediately after the titration was completed, 50 ml of the stock
sulfide solution were pipetted into a clean nitrogen purged 100-ml vol-
umetric flask. The flask was filled to volume with oxygen-free water
and mixed thoroughly. This working solution contained approximately
14.2 ^ H?S equivalent to 10.2 ££• H2S at 25°C, 1 atm. The individual
standards for establishing the calibration curves were prepared using
100-ml volumetric flasks. 0, 1, 2, 3, 4, 5, 6, 7 ml of working solution
were quickly pipetted into each of seven volumetrics containing about
25 ml of the .cadmium hydroxide absorbing solution. An Epindorff pipette
*parts per thousand
17
-------
was used to expedite pipetting. The standards were then filled to the
100-ml mark with additional absorbing solution, and mixed thoroughly.
This corresponded to standards with 0, 1, 2, 3, 4, 5, 6, 7 yl H2S for
each 10 ml of absorbing solution. 3.0 ml of amine solution were then
added to each flask without mixing, followed immediately (within 10
seconds) by 2.0 ml ferric chloride solution. The seven standards were
then stoppered and mixed thoroughly to develop the methylene blue color.
Each standard was developed for color one at a time. Twenty minutes
were allowed for complete color development. Each standard was then
read on a spectrophotometer in the visible range at 670 nm using the
blank standard to zero the instrument, and a calibration curve was drawn
(Curve 1, Figure 1). Curve 1 represents a statistical average using a
least squares analysis of nine calibrations. This curve is described best
by the equation:
Jfjpjjfl = 4.29 (ABS)2 + 5.6 (ABS) + 0.059
where:
ABS is the color absorbance as read on a spectrophotometer
^5—* are the yl of H2S in 10 ml of absorbing solution at 25°C,
1 atm.
To evaluate if oxidizable compounds other than hUS were present in
the sodium sulfide crystals, stock solutions were also prepared by with-
drawing 20 cc of CP grade HgS gas from a lecture bottle fitted with a
pressure regulator, and injecting it through a rubber septum placed on
the top of a 1-liter volumetric flask containing 0.01M sodium hydroxide
made with distilled deionized water. The flask was inverted so the HnS
gas bubbled through the solution. The solution was then mixed thoroughly
and standardized using the same procedure as for the solid sodium sulfide
standard, and a calibration curve was prepared. This curve v/as essen-
tially the same as Curve 1, Figure 1.
2. Use of a Gravimetrically Calibrated HpS Permeation Tube
A permeation tube with an emission rate of 1 yg/min of H?S estab-
lished previously by accurate weighings, was stabilized at a constant
temperature of 30°C in a permeation oven. Nitrogen was used as the
carrier gas over the permeation tube. The total'flow out of the perme-
ation oven was adjusted to one liter per minute. A midget impinger con-
taining 10 ml of cadmium hydroxide-STRactan absorbing solution was con-
nected to the total output of the permeation oven. The oven output was
sampled for seven different, known periods to time in order to collect
18
-------
.8
.5
.4
ABSORBANCE
FIGURE I
• COMPARISON OF HgS STANDARD CALIBRATION
CURVES, CADMIUM HYDROXI DE - STRACTAN
METHOD, 670nm.
I. SODIUM SULFIDE STANDARD
(100ml VOLUMETRIC FLASKS)
2. h^S PERMEATION TUBE
(MIDGET IMPINGERS)
3. SODIUM SULFIDE STANDARD
(MIDGET IMPINGERS)
@
25°C
IOmlIAtm
Figure 1. Comparison of H2S standard calibration curves
Cadmium hydroxide-STRactan method, 670 nm
19
-------
standards with the equivalent of 1-7 yl H2S per 10 ml of absorbing solu-
tion. The methylene blue color was developed in the impingers by adding
0 3 ml of the amine solution followed by 0.2 ml of the ferric chloride
solution. A calibration curve was then prepared (Curve 2, Figure 1).
3. Comparison of Calibration Standards
As can be seen in Figure 1, the calibration curve prepared by the
use of aqueous sulfide solutions in 100 ml volumetric flasks. Curve 1,
is more sensitive (greater color per yl of hLS) than the calibration
curve prepared by the use of gravimetric techniques, Curve 2.
In an attempt to determine the discrepancy between the calibration
curves prepared using aqueous sulfide solutions and gravimetric proce-
dures, a number of investigations have been performed. The use of pure
H9S gas injected directly into an aqueous solution and the resulting
calibration curve similar to that for Na2S-9H20 indicated that other
possible sulfides are not present in the sodium sulfide crystal which
could bias the results of the iodometric titration. A determination of
the sulfide concentration in the standardized aqueous solution made by
actually weighing the sodium sulfide crystal agreed with the iodometric
titration results within 2 percent. Consequently, the iodometric titra-
tion is not a source of error. The concentration of sulfide present in
the aqueous solution when pure hUS was injected into solution was calcu-
lated and found to agree with the iodometric titration within 10 percent.
This 10 percent still does not account for the discrepancy and can
largely be attributed to possible volume-measuring errors occurring
during the gas transfer from the H2S lecture bottle to the flask.
An examination of the calibration procedures using a permeation
tube indicated that the gravimetric determination of the rate of HpS
emission could be in error if other non-HLS compounds were being
emitted from the permeation tube. Discussion with the manufacturer of
the permeation tube indicated that the tubes were filled with high purity
HpS. In an attempt to physically determine if other compounds were
being emitted, a mass spectrometer was used to analyze the gas emitted
from the tube. This analysis, while at its lower range of sensitivity
due to the tube's low rate of hLS emission, did not indicate the presence
of other compounds.
Error due to variations in the rate of HLS emission was also elim-
inated as it did not vary by more than 2 percent during the life of the
permeation tube.
The collection efficiency of the cadmium hydroxide-STRactan solu-
tion in the midget impinger was also investigated by analyzing the
reagent of a second impinger placed in series while sampling a stream
of H2S. No significant H2S was recovered in the second impinger indi-
cating complete recovery of H2$ in the initial impinger
20
-------
The effect of temperatures on color development had been previously
studied in an effort to achieve agreement between the two methods of
calibration. Calibration curves were prepared using 100 ml volumetric
flasks and midget impingers immersed in a 0°C ice bath. All reagents
used were also kept at 0°C. The resulting calibration curves still
remained about 20-30 percent apart in value, but both were higher in
sensitivity. As before, the decreased temperature caused the methylene
blue color to be more intense than at ambient conditions.
Further analysis of the different calibration methods indicates
that a loss of H^S could be occurring during the color-developing pro-
cedures. When tne acidic amine solution is added to cadmium hydroxide-
STRactan reagent that contains H2S, HpS gas is formed and can escape
the solution until the ferric chloride reagent is added. The aqueous
sulfide method of calibration uses 100 ml of reagent of high sulfide
content (by sheer volume alone) in a volumetric flask where'the total
air-exposed surface area to total volume ratio was very small. The
gravimetric calibration used 10 ml of reagent in an impinger where the
air-exposed surface area to total volume was not nearly as small.
Therefore, an investigation was performed to determine the effect of
developing the same cadmium hydroxide-STRactan solution (containing
sulfide) in different glassware of widely different surface area to
total volume ratios.
Curve 3 in Figure 1 shows a calibration curve prepared using
aqueous sodium sulfide solution in a midget impinger; good agreement
was achieved with the permeation tube calibration. Curve 2. Thus,
both the aqueous sodium sulfide solution and gravimetrically calibra-
ted permeation tube procedures can produce the same calibration curves
when the same glassware is used, when the final color is developed.
This comparison of the two different methods for calibration of the
cadmium hydroxide-STRactan method indicates that similar glassware
and volumes of reagent must be used when calibrating and sampling.
Under these conditions gravimetric calibration using permeation tubes
can agree with chemical calibrations based on iodometric titrations,
giving equivalent calibration curves (Figure 1, Curves 2 and 3).
As a final check, a total sulfur analyzer with a flame photometric
detector (FPD) was used to confirm the emission rate of an H~S permea-
tion tube. The analyzer was calibrated with sulfur dioxide fs02).
As S0? and H?S both respond similarly on an FPD detector, this proce-
dure offered an independent check on the determination of the emission
rate of the H?S permeation tube. The emission rate of the H-S permea-
tion tube measured on the FPD analyzer was 0.127 yg/min. The emission
rate determined wet chemically using the gravimetrically prepared
calibration curve was 0.130 yg/min; the emission rate was 0.125 yg/min
using the calibration curve prepared from the addition of sodium
sulfide to impingers. These results further substantiate the modified
standardization procedures as being close to absolute.
21
-------
SAMPLING FOR AMBIENT HgS CONCENTRATIONS IN THE FIELD
1. 30-Minute Sampling
Ambient H~S concentrations have been measured monthly at The
Geysers since T971. 5»6 The cadmium hydroxide-STRactan method has been
used for sampling; it has been modified based on experience from time
to time.5 Nominally, 30-minute samples are taken at each location
using a sampling train as shown in Figure 2. Sample flow is maintained
at'2 liters per minute.; 10 ml of absorbing solution are used. The
effect of sample flow rate on the amount of H2S collected has been
studied. No significant differences in color absorbances were observed
when a series of samples were taken from a permeation tube outlet of
0.5, one and two liters per minute for ten-minute periods. The samples
are kept in the dark at all times to prevent photodecomposition even
during sampling (aluminum foil is wrapped around the impingers to keep
the absorbing solution in the dark); and the rnethylene blue color is
developed immediately after taking a sample instead of returning samples
to the laboratory for subsequent analysis. The color is read within
24 hours of sampling. Tests have shown that the methylene blue color
is very stable up to at least 48 hours if kept at room temperature
and in the dark. 5
The midget impingers are modified to prevent foaming and carry-over
of cadmium hydroxide-STRactan solution by increasing the length of
the impinger by 75 mm over that of the standard midget impinger and
by slipping two Teflon demister discs over the impinger air inlet tube,
about 1-2 inches from the top. Ethanol could also be added to reduce
foaming, but greater evaporation of reagent would occur. Color losses
were observed 74 hours after initial color development when ethanol
is used.
2. Twenty-four Hour Sampling
Since it was desired to simultaneously measure 24-hour average
ambient HLS and SOp concentrations at The Geysers, investigations were
made to determine Recovery of HoS during 24-hour sampling.' The sampling
train (Figure 3) was exactly the same as that for the pararosaniline
procedure employed by the EPA as their Standard Reference Method for
SO-.' Polypropylene bubblers containing the cadmium hydroxide-STRactan
absorbing solution were used instead of the standard glass impingers.
Flow control was accomplished by critical flow orifices which were
27-gauge hypodermic needles. This size gives a nominal 0.2 liter per
minute flow. Constant H?S concentrations of 18, 23, and 40 ppb in
air were provided to the sampling train.
The results obtained, shown in Table 1, were calculated using
the calibration curve prepared with 100-ml volumetric flasks, Figure 1,
Curve 1. The permeation rate from the H2S tube was calibrated against
22
-------
TO
TRAP
CRITICAL ORIFICE FLOW CONTROL
HYPODERMIC NEEDLE
J L
taMMMM^M^
RUBBER^SEPTUM
1/M
MEMBRANE FILTER
TEFLON SAMPLING PROBE
INLET
IMPINGER
( MODIFIED)
TO AIR
IPUMP
TO CRITICAL
ORIFICE FLOW
CONTROL AND
AIR PUMP
TRAP
Figure 2. 30 minute sampling train,
23
-------
CRITICAL ORIFICE FLOW CONTROL
HYPODERMIC NEEDLE
TO AIR
PUMP
RUBBETSEPTUM
MEMBRANE FILTER
FLOWMETER OR CRITICAL ORIFICE FLOW CONTROL
30 mi
20mi
lOmi
\
s
•\
IMPINGER
TO AIR PUMP
GLASS WOOI
Figure 3. 24 hour sampling train
24
-------
. Report 7485.37-76
Date
Table 1
24-Hour h^S Sample Recovery
H2S Concentration
Permeation Effluent as Found, ppm
(No. of measurements)
3/9/76
3/24/76
3/26/76
3/29/76
4/29/76
4/30/76
5/1/76
5/3/76
5/4/76
5/8/76
5/7/76
, £
40
40
40
40
23
23
23
23
18
18
18
22
23
22
26
11
12
13
14
13
11
10
(5)
(8)
(7)
(8)
(7)
(8)
(8)
(8)
(8)
(8)
(8)
25
-------
the same curve so that anomalies resulting from the use of different
calibration curves were cancelled out. A statistical analysis using
a least squares analysis of the data resulted in the line shown in
Figure 4 having a slope of 0.544. This means that using the method
described for 24-hour sampling results in about a 46 percent loss of
H2S.
Although there is a significant loss of H2S during 24-hour
sampling, the loss seems to be linear with concentration over the range
of 18 to 40 ppb. This loss cannot be immediately explained.
CONCLUSIONS
Analysis of procedures used to prepare calibration curves for
the cadmium hydroxide-STRactan method indicates that agreement can be
achieved between the use of aqueous sulfide solutions prepared from
sodium sulfide and the use of HoS permeation tubes calibrated gravi-
metrically if the same glassware and reagent volumes are used. The
use of 100 ml volumetric flasks during calibration results in less
sulfide loss after the addition of the amine reagent than in the midget
impingers used for field testing.
Analysis of the use of the cadmium hydroxide-STRactan method for
24-hour sampling indicates that 54 percent of the input H^S is recovered
from a constant concentration gas stream. This recovery Ts linear
over the range of 18 to 40 ppb.
RECOMMENDATIONS
Results have shown that a gravimetrically calibrated permeation
tube emitting 1.0 yg/min HoS can be used to calibrate the cadmium
hydroxide-STRactan method. Use of a permeation tube will more accu-
rately simulate actual field sampling techniques and will best account
for losses occurring during color development procedures. Similar
procedures are also used for primary calibration of the method for the
determination of SOp. It is therefore recommended that the cadmium
hydroxide-STRactan method be calibrated using a gravimetrically cali-
brated permeation tube emitting approximately 1.0 yg/min. Other tubes
emitting at lower rates and of different design are'more difficult
to calibrate and are not recommended without further investigation.
26
-------
ppb, VOL H2S AS MEASURED AVERAGED OVER A
24 HOUR PERIOD.
30
10
O
O
0 10 20 30 40 50 60 70
ppb, VOL H2S GENERATED FROM PERMEATION TUBE.
Figure 4, Comparison of 24-hour sampling using the cadmium
hydroxide-STRactan method (2 1pm) and concentra-
tions generated from a permeation tube,
0.215 pg/min H2S
27
-------
BIBLIOGRAPHY
1. Intersociety Committee. "Methods of Air Sampling and Analysis."
American Public Health Association, Washington, D.C. 1972. p. 726.
2. U.S. Department of Health, Education, and Welfare. U.S. Public
Health Service Center for Disease Control, National Institute for
Occupational Safety and Health, Division of Laboratories and Criteria
Development. "NIOSH Manual of Analytical Methods." Superintendent
of Documents, U.S. Government Printing Office, Washington, D.C.
1974. Section 126, pp. 1-13.
3. Staff of Air and Industrial Hygiene Laboratory, State of California,
Department of Public Health. "Determination of Hydrogen Sulfide
in the Atmosphere and Source Emissions." Recommended Method (Draft 2),
State of California, Department of Public Health, Berkeley, California.
1969.
4. Personal Communication with S. G. Sharp, Chemist, Pacific Gas and
Electric Company, Department of Engineering Research, 1976.
5. Altshuler, S. L., and S. G. Sharp. "Geysers Air Monitoring Program;
July 1970-November 1972; Progress Report." Report 7485.4-72,
Pacific Gas and Electric Company, Department of Engineering Research,
San Ramon, California. 1973.
6. Altshuler, S. L. "Ambient Air Quality at and in the Vicinity of
The Geysers, 1970-1975." Report 7485.25-75, Pacific Gas and Electric
Company, Department of Engineering Research, San Ramon, California.
1976.
7. "Reference Method for the Determination of Sulfur Dioxide in the
Atmosphere (Pararosaniline Method)." Appendix A. Federal Register,
Vol. 36, No. 228, November 25, 1971.
28
-------
APPENDIX A.
THE DETERMINATION OF HYDROGEN SULFIDE IN AIR
BY THE CADMIUM HYDROXIDE-STRACTAN METHOD
HYDROGEN SULFIDE IN AIR
Physical and Chemical Analysis Branch
Analytical Method
Analyte:
Matrix:
Procedure:
Date Issued:
Date Revised:
1. Principle
Hydrogen Sulfidc
Air
Absorption - Mcthytene
Blue — Spectrophotometric
6/9/72
1/15/74
of the Method
Method No.:
Range:
Precision:
Classification:
P&CAM 126
0.008 ppm — 50 ppm
Unknown
C (Tentative)
i.l Hydrogen sulfide is collected by aspirating a measured volume of air through an
alkaline suspension of cadmium hydroxide (Reference 1 1.1). The sulfide is pre-
cipitated as cadmium sulfide to prevent air oxidation of the sulfide which occurs
rapidly in an aqueous alkaline solution. STRactan 10 is added to the cadmium
hydroxide slurry to minimize photo-decomposition of the precipitated cadmium
sulfide (Reference il.2). The collected sulfide is subsequently determined by
Spectrophotometric measurement of the meihylenc blue produced by the reaction
of the sulfide with a strongly acid solutionofN, N-djmethyl-p-phenylenediaminc
and ferric chloride (References 11.3, 11.4, 11.5). The analysis should be com-
pleted within 24-26 hours following collection of the sample.
1.2 Hydrogen sulfide may be present in the open atmosphere ar concentrations of a
few ppb or less. The reported odor detection threshold is in the 0.7-8.4 ^g/m3
(0.5-6.0 ppb) range (References 1 1.6, 13.7). Concentrations in excess of
(100 ppb) are seldom encountered in the atmosphere.
13 Collection efficiency is variable below lO^ig/m3 and is affected by the type of
scrubber, the size of the gas bubbles and the contact time with the absorbing
solution and the concentration of H2S (References 1 l.S, 11-9, 1 1.10).
2. Range and Sensitivity
2.1 This method is intended to provide a measure of hydrogen sulfidc in the range of
1.1-lOOjBfi/m3. For concentrations above 70 iRg/m3 the sampling period Can be
reduce;! or the liquid volume increased cither before or after aspirating. (This
29
-------
method is also useful Tor the mp/m3 range of source emissions. For example,
100 mS cadmium (OlOj- STRactan 10® media in Greenberg-Smith impingers and
5-minute sampling periods have been used successful!/ for source sampling.) The
minimum detectable amount of sulfide is O.OOS jig/mC, wliich is equivalent to
0.2pg/m3 in an air sample of 1 m5 and using a final liquid volume of 25 m2.
When sampling air at the maximum recommended rate of 1.5 2/ininute for 2
hours, the minimum detectable sulfide concentration is 1.1 pg/m3 at 760 mm Hg
and 25°C.
2.2 Excellent results have been obtained by using this method for air samples having a
hydrogen sulfide content in the range 5-50 ppm.
3. Interferences
3.1 The methylene blue reaction is highly specific for sulfide at the low concentra-
tions usually encountered in ambient air. Strong reducing agents (e.g.. SO2)
inhibit color development. Even sulfide solutions containing several micrograms
sulfide/m2 show this effect and must be diluted to eliminate color inhibition. If
sulfur dioxide is absorbed to give a sulfite concentration in excess of I0pg/m2,
color formation is retarded. Up to 40 ^g/m2,of this interference, however, can be
overcome by adding 2-6 drops (0.5 rrjC/urop) of ferric chloride instead of a single
drop for color development, and extending the reaction time to 50 minutes.
3.2 Nitrogen dioxide gives a pale yellow color with the sulfide reagents at 0.5 A
-------
3.7 The choice of impingvr used to Irap H2S with the Cd(OH), slurry is very
important when measuring concentration in the range 5-50 ppm. Impincers or
bubblers having frittcil-eml gas delivery tubes arc a problem source if the sulfide in
solution is oxidized by oxygen from the atmosphere to free sulfur. The sulfur
collects on the fritted-gljss membrane and may significantly change the flow rate
of the nir sample through the system. One ,way of avoiding this problem is to use a
midget impinger with standard, glass-tapered tips.
4. Precision and Accuracy
4.1 A relative standard deviation of 3.5 per cent and a recover/ of SO per cent has
been established with hydrogen sulfide permeation tubes (Reference 11.2).
5. Advantages and Disadvantages of the Method
5.1 Effect of Light and Storage - Disadvantage
5.1.1 Hydrogen sulfirlo is readily volafili7.ed.from aqueous solution when the pH
is below 7.0. Alkaline, aqueous sulfide solutions are very unstable because
sulfide ion is rapidly oxidized by exposure to the air.
5.4.2 Cadmium sulfide is not appreciably oxidized even when aspirated with
pure oxygen in the dark. However, exposure of an impinger containing
cadmium sulfide to laboratory or to more intense light sources produces
an immediate and variable photo-decomposition. Losses of 50-90 per cent
of added sulfide have been routinely reported by a number of labora-
tories. Even though the addition of STRactan 10® to the absorbing
solution controls the photo-decomposition (Reference 1 1.2), it is neces-
sary to protect the impingsr from light at all times. This is achieved by the
use of low actinic glass impingers, paint on the exterior of the impingsrs.
or an aluminum foil wrapping.
6. Apparatus
6.1 Sampling Equipment. The sampling unit for the impinger collection method
consists of the following components:
6.1.1 A graduated 25-rnC midget impinger with a standard glass-tapered gas
delivery tube containing the absorbing solution or reagent.
6,1.2. A pump suitable for delivering desired flow rates with a minimum capac-
ity of 22pm through the impuiger. The sampling pump is protected from
splashover or water condensation by an adsorption tube loosely packed
with a plug of glass wool ami inserted between the exit arm of the
impinger and the
31
-------
6.1.3 An integrating volume meter such as a dry-gas or wet-test motor. The air
meter must be capable of measuring the.air flow within ±2 per cent. A wet'
or dry-pas meter, with contacts on the 1-fect3 or 10-1 dial to record air
volume, or a specially calibrated rotamcter, is satisfactory. Instead of
these, calibrated hypodermic needles may be used as critical orifices if the
pump is capable of maintaining greater than 0.7 atmospheric pressure
differential across the needle (Reference 11.11).
6.1,4 Thermometer.
6.1.5 Manometer.
6.1.6 Stopwatch.
6.2 Associated laboratory glassware.
6.3 Colorimeter with red filter or spectrophotometer at 670 nm.
6.4 Matched cells, 1-cm path length.
7. Reagents
All reagents must be ACS analytical reagent quality. Distilled water should conform to
the ASTM Standards for Referee Reagent Water.
All reagents should be refrigerated when not in use.
7.1 Amine-sulfuric Acid Stock Solution. Add 50 m£ concentrated sulfuric acid to
30 m£ water and cool. Dissolve 12 g of N, N-dimethyl-p-phenylenediamine
dihydrochloridc (para-aminodimethylaniline) (redistilled if necessary) in the acid.
Do not dilute. The stock solution may be stored indefinitely under refrigeration.
7.2 Amine Test Solution. Dilute 25 m£ of the Stock Solution to 1 liter with 1:1
sulfuric acid.
7.3 Ferric Chloride Solution. Dissolve 100 g of ferric chloride, FeC23 . 6H2O, in
water and dilute to 100 m£.
7.4 Ammonium Phosphate Solution. Dissolve 400 g of diammonium phosphate,
(NH4)2HP04, in water and dilute to 1 liter.
7.5 STRactan 10®, (Arabinogalactan) Available from Stein-Ha!! and Company, Inc.,
385 Madison Avenue, New York, New York.
32
-------
7.6 Absorbing Solution. Dissolve 4.3 g of 3CdSO4. 3 H2O. and 0.3 g sodium
hydroxide in separate portions of water. Mix and add 10 g STRactan 10® and
dilute to \ liter. Shake the resultant suspension vigorously before removing each
aliquot. The STRactan®-c;idmium hydroxide mixture should be freshly prepared.
The solution is only stable for 3 to 5 days.
7.7 H2S Permeation Tube. Prepare or purchase* a triple-walled or thick-walled
Tenon® permeation tube (References 11.10, 11.12, 11.13, 11.14. 11.15) which
delivers hydrogen sulfide at a maximum rate of approximately 0,1 ^g/minute at
25°C. This loss rate will produce a standard atmosphere containing 5Qjug/m3
(36 ppb H2S when the tube is swept with a 2 £/minute air flow. Tubes having
H2S permeation rates in the range of 0.004-0.33 pig/minute will produce standard
air concentrations in the realistic range of 1-90 pg/m3 H2S wjth an air How of
1.5C/min.
7.7.1 Concentrated, Standard Sulfide Solution
Transfer freshly boiled and cooled 0.1M N'aOH to a liter volumetric flask.
Flush with purified nitrogen to remove oxygen and adjust to volume.
(Commercially available, compressed nitrogen contains trace quantities of
oxygen in sufficient concentration to oxidize the small concentrations of
sulfide contained in the standard and dilute standard sulfide standards.
Trace quantities of oxygen should be removed by passing the stream of
tank nitrogen through a Pyrex or quartz tube containing copper turnings
heated to 400-450°C.) Immediately stopper the flask with a serum cap.
Inject 300 rn£ of H-S gas through the septum. Shake the flask. Withdraw
measured volumes of standard solution with a 10 m2 hypodermic syringe
and fill the resulting void with an equal volume of nitrogen. Standardize
with standard iodine and thiosulfatc solution in an iodine flask under a
nitrogen atmosphere to minimize air oxidation. The approximate con-
centration of the sulfide solution will be 440 pg su!fide/m2 of solution.
The exact concentration must be determined by iodine-thiosulfate
standardization immediately prior to dilution.
For the most accurate results in the iodomctric determination of sulfide in
aqueous solution, the following general procedure is recommended:
1. Replace the oxygen from the flask by flushing with an inert gas such as
carbon dioxide or nitrogen.
'Available from Metrom'cs, Inc., 3201 Porter Drive. Palo AUo. California 94304, or PolvSciencc Corp.,
009 Pitncr Avcnuo. Evanston, Illinois 60202.
33
-------
2. Add an excess of standard iodine, acidify, and back titrate with standard
thiosulfatc and starch indicator (Reference 1 1.16).
7.7.2 Diluted Standard Sulfidc Solution
Dilute 10 inC of the concentrated sulfide solution to 1 liter with freshly
boiled, distilled water. Protect the boiled water under a nitrogen atmo-
sphere while cooling. Transfer the dcoxygenated water to a flask pre-
viously purged with nitrogen and immediately stopper the flask. This
sulfidc solution is unstable. Therefore, prepare this solution immediately
prior to use. This test solution will contain approximately 4 ^g sulfidc/mC.
8. Procedure
8.1 Cleaning of Equipment. Ail glassware should be thoroughly cleaned; the following
procedure is recommended:
8.1.1 Wash with a detergent and tap water solution followed by tap water and
distilled water rinses.
8.1.2 Soak in 1:1 or concentrated nitric acid for 30 minutes and then follow
with tap, distilled, and double distilled water rinses.
8.2 Collection and Shipping of Samples
8.2.1 Pipet 10 m2 of the absorbing solution (Section 7.6) into the midget
impinger. The addition of 5 m£ of 95 per cent ethanoi to the absorbing
solutfon just prior to aspiration controls foaming for 2 hours (induced by
the presence of STRactan 10®. In addition, 1 or 2 Teflon demister discs
may be slipped up over the impinger air inlet tube to a height approxi-
mately 1-2" from the top of the tube.
8.2.2 Connect the impinger (via the absorption tube) to the vacuum pump with
a short piece of flexible tubing. The minimum amount of tubing necessary
to make the joint between the prefilter and impinger should be used. The
air being sampled should not be passed through any other tubing or other
equipment before entering the impinger.
8.2.3 Turn on pump to begin sample collection. Care should be taken to
measure the flow rate, time and/or volume as accurately as possible. The
sample should be taken at a flow rate of 1.5 Spm.
34
-------
8.2.4 After sampling, the impingcr stem qan be removed and cleaned. Tap the
stem gently against the inside wall of the impingcr bottle to recover as
much of the sampling solution as possible. V.'ash the stem with a small
amount (1-2 m?) of unused absorbing solution and add the wash lo the
impinger. Then the impinger is sealed with a hard, non-reactive stopper
(preferably Teflon). Do not seal with rubber. The stoppers on the
impingcrs should be tightly sealed to prevent leakage during shipping. If it
is preferred to ship the impingers with the stems in, the outlets of the
stem should be sealed with Para film or other non-rubber covers, and the
ground glass joints should be sealed (i.e., taped) to secure the top tightly.
8.2.5 Care should be taken to minimize spillage or loss by evaporation at all
times. Refrigerate samples if analysis cannot be done within a day.
8.2.6 Whenever possible, hand delivery of the samples is recommended. Other-
wise, special impinger shipping cases designed by NIOSH should be used
to ship the samples.
8.2.7 A "blank" impinger should be handled as the other samples (fill, seal and
transport) except that no air is sampled through this impinger.
83 Analysis
8.3.1 Add 1.5 mfi of the amine-test solution to the midget impinger through the
air inlet tube and mix.
83.2 Add 1 drop of ferric chloride solution and mix. (Note: See Section 3.1 if
SO2 exceeds 10 /Jtg/m2 in the absorbing media.)
83.3 Transfer the solution to a 25 m2 volumetric flask. Discharge the color due
to the ferric ion by adding 1 drop ammonium phosphate solution. If the
yellow color is not destroyed by 1 drop ammon'rum phosphate solution,
continue dropsvise addition until solution is decolorized. Make up to
volume with distilled water and allow to stand for 30 minutes.
83.4 Prepare a zero reference solution in the same manner using a 10 m£
volume of absorbing solution, through which no air has been aspirated.
83.5 Measure the absorbancc of the color at 670 nm in a spectrophotometer or
colorimeter set at 100 per cent transmission against the zero reference.
35
-------
9.- Calibration and Standards
9.1 Aqueous Sulfide
9.1.1 Place 10 m5 of the absorbing solution in each of a series of 25 in?
volumetric flasks and add the diluted standard sulfidc solution, equivalent
to 1, 2, 3, 4, and 5 ^g of hydrogen sulfide to the different flasks.
9.1.2 Add 1.5 m2 of aminc-acid test solution to each flask and mix.
9.1.3 Add 1 drop of ferric chloride solution to each flask. Mix, make up to
volume and allow to stand for 30 minutes.
9.1.4 Determine the absorbance in a spectrophotometer at*670 nm, against the
sulfide-free reference solution.
9.1.5 Prepare a standard curve of absorbance vs. jig H2S/:n2.
9.2 Gaseous Sulfide. Commercially available permeation tubes containing liquefied
hydrogen sulfide may be used to prepare calibration curves for use at the upper
range of atmospheric concentration. Triple-walled tubes, drilled rod and micro
bottles which deliver hydrogen suifide within a minimum range of 0.1-1.2 ^g/
minute at 25°C have been prepared by Thomas (Reference 1 1.10); O'Keeffe
(References 11.12, 11.13); Scaringelli (References 11.14, 11.15). Preferably the
tubes should deliver hydrogen sulfide within a loss rate range of 0.003-0.28
pg/minute to provide realistic concentrations of H2S (1.5-140 Mg/m3,
1.1-100 ppb) without having to resort to a dilution system to prepare the
concentration range required for determining the collection efficiency of midget
impingers. Analyses of these known concentrations give calibration curves which
simulate all of the operational conditions performed during the sampling and
chemical procedure. This calibration curve includes the important correction for
collection efficiency at various concentrations of hydrogen sulfide.
9.2.1 Prepare or obtain a Teflon® permeation tube that emits hydrogen sulfide
at a rote of 0.1-0.2 pg/minute (0.07-0. M /uC/minute at standard conditions
of 25°Cand 1 atmosphere). A permeation tube with an effective length of
2-3 cm and a wail thickness of 0.318 cm will yield the desired permeation
rate if held at a constant temperature of 25°C ± 0.1°C. Permeation tubes
containing hydrogen sulfide are calibrated under a stream of dry nitrogen
to prevent the precipitation of sulfur in the walls of the tube.
9.2.2 To prepare standard concentrations of hydrogen sulfide, assemble the
apparatus consisting of a water-cooled condenser, constant iL-mperaturc
bath maintained at 25°C ±0.1°C cylinders containing pure dry nitrogen
36
-------
and •pure dry oir with appropriate pressure regulators, needle valves and
flow meters for the nitrogen ond dry air, diluent-streams. The diluent
gases are brought to temperature by passage through a 2-meter-long
copper coil immersed in the water bath. Insert a calibrated permeation
tube into the centra! tube of the condenser, maintained at the selected
constant temperature by circulating water from the constant-temperature
bath, and pass a stream of nitrogen over the tube at a fixed rate of
approximately 50 niC/minutc. Dilute this gas stream to obtain the desired
concentration by varying the flow rate of the clean, dry air. This flow rate
can normally be varied from 0.2-15 2/minute. The flow rate of the
sampling system determines the lower limit for the flow rate of the
diluent gases. The flow rate of the nitrogen and the diluent air must be
measured to an accuracy of 1-2 per cent. With a tube permeating
hydrogen sulfide at a rate of 0.1 ju£/minutc, the range of concentration of
hydrogen sulfide will be between 6-400 /ug/m3 (4-290 ppb), a generally
satisfactory range for ambient air conditions. When higher concentrations
are desired, calibrate and use longer permeation tubes.
9.2.3 Procedure for Preparing Simulated Calibration Curves
Obviously one can prepare a multitude of curves by selecting different
combinations of sampling rate and sampling time: The following descrip-
tion represents a typical procedure for ambient air sampling of short
duration, with a brief mention of a modification for 24 hour sampling.
1. The system is designed to provide an accurate measure of hydrogen sulfide
in the 1.4-84 A
-------
where:
C = Concentration of H2S in ppm
P - Permeation rate in pg/minute
M = Reciprocal of vupor density, 0.719
R = Flow rate of diluent air, liter/minute
r = Flow rate of diluent nitrogen, liter/minute
3. The data for a typical calibration curve are listed in Table 1.
TABLE 1
TYPICAL CALIBRATION DATA
Amount of
Concentrations H5S in Absorbanca
H:S, ppb pl/1 88 liters j?f Sample
1 .144 .010
5 .795 .056
10 1.44 .102
20 2.88 .205
30 4.32 -307
40 5.76 .410
50 7.95 .512
60 8.64 .615
4. A plot of the concentration of hydrogen sulfide in ppm (x — axis) against
absorbance of the final solution (y — axis) will yield a straight line, the
reciprocal of the slope of which is the factor for conversion of absorbance
to ppm. This factor includes the correction for collection efficiency. Any
deviation from the linearity at the lower concentration range indicates a
change in collection efficiency of the sampling system. If the range of
interest is below the dynamic range of the method the total volume of air
collected should be increased to obtain sufficient color within the
dynamic range of the colorimetric procedure. Also, once the calibration
factor has been established under simulated conditions the conditions can
be modified so that the concentration of II2S is a simple multiple of the
absorbance of the colored solution.
38
-------
5. For long-term sampling of 24-hour duration, the conditions can be fixed
to collect 1 200 S. of sample in a larger volume o'f STRactnn 10®-cadmium
hydroxide. For example, for 24 hours at 0.83 2/min, approximately
12002 of air are scrubbed. An aliquot representing 0.1 of the entire
amount of sample is taken for the analysis.
6. The remainder of the analytical procedure is the same as described in the
previous paragraph.
9.2.4 The permeation tubes must be stored in a wide-mouth glass bottle con-
taining silica gel and solid sodium hydroxide to remove moisture and
hydrogen sulfide. The storage bottle is immersed to two-thirds its depth in
a constant temperature water bath in which the water is controlled at
25°C±O.I°C.
Periodically, (every 2 weeks or less) the permeation, tubes are removed and
rapidly weighed on a semimicro balanqe (sensitivity ±0.01 mg) and then
returned to the storage bottle. The weight loss is recorded. The tubes are
ready for use when the rate of weight loss becomes constant (within ±2
per cent).
10. Calculations
. 10.1 Gaseous Sulfide
10.1.1 Determine the sample volume in liters from the gas meter or flow meter
readings and time of sampling. Adjust volume to 760 mm mercury and
25°C(VS).
10.1.2 Determine the concentration of HZS in
where:
/igHjS = micrograms hydrogen sulfide determined
10~3 = conversion factor, m3/£
10.2 Caseous Sulfide from Aqueous Sulfide
10.2.1 Determine the sample volume (V) in liters from the gas meter or flow
meter readings and time of sampling. Adjust volume to 760 mm mercury
and 25*C (Vs), using the correction formula:
39
-------
P
V = V x — x —
s 760 (T •»• 273}
where:
V - Volume of air in liters at standard conditions
V = Volume of air sampled in liters
P = Barometric pressure in mm Ilg
T = Temperature of sample air in °C
10.2.2 Using the Be.ers-Law Standard curve of absorbancc vs. jug S= ion, deter-
mine tig S~ ion in the sampling impinger corresponding to its absorbance
reading at 670 nm.
10.2.3 Calculate the concentration of H2S in the aspirated volume of air using
the formula:
„ c _ pg S= x 24.45
ppm.H2S=
*>
where:
j*g S~ = micrograms sulfide ion (Section 10.2.2)
24.45 = molar volume of an idea! gas at 25°C and 760 mm Hgr
MW = mass of sulfide ion, 32.06
1 J . References
11.1 Jacobs, M.B., Bravcrman, M.M., and Hochheiser.S. "Ultramicro determination of
sulfides in air." Anal. Chem. 29; 1349 (1957).
11.2 Ramesberger, W.L., and Adams, D.F. "Improvements in the collection of
hydrogen sulfides in cadmium hydroxide suspension." Environ. Sci. &. Tech. 3;
258 (1.969).
1 1.3 Mecklenburg, W., and Rozcnkronzcr, R. "Colorimetric determination of hydrogen
sulfidc." A. Anorg. Chem. 86; 143 (1914).
11.4 Almy, L.H. "Estimation of hydrogen sulfidc in proteinaceous food products.").
Am. Chem. Soc. 47; 1381 (1925).
11.5 Shcppard, S.E., and Hudson, J.H. "Determination of labile sulfide in gebtin and
proteins." Ind. Eng. Chem., Anal. I-d. 2;73 (1930).
40
-------
11.6 Adams, D.F., Young, F.A., and Lulir, R.A. "Evaluation of an odor percep-
tion threshold test facility." Tappi. 5 1; 62A (1968).
11.7 Leonardos, G., Kendall, D.,and Barnard, N. "Odor threshold determinations
of 53 odorant chemicals." J. Air Pollut. Contr. Assoc. 19; 91 (1969).
II.8 Bostrom, C.E. "The absorption of sulfur dioxide at low concentrations
(ppnni) studied by an isotopic tracer method." Air £ Water Pollut. Int. J, 9;
333 (1965).
11.9 Bostrom, C.E. "The absorption of low concentrations (pprim) of hydrogen
sulfide in a Cd(OH)2 suspension as studied by an isotopic tracer method."
Air & Water Pollut. Int. J. i_p_;435 (1966).
11.10 Thomas, B.L., and Adams, D.F. Unpublished information.
11.11 Lodge, J.P., Pate, J.B., Ammons, B.E., Swanson, G.A. "The use of hypoder-
mic needles as critical orifices." J. Air Poll. Control Assoc. 16; 197 (1966).
11.12 O'Keeffe, A.E., and Ortman, G.C. "Primary standards for trace gas analysis."
Anal.Chem. 3_8; 760 (1966).
11.13 O'Keeffe, A.E., and Ortman, G.C. "Precision picograrn dispenser for volatile
substances." Anal. Chem. 3_9; 1047 (1967).
11.14 Scaringelli, P.P., Frey, S.A., and Saltzman,B.E. "Evaluation of Teflon
permeation tubes for use with sulfur dioxide." Am. Ind. Hyg. Assoc. J. 28;
260 (1967). ~~
11.15 Scaringelli, P.P., .Rosenberg, E., and Rehme, K. "Stoichiometric comparison
between permeation tubes and nitrite ion as primary standards for the
eobrimctric determination of nitrogen dioxide." Presented before the Divi-
sion of Water, Air and Waste Chemistry of the American Chemical Society,
157th National Meeting, Minneapolis, Minn., April 1969.
11.16 Kolthoff, I.M., and Elving, P.J., Eds. Treatise on Analytical Chemistry. Part
II, Analytical Chemistry of the Elements, V. 7. Intcrscience Publishers, New
York,1961.
11.17 Bock,' R., and Puff, H.J. "Bestimnumg von sulfid mir. cirier sulfidion-
encmpfindlichcn clcktrode." Z Anal. Cham. 240; 381 (1968).
41
-------
ABSTRACT
THE MEASUREMENT AND DISTRIBUTION OF
GEOTHERMAL SULFUR POLLUTANTS IN THE GEYSERS AREA
L. A. Cavanagh
Stanford Research Institute, Menlo Park, California
The use of geothermal energy for the production of electricity
has been viewed with increased interest as the availability of
natural gas has decreased and cost of petroleum has increased.
The Geysers Area is the location of ten Pacific Gas and Electric
Company (PG&E) geothermal power plants producing some 502 mega-
watts of power. Hydrogen sulfide (H2S) complaints frequently
arise in the populated regions downwind of the area. Due to
its concern that geothermal expansion would adversely affect air
quality in the surrounding region, PG&E and steam suppliers have
contracted with Stanford Research Institute to install and
operate an air quality and meteorological monitoring network in
The Geysers Area for a period of two years. The objectives of
the study are to analyze H^S distribution and the meteorological
conditions that lead to episodes or intervals of numerous com-
plaints about ^S odors in the populated areas to the east. The
network has been in operation for the first year of a two-year
program analysis of meteorological conditions conducive to
transport of H^S from The Geysers Area to the populated areas
to the east will be described.
ABSTRACT
FIELD DETERMINATION OF HYDROGEN SULFIDE
G. A. Frye and D. W. Wheeler
Aminoil USA, Inc., Santa Rosa, Calif.
Aminoil has been collecting and analyzing samples from geother-
mal wells since 1969. Due to the potential environmental impact
of the hydrogen sulfide in the geothermal fluids, a continuing
effort has been made to determine the sulfide concentration in
the fluids. After comparing several methods for accuracy, inter-
ference of other constituents, and adaptability to field oper-
ations, Aminoil believes the silver—ion,selective electrode
titration is well suited for geothermal operations.
42
-------
ABSTRACT
CHEVRON PLANS FOR ON-SITE SAMPLING AND
ANALYZING OF GEOTHERMAL WATERS
W0 J. Subcasky
Chevron Oil Field Research Co., La Habra, California
In evaluating geothermal waters for power generation, it is im-
portant to know the composition of the fluid and how that fluid
behaves when temperature and pressure are changed during process-
ing. Chevron has designed and constructed on-site equipment to
continuously sample fluid from a production stream and evaluate
the fluid under various temperature and pressure conditions in
either a flash mode or a binary heat-exchange mode. Some meas-
urements to be made under various conditions are (1) composition
and volume of noncondensible gases, (2) types and rates of cor-
rosion, (3) types and rates of scale deposition on equipment sur-
faces, and (4) types and amounts of colloidal precipitates form-
ing in effluents and their effect on brine reinjectivity. The
presentation includes a discussion of the ^equipment design and
preliminary plans for its use in evaluating produced geothermal
water at the Heber site.
43
-------
ABSTRACT
DEVELOPMENT OF A STANDARD METHODS MANUAL FOR SAMPLING
AND ANALYSIS OF GEOTHERMAL FLUIDS AND GASES
E. M. Woodruff
Battelle-Northwest Laboratories, Richland, Washington
Work on the manual of standard methods began early in 1976 under
ERDA-DGE sponsorship in response to needs apparent in our own
programs as well as those expressed by other organizations. The
scope of this program and progress to date are reported here.
The objective is to prepare a manual of recommended methods for
sampling and analysis of geothermal fluids and gases which will
assure accuracy, reliability and intercomparability of reported
results, and to assist industry by reducing the analytic methods
research required by each organization. Our approach has been
to first identify properties and constituents of interest to the
different aspects of geothermal energy exploration, development
and utilization; then summarize available methods with comments
on sample and instrument requirements and other factors which
would influence selection including cost, accuracy and precision,
interferences, or other limitations,, This preliminary review
was distributed to interested organizations in August 1976 ' '
for comment and additional input.
Those methods requiring verification or modification for appli-
cation to geothermal brines will then be tested by a small group
of participating laboratories. Finally, a manual summarizing
these tests and recommendations will be published.
The preliminary review of constituents of interest and published
analytic methods have been compiled in table form. During this
effort, two additional data formats were developed to aid the ad-
ministrator as well as the scientist in establishing geothermal
J. G. Douglas, et al., "Geothermal Water and Gas - Collected
Methods for Sampling and Analysis, Comment Issue," BNWL-2094,
Battelle-Northwest, Richland, Washington, August 1976.
44
-------
capabilities. A graph was prepared to illustrate concentration
ranges for constituents reported for geothermal brines through-
out the world. Individual data points are included and although
they are not completely comprehensive, the analyst can, at a
glance, become familiar with expected ranges, distribution
trends within ranges, and how frequently an individual constit-
uent was included in brine analyses. To relate analytic require-
ments to specific objectives or tasks encountered in geothermal
energy development and utilization, a matrix format has been
devised. Tasks are matched versus properties and constituents
with notations are incorporated to identify several analytic
requirements.
This preliminary draft contains many methods and options that
have not been evaluated for brine applications,however, they
should serve to stimulate interaction between interested organ-
izations and produce refinements that will result in a final
working document aimed at reducing the research and decision-
making burden of laboratories and industry. Two critical re-
views of this work have been obtained by contract, and formu-
lation of the analytic phase of the program is in progress.
45
-------
Sampling and Characterization of Suspended
Solids in Brine from Magmamax #1 Well
J. H. Hill and C. H. Otto, Jr.
University of California
Lawrence Livermore Laboratory
Livermore, California 94550
ABSTRACT
When high temperature, high salinity geothermal brines are cooled,
appreciable quantities of solids precipitate. These suspended solids
can cause extensive erosion and plugging in power plant components. In
addition, there is considerable concern that reinjection of these solids
with spent brine may plug the geologic formation adjacent to the re-
injection wells. This report describes work done to sample these solids
and characterize them so that processes can be devised for their removal
or control.
Suspended solids produced in brine from Magmamax #1 well consist
primarily of an iron-rich amorphous silica gel. A sampling apparatus to
filter these solids out of high temperature, high pressure brine streams
is described. Their chemical composition and physical description is
discussed. Rates of production of suspended solids are given for brines
flashed through nozzles and flashed in steam separators. Control of
solids production by acidification to a pH <4.5 is described. At this
time, control by acidification seems to offer an attractive alternative
to conventional methods (filtering, settling, etc.) for removing suspended
solids from spent brine prior to reinjection.
INTRODUCTION
When high salinity brine from the Salton Sea Geothermal Field (SSGF)
is cooled, sparingly soluble constituents precipitate. Most of these
solids remain in suspension as brine flows through energy conversion
equipment. Unfortunately, process equipment handling high velocity brine
and solids can be rapidly and severely damaged by erosion. Also, these
solids plug valves and piping and may shorten the lifetime of reinjection
wells by destroying geologic formation porosity. This report describes
work done to sample these suspended solids and characterize them both
physically and chemically so that problem areas can be identified and
control techniques can be evaluated. This work was done with brine from
Magmamax #1 well which is located in the Salton Sea Geothermal Area and
operated by Magmamax Power Co. Some of the samples were taken from the
San Diego Gas and Electric Co. Geothermal Loop Experimental Facility (GLEF).
46
-------
Others were obtained from the Lawrence Livermore Laboratory Field Test
Unit (LLLFTU).
RESULTS AND DISCUSSION
The apparatus used to collect suspended solids from high tempera-
ture, high pressure brine lines is shown in Figure 1. Originally, three
sintered stainless steel filters were used in this apparatus to obtain a
particle-size distribution for the solids. The pore size of the filters
was 165, 65, and 10-ym. After examination of the collected solids in-
dicated that they were composed primarily of an amorphous iron-rich
silica gel, only the 10 ym filter was used. In use, the apparatus is
fitted with valves and a pressure gauge so that the differential pressure
across the filter can be regulated to <10 psi. Samples were taken from
10-in. brine lines by attaching this apparatus to a traversing sample
probe (1). This probe is designed so that it can be inserted through a
valve and traversed from one side of the pipe to the other. Thus it is
possible to obtain a sample of the brine that is relatively free of solids
that have collected on the wall of the pipe or around the valve. To avoid
errors caused by the precipitation of solids when brine contacts the cold
probe, the probe is heated by running the hot brine through a bypass
before the valves on the sampler are opened.
Chemical analyses of samples-of suspended solids taken from the
pipeline near the reinjection well of the GLEF are shown in Table I. These
results indicate that the solids are rich in Fe and,Si. Minor amounts of
heavy metals such as Pb and Zn are also present. X-ray diffraction
analysis indicates that the iron-rich silica matrix is amorphous. Also,
the heavy metals occur primarily as sulfides. These samples also contain
a considerable amount of Ca and Mn which occur primarily as carbonates.
Their presence is attributed to the reinjection of high pH (9 to 10)
condensate into the spent brine during this period of operation.
Suspended solids taken from the brine at Magmamax #1 wellhead and
from the separator on the LLLFTU were found to be almost pure PbS. Electron
micrographs of these solids are shown in Figure 2. These micrographs and
chemical analyses indicate that these solids are submicron sized-crystals of
PbS which are cemented into larger agglomerates.
Electron micrographs of suspended solids taken from the reinjection
pipeline of the GLEF are shown in Figure 3. These micrographs show the
typical silica gel structure of the amorphous iron-rich silica matrix.
This is the same structure previously found in silica scale deposited from
geothermal brines.
The extent of solids production in the GLEF and the LLLFTU was deter-
mined by sampling suspended solids at the Magmamax #1 wellhead, the
separator of the LLLFTU, the reinjection pipeline of the GLEF, and from
•brine which had been expanded through nozzles to atmospheric pressure in
47
-------
ULTEK CURVAC
FLANGES
1.875'I.D. SST
TUBING
SPACER (3)
GASKET
SST FILTER
Figure 1. Suspended solids sampler
48
-------
90° C wt. % (normalized
150-300 psi „
ELEMENT FILTER SIZE
65 ym 10 ym
Si 50 43
S 1.7 0.9
Ca 10 5.4
Mn 3.0 4.6
Fe 28 37
Cu 0.5 0.6
Zn 1.6 2.1
As 0.3 0.6
Sr 0.03 0.09
Mo 0.01 0.009
Ag 0.009 0.06
Cd 0.003 0.006
Sn 0.01
Sb 0.003 0.15
Ba 0.13 0.6
Pb 1.4 4.5
to 100)
MX-3-L1
4
FILTER SIZE
65 ym
32
0.7
4.4
4.1
48
0.8
1.2
0.5
0.03
0.03
0.003
-
0.003
0.03
0.08
1.0
10 ym
30
0.4
1.9
3.0
57
0.3
0.8
0.8
0.05
0.04
0.003
-
0.004
0.04
0.14
1.1
TABLE I. SEMI-QUANTITATIVE ANALYSIS OF SUSPENDED
SOLIDS FROM THE GLEF.
Brine
Sample
Magmamax #1 Wellhead
Separated Brine
Expanded through nozzle
Magmamax #3 Reinjection
line from GLEF Operation
Concentration of Solids
mg solids/kg brine
26
20
475
480
Composition of
Solids
PbS
PbS
Fe-Si matrix with
minor PbS
Fe-Si matrix with
minor PbS and ZnS
For a 10-MW powerplant using 800,000 Ib/hr of brine,
solids production would be approximately 5 tons/day.
TABLE II. SUSPENDED SOLIDS PRODUCTION
49
-------
(a)
(b)
U1
O
56.000X
179 nm
(c)
80.000X
70.000X
143 nm
(d)
52.000X
(b)
I 1 100.000X I 1 50.000X
100 nm 200 nm
100.000X
100 nm
(c)
125 nm
190 nm
Figure 2. Electron micrographs of
solids from Magmamax #1
wellhead
Figure 3. Electron micrographs of
solids from GLEF
reinjection line
-------
the LLLFTU. These results are shown in Table II. These data indicate
that brine from Magmamax #1 wellhead contains about 26 ppm of suspended
solids. However, about 475 ppm of suspended solids are produced when
brine is flashed to remove steam in the GLEF or when brine is flashed
through nozzles to atmospheric pressure in the LLLFTU. For a 100-MH
powerplant, 475 ppm of suspended solids corresponds to approximately
5 tons/day which would be reinjected with the spent brine.
The time elapsed before solids will plug a reinjection well depends
largely on the porosity and fracturing that exists in the formation.
However, there can be little doubt that plugging will eventually occur.
Therefore, methods must be developed either to prevent the formation of
solids in process brine or to remove solids from the spent brine. One
technique that we have tried to inhibit the formation of solids is
acidification.
The rate of suspended solids production in Magmamax #1 brine was
determined as a function of pH during acidification experiments conducted
in October and November, 1976. In these experiments, separated high
temperature (220°C), high pressure (250 psi) brine was acidified and
expanded through nozzles in the LLL brine modification apparatus to
atmospheric pressure. Samples used to determine suspended solids pro-
duction were taken from elbows in the exhaust system about 16 inches
downstream from the nozzles. The temperature of the brine at the sampling
point was about 95°C. All samples were transferred to a constant tem-
perature bath a-t 85°C immediately after they were taken. To determine the
rate of solids production, the samples were removed from the constant
temperature bath at timed intervals and immediately filtered through glass
frits with a nominal pore size 4 to 5.5 ym. The solids were then washed
with distilled water, dried at 110°C and weighed.
Initial experiments were conducted over intervals up to 24 hrs. with
samples in the pH range from 1.6 to 5.8. Results are shown in Figure 4.
These data indicate that the precipitation of suspended solids from
Magmamax #1 brine in this pH range is essentially complete in 4 hrs. or
less. They also- show that acidification of the brine to a pH <4.5 greatly
inhibits the production of suspended solids. The amount of solids pro-
duced at a pH <4.5 is <1/10 the amount produced in unmodified brine
(pH 5.8). These results were confirmed by additional experiments conducted
over shorter time intervals.
The quantity of suspended solids produced as a function of pH in a
4-hr time interval is shown graphically in Figure 5. This figure
emphasizes the decrease in solids production as the brine is acidified
to a pH <4.5. It shows that over 400 ppm of solids precipitated from un-
modified brine (pH >5.8). However, acidification of the brine to a
pH <4.5 reduced solids production to <30 ppm.
51
-------
Ul
to
500
400
~ 300
a.
CO
C3
g.2M
100
1 1 1 1 1 1 —
. pH 6.0
° ^"
?
•
•
.
" ^~U P« 2.5
1 1 1 — *~~\ 1 \ — \ i
600
12345678
ELAPSED TIME (HOURS)
Figure 4. Rate of solids production at
85°C in separated Magmamax
No. 1 brine
500
400
300
200
CO
100
T
1 I I I T
SEPARATED BRINE
TEMPERATURE 85°C
TIME 4 HOURS
5678
Figure 5. Suspended solids produc-
tion as a function of pH
-------
The solubility of precipitated solids in acid solutions was also
investigated. In this experiment, a 5 gallon sample of brine was
taken from Magmamax #1 well and allowed to cool to ambient temperature.
Aliquots of the brine together with its suspended solids were then
treated as shown in Table III. These data indicate that the unmodified
brine contained 580 ppm of suspended solids. They also show that
-------
Treatment
Concentration of solids
after treatment9
mg/kg brine
None
Acidified to pH 4.1, allowed to
stand 5 hrs. at room temperature.
Acidified to pH 2.0, allowed to
stand 5 hrs. at room temperature.
Acidified to pH 2.0, allowed to
stand 3 hrs. at 90°C.
Leached 1 hr. in 1 M HC1, 0.1 M
at 50°C.
580
615
460
400
390
All samples were filtered, washed with
and dried at 110°C.
TABLE III. RESPONSE OF SUSPENDED SOLIDS TO ACID TREATMENT
NOTICE
This report was prepared as an account of work
sponsored by the United States Government. Neither the
United States nor the United States Energy Research
& Development Administration, nor any of their
employees, nor any of their contractors, subcontractors,
or their employees, makes any warranty, express or
implied, or assumes any legal liability or responsibility
for the accuracy, completeness or usefulness of any
information, apparatus, product or process disclosed, or
represents that its use would not infringe
privately-owned rights.
NOTICE
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.
54
-------
ABSTRACT
SAMPLING AND ANALYSIS OF HOT AND COLD SPRING WATERS AND ASSOCIATED
ROCK AND SOIL SAMPLES FROM POTENTIAL GEOTHERMAL RESOURCES AREAS IN
NORTH CENTRAL NEVADA.
H. R. Bowman, H. Wollenberg, F. Asaro and A. Hebert
Lawrence Berkeley Laboratory, Berkeley, California
The principal objective of the geochemical section of the geo-
thermal resource assessment program at L.B.L. is to develop and
evaluate new techniques and methods for sampling and analyzing
materials associated with hot spring activities. The analytical
methods were mainly (1) neutron-activation-analysis (NAA), (2)
non-dispersive xray-fluorescence-analysis (XRF), and (3) gamma-
ray- spectrome try (GRS). These methods together are capable of
testing for over 50 elemental abundances in rock, soil and eva-
porated water samples. The methods complement one another with
NAA primarily used for trace and minor elements, XRF for the more
abundant elements and GRS for the radioactive elements. The
rather high precisions of these methods allow considerable cross-
checking of elemental abundances determined by the different
methods.
The analytical results indicate that: (1) H2S (determined by
XRF) and W (determined by NAA) both correlate with down hole
temperatures within a specific geothermal system; (2) the uranium
daughters, determined by GRS measurements of spring waters and
deposits, are most abundant in CaCO- dominated systems; (3)
uranium may be concentrating at depths within hydrothermal systems.
55
-------
RESULTS OF SAMPLING AND ANALYSIS OF HOT AND COLD SPRING WATERS
AND ASSOCIATED ROCKS AND SOILS FROM POTENTIAL
GEOTHERMAL RESOURCE AREAS IN NORTH CENTRAL NEVADA*
H, Bowman, H. Wollenberg, F. Asaro, and A, Hebert
INTRODUCTION
This work is concerned with the sampling, analytical procedures, and
analytical results from a study of four hot spring areas in the Battle
Mountain high heat-flow area of north-central Nevada, These areas
include Kyle Hot Springs (Buena Vista Valley), Leach Hot Springs (Grass
Valley), Buffalo Valley Hot Springs, and Beowawe Hot Springs (Whirlwind
Valley). This effort is part of a more general study at the Lawrence
Berkeley Laboratory and the University of California at Berkeley on
applications of geoscience techniques to geothermal resource area
assessment (Wollenberg, 1975).
We have collected and analyzed water samples from both hot and
cold springs in these areas as well as rock samples from associated
formations. Our sampling and analytical techniques have been described
in detail earlier (Bowman, 1975, Wollenberg, 1976) and will be discussed
here only briefly.
A number of water sampling methods were tried. The most repro-
ducible results were obtained by inserting a Tygon tube down into a
spring or spring orifice, drawing water up through a 0,45-micron filter
and into a half-liter nalgene bottle using a hand-operated vacuum
pump (Fig. 1). The half-liter water samples were evaporated in the
*Work was performed under the auspices of the U.S. ERDA.
56
-------
Ul
-J
Figure
-------
laboratory (at 80°C) in the original plastic collecting bottles. These
collection methods were devised to retain all solid materials, including
those which would normally precipitate on cooling.
At the sites, a drop of water from each spring was evaporated onto
a Lexan disk with a fixing solution. These samples were analyzed in
the laboratory using soft x-ray fluorescence (XRF). A silver disk
was placed in a separate sample from each spring for ^S determination
using the same XRF method (Hebert and Street, 1974).
Separate aliquots of water evaporates and crushed rocks were fused
into glass disks for XRF measurements. Samples for neutron activation
analysis (NAA) were prepared by mixing these materials with a cellu-
lose binder and pressing them into small pellets (Bowman, 1973).
RESULTS
Buffalo Valley Hot Springs
Water Analysis
Four separate hot pools were sampled at Buffalo Valley Hot
Springs, and the results are given in Table I. The last column in
Table I gives the average abundance for the four pools along with the
measured RMS (la) deviations. The average abundances for the elements
Na, Rb, and Cs have been underlined and the (la) deviations are 2.2%,
3.8%, and 3.2%, respectively.
58
-------
TABLE I
Buffalo Hot Springs
mg/liter
Temp°C
Si02
Na
Cl
K
Ca
15-6
72°
75
268
28
29
24
15-1
72°
64
269
28
27
25
15-4
65°
81
277
26
27
28
15-5
68°
84
280
25
36
20
Ave.
76±9
274±6
26.8±1.5
30±4
24±3
yg/liter
U
Ba
W
Br
Sb
Mo
Rb
Cs
As
Fe
Sc
Mn
<.08
160
28
62
37
4
124
150
<10
<100
<.02
30
<.16
150
30
65
57
<1
133
155
<10
<100
<.02
30
<.16
140
33
70
35
<1
130
160
-------
Rock Samples
A rhyolitic ash flow unit, the Fish Creek Mountain Tuff (McKee,
1970), was sampled and the results are shown in Table II. These tuffs
erupted from a center only about 10 miles south of Buffalo Valley Hot
springs, in the early Miocene. Although 75 cubic miles of ash-flow
material erupted, it did not spread more than about 10 miles from its
source. The unit was uplifted during Basin and Range faulting and, at
present, can be sampled from the bottom contact upward. The chemical
compositions of the ash layers are shown in Table II with the lower
ash flow sheets on the right. The errors shown in col. 1 are the
la values or the precision of measurement. Accuracies can be
obtained by referring to Bowman (1973).
Buffalo Valley Heat Flow Holes
Data from a heat flow hole (BV-2) cuttings, drilled just south
of Bufallo Valley Hot Springs and a holei drilled about 2 miles north-
west of the springs (BV-3) are shown in Fig. 2. The Si02 contents
for these holes are plotted as a function of hole depth. At the bottom
of this figure we have plotted the temperature profile of these wells.
The BV-2 data in Fig. 2 suggest that Fish Creek Mountain Tuff, or sedi-
ments from this ash, may extend down to 300 ft. at Buffalo Valley Hot
Springs. The data from BV-3 indicates the presence of much more basic
sediments that were derived primarily from the Tobin Range to the west.
60
-------
TABLE II
Comparison of Chemical Analysis of Fish Creek Mt. Rhyolitic Tuft Sequence
Si02
Ai203
I FeO
MgO
CaO
Na2°
K20
Ti02
MnO
Total
U
Th
Ta
Hf
La
Ce
Nd
Sm
Eu
Tb
Dy
Yb
Lu
Rb
Cs
W
Ba
Sb
Mo
Sc
Co
Zn
Cr
Ni
7
75.08
13,0
0.51±,01
0.03
0.80
3.38±;03
5,40
0,20
0,004
98.40
6,70±.05
30.5±,2
2.01±.01
5,6±.l
97±1
180±1
65±2
9.88±.02
0,35±.02
0.84±,04
S.l±.l
2.54±.04
0.30±.02
180±4
4.2±.l
2.9±.3
423±20
0.2±.l
<0.7
3,2±.l
0.21±.04
34±5
<9
<10
4
74.00
13,2
0,28
0,12
0.57
4.20
5.09
0,11
0.006
97,53
11.3
66.8
4,49
10.5
70
137
61
11.6
0.18
1.6
9,8
5.0
0,68
498
19
3.3
201
1.0
<.5
1.8
0.4
100
<10
<10
3
71.23
13,4
0,87
0.26
1,86
4.06
4.59
0,19
0-019
96.48
13.0
74.5
4.80
10.0
59
120
60
12.4
0.29
2.0
12.1
6.2
0.76
495
25
3.5
330
1,7
<.5
3.7
1.0
200
<10
<10
2
74:70
12,5
0.64
0,20
0.47
4.42
4.58
0.06
0.015
97.54
16.1
113.5
6.42
10.3
23
50
38
10.6
<.01
2.2
13.6
6,2
0.80
685
32
3,8
31
1.6
1.2
1.7
0.3
140
<10
<10
1
73.74
12.4
0.65
0.13
0.47
2.84
6.18
0.05
0.019
96,50
23.4
116.3
5.49
9.9
27
72
51
16,2
<,01
3.4
21.3
8.9
1.12
884
30
4.3
105
1.7
3.2
1.0
0.7
137
<10
<10
6A
1.34
2,60
6.0
0.11
0.024
20.8
84.2
5.58
10.1
26
66
54
17.4
0.09
3,1
21.8
9,0
1.08
476
20
4.6
<30
0.85
7.2
1.9
0.3
130
<7
<7
6B
.121
2.79
6.0
0.11
0.025
22,2
84.7
5,46
9.7
29
68
66
18,3
0.06
3.1
22.8
9.1
1.17
481
20
4.4
<30
0.75
7.7
1.4
0.2
115
<7
<7
6C
1.28
2,61
5.9
0.11
0.024
21.3
83.3
5.54
10.1
27
67
53
17.4
0.05
3.1
21.7
9.0
1.12
489
20
4.2
<30
1.06
5.6
1,8
0,3
80
<7
<7
61
-------
to
80
60
(
40
O
o
CD
l~
±3
O)
Q.
E
20
100 200 300
Depth (feet)
BV-2
BV-3
BV-3
BV-2
300h
Q.
O.
e
.E 200
tn
(U
O
100
0
\- /
0
200 400
Sodium (ppm)
600
Figure 2. Buffalo Valley heat flow,
holes 2 and 3
Figure 3
-------
The slight job in the temperature profile of BV-3 seems to be associated
with the Si02 high of the BV-3 data above,
Grass Valley (Leach Hot Springs area)
Water Analysis
Five separate pools in the Leach Hot Springs area were sampled
along with three nearby cold springs. The chemical compositions
of the hot waters at Leach Hot Springs varied considerably from
spring-to-spring. The cesium, rubidium and chloride content
of the hot spring waters are plotted in Fig. 3 against sodium concen-
trations, The chloride content correlates quite well with the sodium
content and a general trend can be seen in the Cs and Rb versus Na data.
The spring surface temperatures are inversely related to the Na and
Cl contents of these waters,
DISCUSSION
The overall cesium and sodium data for all hot springs are plotted
in Fig. 4. The Buffalo data (B) seems to line up with the Kyle data (K)
and most of the Leach data (L) . The maverick pool at Leach seems to be
off this line as are the two data points from Beowawe (W).
The data from Buffalo Valley seems to follow this general trend
and although the variations are small they may be real. The signifi-
cance of these correlations is not understood at present and is being
studied.
The overall tungsten data is shown in Fig. 5. The tungsten
abundances in hot and cold spring waters are plotted on the vertical
63
-------
100 -
f "50
0)
o
o
I '00
o
c.
0)
1/5
o>
50
50 100 150 200
100 150 200
Figure 4. Leach Hot Springs
<§>Bebwawe area
© Leach area
EKyle area
• Buffalo Valley
Cold and hot spring waters
Steam
**
100 200
Estimated underground water temperatures (°C)
Figure 5. Cold and hot spring waters.
64
300
-------
scale and calculated aquifer temperatures are shown along the horizontal
axis. Not much is known about the tungsten rock-water interactions,
but these data suggest that tungsten prefers the liquid phase at ele-
vated temperatures.
The H2S dissolved in these waters was tested by placing a silver
disk into bottles of spring water and then testing for sulfur later
by using XRF. The dissolved H2S seems to vary with the calculated
aquifer temperature (Fig. 6) suggesting that it may be a useful chemical
g e othermometer.
The uranium concentrations measured in both hot and cold springs
for this region are shown in Fig. 7. The hot spring areas are labeled
at the top of the figure. The bars labeled C are cold springs, and
those labeled H are hot springs. The higher uranium concentrations
are associated with the cold springs and not the hot. If we assume
that the cold ground waters in these areas are feeding the hot spring
aquifers at depth, then we must conclude from these results that the
uranium is accumulating at depths. The radon contents of the hot
water at Kyle and Buffalo Valley Hot Springs (Wollenberg, 1974) seem
to reinforce these conclusions. Data such as these can be used to make
estimates on the amount of uranium that might have accumulated at
depths as well as the ages of the hot spring areas.
65
-------
ion
1 V V
~ 10
Q.
Q.
CO
OJ
I
1 1.0
0
.5
13
cr
LJ
O.I
1 1 1 1 1
silver B
disk color *^ Boiling or bubbling springs
A Hot and cold springs
\° Calibration points
"
B
B *^B
-.„,... ^ __
:^A
A X *
_ Brown "N.
\
\
Golden \ A
\rp
1\ A
- Silver t\ ^ A A*
\\ A A
V
^ \
Calibration lineA
i i i i i-
d)
.is o
Q.O
£l
200 | S
a> ^_
o ~
E ^
ai 3
£ c
0 5
a; o
100 ^"^
0) O
«5
s
1
i_ O
<0 -£.
0 c-
o
U.
100 10 1.0
Observed relative sulfur Xray intensity
(6min ; 2.54cm disk; 275 ml water)
Figure 6.
66
-------
URANIUM (PPB)
Hot and Cold Springs
5h
4h
GO
Q.
Q.
c
o
—
r — Kvle — - 1 onrh
p i\yic l_CULn
C
m
c
P
_H
H
u
c
w,
I
c
%
c
H
H H T
T T 1
1
C
3
C
2
H
2k
iHr
Beowawe—-|— Buff a lo—H
H
Figure 7.
67
-------
References
Bowman, H.R., Asaro, F., and Perlman, I., 1973. On the uniformity
of composition in obsidians and evidence for magmatic mixing:
Jour. Geology, v.81, p.312.
Bowman, H., Hebert, A,, Wollenberg, H., and Asaro, F., 1975. Trace,
minor, and major elements in geothermal waters and associated
rock formationa (north-central Nevada); for Proceedings of Second
United Nations Geothermal Symposium, San Francisco.
Hebert, A.J., and Street, K., 1974., A nondispersive soft x-ray fluore-
scence spectrometer for quantitative analysis of the major elements
in rocks and minerals: Anal. Chem., v. 46, p. 203.
McKee, E.H., 1970, Fish Creek Mt. Tuff and volcanic center, Lander
County, Nevada, Geol. Survey Prof. Paper 681.
Wollenburg, H., 1974, Radioactivity of Nevada hot-spring systems: Law-
rence Radiation Laboratory Internal Report LBL-2482.
Wollenberg, H,, 1975, Geothermal energy resource assessment, UCID-3762.
Wollenburg, H., 1976, Proceedings of the U.S. EPA Workshop on Sampling
Geothermal Effluents, pg.143-164.
68
-------
ABSTRACT
METHODOLOGY FOR SAMPLING AND ANALYSIS
OF HIGH PRESSURE STEAM LINES
W. Hamersma
Systems Group of TRW Inc., Redondo Beach, California
This paper is based on work at TRW which developed methodology
for sampling and analysis of energy and synthetic fuel plants
for an environmental assessment.
An experimental approach and plan will be presented adapted from
existing technology for the sampling and analysis of high pres-
sure steam lines. The sampling apparatus will allow for the
collection of fluids and non-condensable gases at atmospheric
temperature. An analysis scheme for these fluids and gases will
be briefly described.
The sampling portion of this plan will be adpted from Chapter
V of "IERL Procedures Manual: Level 1 Environmental Assessment".
69
-------
ABSTRACT
HOW CAN STANDARD METHODS FOR SAMPLING AND ANALYSIS
SUPPORT THE MOTIVES BEHIND THE REGULATIONS?
Do E. Michels
EG&G Idaho, Inc..Idaho Falls, Idaho
The basic concern about geothermal effluents lies in their po-
tential detriment to the biological community. Thus, we need
good regulations for real protection. Of course, regulations
about the effluents would be foolish if they turned out to be
irrelevant to ecological health. Unfortunately, basing measures
of regulatory conformance on real biological effects risks demon-
strating facts too late.
Environmental sampling can give earlier warnings about ecological
harm, but relies on abstract models. On one hand these models re-
late environmental contaminations to biological effects. On the
other hand they require special concepts of sampling and data in-
terpretation to yield data usable for the models of bio-effects.
Basing regulations on these models, sampling concepts, and in-
terpretational methods leads to a judgemental jungle about
whether factors are over- or under-emphasized. This in turn can
lead to acrimonious debate over whether regulations about envi-
ronmental contaminations are either relevant or actually being
met.
The least ambiguous place to measure effluents is at the source
and there is some practicality in enforcing regulations on such
measures. However, it may be an indefinite path from an effluent
to a biologic effect. Thus, effluent regulations are at a great
risk of being inappropriate.
Effluent sampling, environmental monitoring, and bio-effect meas-
ures have distinct methods of collecting data and interpreting it.
Partly, these result from the different aspects of the media, but
also because the motives for taking samples are different. Par-
allel with these different motives, value judgements about merit
differ among workers. These differences can lead to conflict
about whether a (proposed) regulation is appropriate.
How might regulations be developed so as to be judged fair and
relevant to all three points of view? Could one point of view
be accepted as philosophically superior? Could standard methods
of sampling and analysis be chosen with respect to how the data
will be interpreted?
70
-------
HOW CAN STANDARD METHODS FOR SAMPLING AND
ANALYSIS SUPPORT THE MOTIVES BEHIND THE REGULATIONS?
Donald E. MicheIs
Idaho National Engineering Laboratory
INTRODUCTION:
Three kinds of questions relate to standard methods for sampling geo-
thermal fluids:
"What is it like?" (posed in the academic sense of inquiry
about the phenomena irrespective of social, engineering, or
health effects),
"What will it do to the hardware?" (an engineering question
concerned with features of design and economics),
"What is the environmental effect?" (a social and health
issue).
The requirements and techniques are different for each evaluation and,
to a possibly serious degree, workers asking one of the questions above may
not fully hear the other questions or their answers.
The drive to establish regulations and the need to support enforcement
with standard methods concern the last item. Curiously, only a few of the
methods presented at this meeting so far apply to measuring environmental
effects. Environmental measures deserve more attention, not only because
of health and environmental issues, but also because such monitoring will
be a significant overhead cost item on operational systems, irrespective
of how well the installations may be accepted by the public. It is impor-
tant that both the regulations and the sampling/analysis (R&S/A) be biolog-
ically astute. The intent of this presentation is to show the kinds of in-
formation needed. Additional work will be required to identify specific
R&S/A that are both effective and low-cost.
INPUTS:
The construction of regulations requires input of several kinds, as
diagrammed in Figure 1. We perceive effluent materials getting into the
environment to an extent that the amounts picked up by biota might have
detrimental effects. Thus, our ultimate concern is biological. The en-
vironmental concentrations, tissue concentrations, and bio-effects are
indirectly connected to effluent regulations. Data obtained from studies
71
-------
Postulates
Law, Public Opinion
Figure 1. Flow sheet for regulations
Currently
Eventually
Figure 2. Inputs to regulations
72
-------
of the effluent, environment, and bio-effects need to be incorporated into
models. Those models, combined with informed judgement (postulates), might
yield regulations that are fair and effective because of being biologically
astute -- if the several inputs are well-balanced.
Figure 2 shows two ways in which the inputs to regulations might be
balanced; the widths of the several arrows indicate the relative weight-
ing. The upper part shows my current perception of how the inputs have
been apportioned; the lower part is my suggestion for how they should be.
I won't insist that the widths of all the arrows are exactly right, but
there are three points worth making in this regard: 1) we do not now have
an optimum apportionment, 2) an optimum apportionment would still be dia-
grammed by arrows of unequal width, and 3) the optimum input would empha-
size models of environment and bio-effect.
These models deserve to be elaborated so as to show the kinds of
thought processes used to deal with the transfer of toxins between environ-
mental compartments, since some of those compartments are accessible for
monitoring.
ENVIRONMENTAL PATHWAYS
Effluents eventually get into man and other species through a some-
times intricate sequence of transfers that have come to be called environ-
mental pathways. Figure 3 shows a diagram of such a networkt1!. Each box
in the diagram represents an environmental compartment, each arrow a trans-
fer between compartments. Such a diagram can be quantified by sampling
which leads to estimates of effluent (concentrations, amounts, etc.) in
the separate compartments and possibly the rates of transfer. Some com-
partments are less difficult to sample than others so that an S/A program
might focus on them. Unfortunately, the easier sampling tends to be to-
ward the left side of the diagram while the items of interest, man and
what he ingests, are toward the right side.
It is sometimes a long environmental path from effluents to man so
that sampling effluents risks obtaining data that are ambiguous. The
precision gained in sampling the upstream compartments can be dissipated
by uncertainties in the transfer coefficients between the downstream com-
partments leading to man. The tradeoffs regarding which compartments to
sample require careful weighting when setting up regulations intended to
protect the biota, and man in particular.
Our attention to effluents and environmental compartments provides
only a means to an end. Figure 4 shows that the effluent materials may
get involved as nutrients, poisons, carcinogens, and mutagens. More
specifically, the effects may be identified with primary biologic func-
tions, cell development, or the nervous system. These latter aspects
comprise our real concern. The upstream compartments and transfers are,
73
-------
Figure 3. Pathways to man.
Pathways
Primary
Functions
Cell
Development
Nervous
System
Figure 4. Bio-effect models
74
-------
in this sense, incidental to these bio-effects. We study the upstream
factors to learn which pathways are amenable to control efforts that might
be effective in short-circuiting or avoiding dangerous materials We
should be careful not to lose sight of this aspect of our studies Even
though effluent measurement risks being irrelevant or excessively remote
from bio-effects, a thorough study of the bio-effects in real time risks
learning too late what we need to know. These are the most compelling
reasons to follow effluent sampling with pathway analysis.
TRANSFER COEFFICIENTS
An important concept for transfers through the environmental path-
ways is the discrimination factor (F)[21. Ideally, it is defined as the
ratio of concentrations (Cu, Cd) in compartments that are adjacent in the
upstream-downstream sense.
C.
F - / eq. 1
u
This equation has an appealing simplicity for it appears that the factor
F might be measured in one place and then applied elsewhere. In the new
place, the concentration in one compartment might be estimated by measur-
ing the concentration in an adjacent compartment and multiplying by the
imported discrimination factor (or its reciprocal). Unfortunately, this
approach, which is useful in liquid-liquid chemical systems, for example,
does not work well in the environmental context. There are at least three
reasons: 1) There is no micro-reversibility in the environmental systems.
For example, a cow eats grass and the gut exchanges nutrients and other
chemicals, but the exchange is only one way — none gets directly back to
the living grass; 2) The feedback which does occur* is cyclical, not back
and forth -- the cow's feces enter the soil compartment and its decompos-
ing organisms before part returns to the grass which is then eaten in a
subsequent cycle; 3) Within any one compartment the input-output involves
more than one neighboring compartment, in contradiction to the form of
equation 1.
QUANTITATIVE FEATURES
Still, we need to know the buildup of toxic materials in downstream
compartments and this involves some kind of data reduction. The buildup
would be well described by quantifying a selected small number of features.
1) a central value for concentration, roughly or perhaps exactly an aver-
age value, 2) a measure for the range of concentrations as they occur
throughout the small parts of the compartment irrespective of dispersion
patterns, 3) the full inventory of effluent material that resides in a
compartment, 4) estimates of precision vs. accuracy and 5) changes in the
above related to dynamics of the situation and passage of time. Elaboration
75
-------
of the above concepts would require much more space than is available here,
so I will deal with some of the important subtle factors.
AVERAGE CONCENTRATION
The abstract concept of average concentration applied to an environ-
mental compartment usually is difficult to match with reality and adapt to
the physical problems of obtaining representative samples, as the follow-
ing discussion will show. Definition of average concentration (Cavg) re-
lates to both the full inventory (I) of the component and the entire volume
(V) of the compartment.
Cavg ' V- eq' 2
The sampling problem arises because of real variations in concentration
that occur throughout the compartment. A single sample, however well-
analyzed, gives no clue in itself about how well its concentration matches
the average for the whole compartment. Several samples can show a pattern
of concentrations, and features of the pattern may be compelling for de-
cisions about how Cavg is to be derived from the several data.
The simplest approach is simply to average the analytical concentra-
tions and propose that it equals Cavg as in eq. 3.
Cavg = n" E Ci eq. 3
This method is commonly used and commonly is not valid. Most significantly,
it presumes that each Cj represents a uniform fraction of the environmental
compartment and that all fractions are represented fairly at the time of
sampling. Such a situation can be set up in sampling a water discharge
where a device extracts a pre-selected small volume of water at intervals
corresponding to the passage of a much larger increment of flow. This kind
of proportional sampling is more difficult in natural systems. Air samp-
ling, for example, appears to be proportional, but seldom is. Even at a
constant volume rate of filtering air the wind vectors around the sampler
vary with time so that faster moving air masses are undersampled compared
to their slower moving counterparts. This kind of problem might be accom-
modated mathematically by weighting factors (Wj). These represent fractions
of the compartment which are represented by a t^ as in equation 4.
Cavg =Etf. ?F * Vi eq. 4
This kind of weighting may be fair, but it does imply that one must acquire
some number to represent the volume (or mass) of the compartment. That
number would correspond to V in eq. 2.
76
-------
BACKGROUND
The notion of sampling contaminated compartments implies getting two
kinds of samples -- those dosed only with a pre-existing (background)
amount of the material of interest and those which are dosed with both
background and the effluent. The higher analytical values might be easily
recognized as contaminations, but there may be substantial trouble in
making a Convincing identification of low-level contamination in the face
of a variable background. Points at issue may involve either or both the
dimensions of the infected area and the concentration level which best
divides the contaminated group from the uncontaminated. The statistics
of mixed distributions apply here. The increment of contaminant inven-
tory that hangs in the balance during disputes over the true value of
background can be large since the volumes of involved soil, air, or water
are often huge.
STATISTICAL APPROACHES
Some statistical methods are useful in sorting heterogeneous distri-
butions into their homogeneous components[3]. The latter are character-
ized by specific central values* and measures of dispersion (standard
deviations). In applying these methods to environmental cases the stan-
dard deviation should not be identified as an error term, although of
course it does contain the errors. More importantly, it gives a measure
of real variations in the environmental concentration from place to place.
These natural variations tend to be very large in relation to the average
concentration, particularly for those materials which are especially scarce.
They often are large compared to analytical errors; hence the justification
for ignoring error at this stage of the interpretation.
For most trace materials, Gaussian statistics (the normal distribu-
tion) are a poor model[4]. Often, better results are obtained if the data
are transformed mathematically before matching to the Gaussian distribu-
tion. Among the several possible transforms the logarithmic has advantages
of being both simple and broadly useful. That is, the logarithms of many
data sets are distributed in (or near) a Gaussian way even though the un-
transformed data are not. Many standard statistical treatments can be
applied to the transformed data with no further concern for validity. After
the statistical treatments, the original units of the data can be restored
to the calculated measures by taking inverse logarithms.
Graphical procedures are simple to use in identifying distribution
5! and for deciding how to unmix distributions1-^. These work well
median, mode, arithmetic average, geometric average, and others are all
useful measures of central value, but statistical treatments and exten-
sions to inventory, flux, etc., demand a proper selection of central
value to yield valid results.
77
-------
for small sets of data (12 to 200 points), but become increasingly tedious
for larger sets. Computer programs which perform a digital version of the
graphical procedures can be prepared^.
CONTAMINANT FLUX
Assessment of inventories may be approached through fluxes by equating
inventory (I) to the algebraic sum of inputs (Qj) and outputs (Qg). The
inputs and outputs can be incremented as in equation 5.
Qj = E E CiMi eq. 5
a. L.
There, Mj represents the mass of the i increment of input. The factor
E represents an inefficiency for a transfer, i.e., its value is less than
unity. An example involving E is given by the feed eaten by a cow. Al-
though the entire feed mass is at some time contained within the cow, only
a portion becomes part of the cow's tissues. This kind of partitioning
applies also to the trace materials which are incorporated in feed as
effluent contamination. Thus, the input-output approach requires a back-
ground of special studies that give estimates for E-values. E-values are
less straightforward to measure than first appearances might suggest. Be-
sides the mechanical and analytical problems in setting up a balance sheet
and collecting data about feed-in and feed-out, one has to deal with inter-
actions. That is, a value of E for a particular effluent material involves
more subtle factors than chemical composition of the effluent and biologic
species. General dietary balance is known to affect uptake of trace ele-
ments which are imperfectly similar to essential elements^7].
Relating effluents to feed inputs of foraging animals involves still
another wrinkle. They are selective feeders -- when they can be selective.
Figure 5 shows the distribution of botanic families obtained from system-
atic sampling of the rumen contents of steers grazing freely in a botanic-
ally limited areatsl. Contaminants are dosed differently on the several
species involved, with the result that the ingestion rate for a contaminant
varies seasonally. Note in Figure 5 that the seasonal variation is imper-
fectly aligned with the calendar. Estimating inputs to foragers by sampling
feedstock species implies that the human samplers discriminate in much the
same way as do the foragers. This is to say that, since foragers, like
cattle, are non-random samplers, their human counterparts should be non-
random also. Mimicking the selectivity exercised by foragers requires a
great deal of observation. It is important to do because in some cases,
the bulk of an annual dose to an animal comes in just a few months from one
or two species of feedl9J. This kind of complication, once understood, and
quantified, might be used to streamline effluent impact monitoring. Time
and money spent sampling plant species, etc., could be apportioned according
to relevance for inputs rather than according to physical prominence.
78
-------
100
June
Figure 5. Botanical composition of rumen content from
fistulated steers grazing on area 13,
Nevada Test Site (from Ref 9 by permission)
79
-------
GEOSTATISTICS
There is one more feature of natural distributions of materials that
deserves mention because of both obscurity of the concept among workers
and the potential for enhancing cost-benefit relationships in environmen-
tal studies. Development of this point has occurred in the mining and
mineral prospecting industry under the name "geostatistics"[10J. In
natural distributions the concentration indicated by a sample taken from
one point of space is not entirely independent of the concentrations that
might be shown by other samples (that could be) taken nearby. This notion
is important to the idea of "representative sampling". Also, it is fun-
damentally different from most of classical statistics which have resulted
from defining (assuming) the samples to be independent. Samples taken
across space tend to show decreasing similarities to some reference sample.
In other terms, the contaminant levels and the factors which were involved
with the contamination are increasingly independent. This feature can be
handled with a term that is calculated almost identically to standard de-
viation, but defined so that its value may be correlated with a distance
dimension. A diagram for describing this analytical approach is shown in
Figure 6 A and B. The circles represent the pattern of sample points in
a stationary medium. Consider that pattern B is superimposed over pattern
A so that the sample points coincide except that the spacing for pattern
A is twice as great. Each interior sample point has four nearest neighbors.
These can be compared to the central point by their difference in concen-
tration. There is a characteristic difference which would be calculated
in a statistical style by squaring the individual differences, summing,
and averaging as in equation 6.
a2 = £E (XC-X.)2 eq. 6
In this application Xc is successively each interior point so that the en-
tire summation involves about 4n terms in (Xg-Xi). Thus, equation 6 could
be represented alternatively as in equation /.
n 4
c=l
E (X -X.) eq. 7
1=1 c 1
The same calculational procedure could be applied separately to pattern A
and B but the outcomes, a-values, will be systematically different. One
expects CJA > CTB and tne difference is related more to the systematically
closer spacing of sample points than to the edge difference of the patterns
plus random errors. The difference a^-ae and the different spacings are
potent descriptors of the dispersion and are called variograms. There are
several classes of variograms and some are represented in Figure 7; the con-
texts are described
The implications for economic sampling shall be described with the
proviso that sampling with separation in time for a flowing mass (like a
river or free air) can yield a variogram as does sampling across space
for stationary media. Economy is favored by longer separations between
80
-------
CO
OOoOOOoO
O OOO O O o
Boo
O O O o
o o o o
000 o-«-
o o
oo o o o o o
OOOOOOOQ o
Separation Increasing -
Most Common
Laminar Context
Nugget Effect
Classical Statistics
Figure 6. Geostatistics
Figure 7. Classes of variogr
ams
-------
samples; however, one makes the separations larger at the risk of having
poor represent!vity for material identified with the middle spaces between
samples. In terms of the variogram, the separation should correspond to
an interval taken rather far up on the sloped portion of the curve. Smaller
separations are somewhat redundant, larger separations are less represen-
tative.
SUMMARY
In summary, the driving force to establish effluent regulations comes
from both public concern and an obligation to minimize undesired environ-
mental effects. The standard methods for sampling and analysis are re-
quired for legal enforcement of the regulations. The sampling and analysis
of concern here relate less to the geothermal phenomena, and hardware
engineered to exploit them, than to environmental pathways and biologic
effects.
The direct sampling of effluents could greatly simplify the enforce-
ment objective. However, it seems reasonable to ask that the regulations
be astute in the biologic sense, since such is the focus of concern. Be-
tween effluent and biologic effect lie environmental pathways through which
the contaminants move. The astuteness of regulations can be judged by how
well knowledge about pathways is both integrated into regulations and adap-
ted to the environmental sampling -- while avoiding problems with inter-
pretational models.
The descriptions of pathways and their relative importance in terms
of effluent materials getting into the biota require quite extensive
specialized study. There seems to be no easy substitute; astute regu-
lations require input from pathway and bio-effect studies.
Dispersion patterns presumably will be investigated by sampling at
selected points in time and geography. Regulatory sampling for effluents
implies that methods will be used which sharply distinguish background
from only slight contaminations. Interpretation of the data in this re-
gard requires somewhat uncommon statistical methods. Additional inter-
pretive models are needed to track the effluents since processes which
disperse them are not truly random or uniformly potent.
Represent!'vity is a troublesome feature to demonstrate easily in
environmental sampling. One approach that can be convincing involves
identification of, in the distributions, of patterns that can be related
to mechanisms that drive the dispersion of effluents. Such models of
environmental dynamics and their counterparts in bio-effects deserve to
become the dominant technical input for the regulation of effluents.
82
-------
References
1. Corley, J. P., et al, Environmental Surveillance for Fuel Fabrication
Plants: US AEC Report BNWL-1723, 1973.
2. Denham, D. H.s et al, Radiological Evaluations for Advanced Waste
Management Studies: US AEC Report BNWL-1764, 1973.
3. Michels, D. E., Log-normal Analysis of Data for Plutonium in the
Outdoors: iji Proceedings of Environmental Plutonium Symposium, Los
Alamos, US AEC Report LA-4756, pp. 105-111, 1971.
4. Larsen, R. I., A Mathematical Model for Relating Air Quality Measure-
ments to Air Quality Standards: Env. Prot. Agency, Report AP-89, 1974.
5. Hahn, G. T. and Shapiro, S. S. Statistical Models in Engineering:
Wiley, 355 pp., 1967.
6. Michels, D. E. and Loser, R. E., unpublished report, (1973).
7. Hogue, D. E. et al, Comparative Utilization of Dietary Calcium and
Strontium-90 by Pigs and Sheep: I. Animal Science, V. 20, n. 3,
August 1961.
8. Smith, D. D., Status Report on Grazing Studies on a Plutonium-contami-
nated Range of the Nevada Test Site: in^ Studies of Environmental
Plutonium and Other Transuranics in Desert Ecosystems, USERDA Report
NVO-159, pp. 41-49, 1975.
9. Smith, D. D. personal communication.
10. Matheron, G., Principles of Geostatisties: Econ. Geol. V. 58, pp. 1246-
1266, 1963.
11. Davis, J. C., Statistics and Data Analysis in Geology:-John Wiley &
Sons, New York, pp. 374-411, 1973.
83
-------
A REVIEW OF THE CHEMICAL COMPOSITION OF GEOTHERMAL EFFLUENTS
F. Tsai, S. Juprasert and S. K. Sanyal
Geonomics, Inc., Berkeley, California
INTRODUCTION
This report presents a literature survey on the chemical
composition of geothermal effluents. We have tried to cover
as wide a geographical area as possible in the hope of getting
a general idea of the concentration ranges of various constit-
uents. The geothermal areas considered in this report are from
New Zealand, Japan, Italy, Iceland, USSR, Central America,
Mexico and the United States. The data were obtained mainly
from published sources. We realize that it is difficult to make
a meaningful comparison of the chemistry of the geothermal ef-
fluents from various parts of the world, because on one hand
the surface and subsurface environments may vary drastically
from place to place, and on the other hand the techniques of
sampling and analysis of effluents may be different in differ-
ent reported cases. It is conceivable that some of the data
reported in the literature are in error.
84
-------
REVIEW
Tables 1 and 2 show all the chemical constituents reported
to have been found in geothermal effluents with their minimum
and maximum concentrations and some of their significant en-
vironmental impacts. It can be seen that, in general, the
chemical composition of geothermal effluents covers a very wide
range, in terms of both the number of chemical elements or
groups present and their concentrations. For example, the total
dissolved solids (TDS) range from 50 to 400,000 ppm and pH from
1 to 10 units. From the environmental impact point of view the
geothermal waters can vary in character from entirely benign
and potable to highly corrosive and saline.
The constituents can be divided into 4 groups according to
their relative abundance (see Table 3). The "major constituents'
are the most concentrated and commonly found in geothermal
systems and play the most important role in chemical reactions
occurring in the system. The "secondary" and "minor" species
also participate in the chemical reactions but with smaller
roles. Trace elements contribute very little to the chemical
reactions in the system but may have considerable implication
in environmental impact. For example, trace elements such as
Ni, Zn, As, Rb, Sr, Ba may be harmful to plant and animal life;
most of the heavy metals are involved in formation of scales in
pipes.
85
-------
TABLE 1
CHEMICAL COMPOSITION OF GEOTHERMAL WATERS
Constituent
Aluminum ( Al )
Ammonium (NH^)
Arsenic (As)
Barium (Ba)
Boron (B)
CHB02)
Bromide (Br)
Cadmium (Cd)
Calcium (Ca)
Carbon Dioxide (C02)
CHC0)
(HC03
(C02 + HC03
Cesium (Cs)
Chloride (Cl)
Cobalt (Co)
Copper (Cu)
Fluoride (F)
Germanium (Ge)
(C03)
+ C03)
+ C03)
Concentration in ppm
0 - 7,140
0 - 1,400
0-12
Q - 250
0 - 1,200
13.6 - 4,800
0.1 - 3,030
0-1
0 - 62,900
0 - 490
0 - 10,150
0 - 1,653
20 - 1,000
15 - 7,100
0.002 - 22
0 - 241,000
0.014 - 0.018
0-10
0-35
0.037 - 0.068
Comments
Health hazard
Human death if
>550mg dosage
Detelerious to plants
Toxic to fish if
>0.2ppm
Clogging scale
Clogging scale
pH control
Hydrogen Sulfide (H2S, 0.2 - 74
total)
Major corrosion
constituent
Toxic to life if
large amount
Health hazard if >lpptn
Healthy if <1.5ppm
pH control, corrosion-
scale agent
86
-------
Chemical Composition of Geothermal Waters, cont'd.
Constituent
Iodide (I)
Iron (Fe)
Lanthanum (La)
Lead (Pb)
Lithium (Li)
Magnesium (Mg)
Manganese (Mh)
Mercury (Hg)
Molybdenum (Mo)
Nickel (Ni)
Nitrate (N03)
Nitrite (N02)
Oxygen (02, dissolved)
Phosphate (PO^)
Potassium (K)
Rubidium (Rb)
Silica (Si02, total)
Silver (Ag)
Sodium (Na)
Strontium (Sr)
Sulfate (SO,,)
Concentration in ppm
0 - 105
0 - 4,200
20
0 - 200
0 - 300
0 - 39,200
0 - 2,000
0-10
0.029- 0.074
0.005 - 2
0-35
0-1
0-10
0 - 0.3
0.75 - 2.05
0.02 - 0.22
0.6 - 29,900
0 - 169
3 - 1,441
0-2
2 - 79,800
0.133 - 2,000
0 - 84,000
Comment s
May precipitate on
oxidation
Cumulative poison
Clogging scale
May precipitate on
oxidation
Drinking standard,
45 ppm
Organic pollutant
Corrosion-related
Eutrophication agent
Scale-corrosion
accelerator
Major scale constitu-
ent, corrosion inhibi-
Mandatory tor
limit; 0.05 ppm
Scale-corrosion
accelerator
Clogging scale
87
-------
Chemical Composition of Geothermal Waters, cont'd.
Constituent Concentration in ppm Comments
Sulfur (S) 0 - 30
Total Dissolved Salts 47 - 387,500
Zinc (Zn) 0.004 - 970 Toxic to fish if
> 0.3 ppm
Zirconium (Zr) 24
The following are trace elements found at Sinclair #4 well,
Salton Sea, California (Lit. 27):
Antimony (Sb), Beryllium (Be), Bismuth (Bi), Cerium (Ce),
Dysprosium (Dy), Erbium (Er), Europium (Eu), Gadolinium (Gd),
Gallium (Ga), Gold (Au), Hafnium (Hf), Holmium (Ho), Indium (In),
Iridium (Ir), Liutetium (Lu), Neodymium (Nd), Niobium (Nb),
Osmium (Os), Palladium (Pd) .Platinum (Pt), Praseodymium (Pr),
Rhenium (Re), Rhodium (Rh), Ruthenium (Ru), Samarium (Sm),
Scandium (Sc), Selenium (Se), Tantalum (Ta), Tellurium (Te),
Terbium (Tb), Thallium (Tl), Thorium (Th), Thulium (Tm),
Titanium (Ti), Tungsten (W), Uranium (U), Vanadium (V),
Ytterbium (Yb), Yttrium (Y).
88
-------
TABLE 2
GAS COMPOSITION OF GEOTHERMAL VAPORS
Constituent
Ammonia (NH3)
Argon (Ar)
Arsenic (As)
Boric Acid (H3B03)
Carbon Dioxide (C02)
Carbon Monoxide (CO)
Helium (He)
Hydrocarbon C2 and
greater
Hydrogen (H2)
Hydrogen Fluoride (HF)
Hydrogen Sulfide (H2S)
(H2 + H2S)
Mercury (Hg)
Methane (CH,,)
Nitrogen (N2)
(N2 + Ar)
Oxygen (02)
Sulfide Oxide (S02)
Concentration in volume percent
0 - 5.36%
0 - 6.3
0.002 - 0.05
0 - 0.45
0-99
0-3
0 - 0.3
0 - 18.3
0-39
0.00002
0-42
0.2 - 6
0.007 - 40.7 (ppb)
0 - 99.8
0 - 97.1
0.6 - 96.2
0-64
0-31
Remarks
Noxious gas
Minor inert gas
Scale formation
Health hazard
Noxious gas,
environmental
Hazard, corrosion
agent
Health hazard
Major inert gas
Corrosion agent
89
-------
TABLE 3
RELATIVE ABUNDANCE OF CHEMICAL COMPOSITION IN GEOTHERMAL WATERS
Major Constituents
(maximum>10,OOOppm)
Chloride
Sulfate
Sodium
Calcium
Magnesium
Potassium
Bicarbonate
Secondary Constituents
(maximum>1,OOOppm)
Aluminum
Iron
Bromide
Manganese
Strontium
Carbonate
Silica (total)
Ammonium
Boron
Minor Constituents
(maximum>Ippm)
Arsenic
Barium
Cadmium
Cesium
Copper
Fluoride
Hydrogen Suflide(total)
Iodide
Lanthanum
Lead
Lithium
Mercury
Nickel
Nitrate
Phosphate(total)
Rubidium
Silver
Zinc
Zirconium
Trace Constituents
(Renerally<0.01ppm)
Antimony Platinum
Beryllium Praseodymium
Bismuth Rhenium
Cerium Rhodium
Dysprosium Ruthenium
Erbium Samarium
Europium Scandium
Gadolinium Selenium
Gallium Tantalum
Germanium Tellurium
Gold Terbium
Hafnium Thallium
Holmium Thorium
Indium Thulium
Iridium Titanium
Liutetium Tungsten
Molybdenum Uranium
Neodymium Vanadium
Niobium Ytterbium
Osmium Yttrium
Palladium
-------
With reliable data, chemical composition of geothermal ef-
fluents can provide much useful information about the reservoir
since the kinds and amounts of constituents depend on the reser-
voir environment: formation lithology, rock-water interaction,
rock mineral chemical equilibria as well as pressure and temper-
ature. The major environmental concerns in a geothermal system
are the presence of noxious constituents in the effluent, cor-
rosion and scale-formation in pipes. Corrosion in casing and
surface plumbing may cause contamination of the environment by
geothermal fluid; scale formation may make disposal of spent
geothermal effluent difficult,, and creates a solid waste dis-
posal .problem. Chloride, oxygen, pH and sulfide are the major
constituents in geothermal effluent responsible for corrosion
and calcite scale formation. Silica, sulfide and hydroxide
are the other causes of scaling. Corrosion and scaling being
chemical processes, not only the chemical composition but also
other thermodynamic factors such as temperature and pressure
will play a role in determining the nature and extent of cor-
rosion and scaling.
It is of interest to compare the chemistry of geothermal
waters with other types of waters. Figure 1 compares the ranges
of major chemical constituents in geothermal water and in potable
water. In general, the reported maximum concentrations of dis-
solved constituents in geothermal water exceed those in potable
water. The situation is similar when geothermal water is com-
pared with drinking, irrigating, feeding or sea water (see Tables
4 and 5).
91
-------
K)
10
-1
10'
10'
10 *
10'
AI3V
-c:
B -C
Br -
Ca2X
HC03-<
o
F-«
* I ~
Fe2+
*Pb2+
Mn2+<
NOs~-C
K + -C
Si02-(
Sr2+
*Zn2+
TDS-(
i i > iniij—i i i mii[ i i i IIMI|—i i i niiij—i i 11 MII|—i i i iniif
TTTI—i i i mi
Li-crrz
"H
i ml i i 1 1 mi I i i 1 1 1 ml i i 1 1 1 ml i i 1 1 1 ml
1 mil i i i 1 1 ml i
10'
10~1 1 101 102 103 104 10 5 1Q6 1Q7
(parts per million)
CONCENTRATION RANGES in —^^ Geothermal Waters
mmimmum Potable Waters
* Potable Water Values not available
Figure 1. Chemical concentration ranges in geothermal and,
potable waters
-------
TABLE 4
COMPARISON OF DRINKING, IRRIGATING, FEEDING AND GEOTHERMAL WATERS
OJ
Substance
Drinking Water (ppm)
Recommended Mandatory
Irrigating Water
Threshold
Feeding Water(ppm)
Threshold Limiting
Geothermal Watex
(ppm)
Ranee
Arsenic
Barium
Bicarbonate
Boron
Cadmium
Calcium
Chloride
Copper
Fluoride
Iron
Lead
Magnesium
Manganese
Nitrate
Selenium
Silver
Sodium
Sulfate
Zinc
TDS
0.01 0.05
1.0
-
-
0.01
-
250
1.0
1.7 2.2
0.3
0.05
-
0.05
45
0.01
0.05
-
250
5
500
1.0 5.0 1
-
500 500
0.5 2 -
5
500 1000
100 350 1500 3000
0.1 1.0
1 6
_
- -
250 500
- -
200 400
_
_
1000 2000
200 1000 500 1000
_
500 1500 2500 5000
0- 12
0- 250
0-10,000
0- 1200
0- 1
0-63,000
0-240,000
0- 10
0- 35
0- 4200
0- 200
0-39,000
0- 2000
0- 35
trace
0- 2
0-80,000
0-84,000
0- 970
0-390,000
-------
TABLE 5
COMPARISON OF SEA WATER WITH GEOTHERMAL WATER
Typical Sea Geothermal Water
Substance Water (ppm) Range (ppm)
Ammonium NH4
Arsenic As
Barium Ba
Boron B
Bromide Br
Calcium Ca
Chloride Cl
Carbonate C03
Fluoride F
Hydrogen Sulfide H2S
Iron Fe
Magnesium Mg
Nitrate N03
Potassium K
Silica Si02
Strontium Sr
Sodium Na
Sulfate S04
TDS
0.05
0.02
0.05
4.6
65
400
19,000
140
1.4
60
0.02
1,270
0.7
380
7
13
10,600
2,650
34,500
0
0
0
0
0.
0
0
0
0
0.
0
0
0
0.
3
0.
2
0
0
1 ,
1,
1 — 3,
62,
241,
1
-1- >
2
4,
39,
6 — 29,
1,
133 - 2,
79,
84,
390,
400
12
250
200
080
900
000
653
35
74
200
200
35
900
441
000
800
000
000
94
-------
SELECTED LITERATURE
1. Axtmann, R.C., Science 187 795 (1975)
2. Coplen, T.B.,'and Kolesar, P., Univ. Cal. Riverside-74-18 (1974)
3. Coplen, T.B., Univ. Cal. Riverside-76-1 (1976)
4. Cusicanqui, H. et al, 2nd UN Symp. Dev. Use Geothermal
Resources 703 (1975)
5. Douglas, J.G. et al, Battelle Northwest Lab - 2094 (1976)
6. Ellis, A.J., UN Conf. New Sources Energy, 208 (1961)
7. Ellis, A.J., Bull. Vole., 2£ 575 (1966)
8. Ellis, A.J., Geothermics - special issue, 2_ 516 (1970)
9. Gupta, M.L., et al, Proc. 2nd UN Symp. Dev. Use Geothermal
Resources, 741 (1975).
10. Hoffman, M.R., EQL Memorandum no. 14 (1975)
11. Koga, A., Geothermics - special issue, 2 1422 (1970)
12. Kuwada, J.T., AIChE Symp. Series 70_ 772 (1974)
13. Mahon, W.A., Geothermics - special issue, 2 1310 (1970)
14. Mercado, G.S., Proc. 2nd UN Symp. Dev. Use Geothermal Re-
sources 1394 (1975)
15. Molina, B.R., and Banwell, C.J., Geothermics - special
issue, 2_ 1377 (1970)
16. Noguchi, K. , et al, Geothermics - special issue, 2_ 561 (1970)
17. Reed, M.J., and Campbell, G.E., Proc. 2nd UN Symp. Dev. Use
Geothermal Resources 1399 (1975)
18. Rowe, J.J. et al, U.S. Geol Survey Bull. no. 1303 (1973)
19. Siegel, S.M., and Siegel, B.Z., Environ. Sci. Tech. 9 473
(1975).
95
-------
20. Sigvaldason, G.E., and Cuellar, G., Geothermics - special
issue, 2_ 1392 (1970)
21. Sigvaldason, G.E., Unesco Geothermal Energy, 49 (.19.73)
22. White, D.E., et al, U.S. Geol. Survey Prof. Paper 440-F (1963)
23. White, D.E., Am. Assoc. Petrol. Geol. Memoir 4 342 (1965)
24. White, D.E. et al, Econ. Geol., 66 75 (1971)
25. Yanagase, T. et al, Geothermics - special issue, 2_ 1619 (1970)
26. Zinder, S. and Brock, T.D., Geochim Cosmochim Acta 41 73
(1977). ~
27. Hill, J.H., and Morris, C.J., Proc. 1st Workshop on Sampling
Geothermal Effluents (EPA) 1975.
96
-------
ABSTRACT
DEVELOPMENT OF GEOTHERMAL GAS SAMPLING EQUIPMENT
F. B. Tonani
GeoChemex, Berkeley, California
This paper summarizes the experience of the author in developing
equipment for geothermal gas sampling under a wide range of con-
ditions. Considered conditions have been: weight; ruggedness
for operation in tough access sites by one or two man parties;
capability to sample tiny seepages of gas from loose ground, scoriae,
fissures, hand-drilled holes in fumarole areas; and independence
from assumptions regarding what happens to a gas sample on storage
(for example, no "air corrected" analyses).
These requirements have led to the development of small sampling
tubes pressurized with gas carrier (the same to be used on GC
analysis), then pressurized with the sample gas. Tubes are
rugged and mailable to the laboratory even from far away countries
in a matter of days. The accompanying gear is devised to allow
careful aspiration of a small amount of sample that will not
disrupt the naturally occurring stream. The apparatus allows
not an abrupt sucking-in of uncontrolled amounts of gases, just
what the specific hole, or fissure can supply without trivial con-
tamination by the surrounding air. Glass syringes with Luer Lock
have been the final choice, in spite of their fragility. But,
plastic syringes are handy whenever usable. Standard items for
GC have been tried and, in some cases, adjusted for custom use.
Large volumes of sample can be safely handled in sampling tubes
equipped with new Teflon-stem valves, to which neoprene 0-rings
have been added for better security on long term transport and
storage.
Successful use of the Draeger detector tubes to analyze geothermal
and volcanic gases in the field has been made possible by accurate
study of their behavior. Questions, such as which components
work under certain conditions, how to proceed under conditions
entirely different from the ones envisioned by the manufacturer,
etc. have been answered after years of only partially successful
field missions. Water, carbon dioxide, hydrogen sulfide and
H2S+S02 tubes have proved to be dependable. For more precision
at high concentration of carbon dioxide, pocket sized absorbing
devices have been built. Commercially available materials have
often needed to be complemented by specially-designed ones.
97
-------
INTRODUCTION
The present paper summarizes the concepts that stem from
the author's own experience in developing equipment for sam-
pling geothermal gases. That experience has been mainly related
to large scale geothermal exploration and to volcanic surveil-
lance projects. Many aspects of it are relevant to the protec-
tion of environment and to the preparation of regulations, name-
ly the assessment of natural discharges and the prediction of
reservoir chemistry, i.e., of expected effluents on exploitation,
at the very early stages in the development of the resource.
In fact, geothermal effluents are special in this respect.
Many other industrial effluents are plainly man made. Others,
e.g.,tailings and effluents from mining operations, are not
strictly man made, however, the rate at which they are released
to the surface environment is much faster than under natural
conditions.
One cannot be as positive about the fact that geothermal
effluents are, in the essential, a net addition to the environ-
ment. Under certain plausible models of a geothermal field, a
geothermal power plant might be just diverting essentially the
same flow that takes place under natural conditions, rather than
generating new discharge, and it is fair to say that actual con-
ditions are likely to be intermediate. It is a matter of fact
that thermal manifestations disappear as a consequence of har-
nessing geothermal energy in the same general area, and we know
that many other unapparent manifestations were probably there
and died off as well. Furthermore, the commonly held view that
geothermal production is just tapping from a static storage of
energy and carrier fluid is very questionable. Many volcanic
and geothermal "systems11 discharge to the surface energy and
chemicals at a rate that compares well with what takes place
during full-scale production of geothermal energy.
Therefore, the rate and distribution of flow of energy and
chemicals under natural conditions as well as in production must
be evaluated in order to assess the impact of geothermal develop-
ment on the environment. Even in cases where the natural out-
put does not greatly affect the final results, predicting reser-
voir chemistry and therefore the expected effluents is of much
use in making decisions as to whether or not development is ad-
visable and how development should be pursued, before sub-
stantial investments are made.
It is worth noting that the overall geochemical surveys ,
necessary to achieve the goals outlined here above,include first
98
-------
of all much water sampling and analysis. The reason for bringing
gas sampling alone to the foreground in the present paper is
that the state of the art in the latter domain seems far less
satisfactory than in the former one. Actually, the gas sam-
pling and analysis procedures that are mentioned here in connec-
tion with the described equipment are closely related to water
study. For example, part of the interest in small samples stems
from the significance of dissolved gas studies in the overall
picture.
CONSTRAINTS ON THE DESIGN OF GEOCHEMICAL EQUIPMENT
Constraints on design of geochemical eciuipment fall in two
broad categories, analysis specifications and constraints related
to the working environment. Analysis specs include precision
(also termed repeatability), accuracy, dependability and cost
in as much as it is related to these specs. Constraints related
to the working environment include other sources of cost, e.g.
allowed capital investment, ruggedness, and more generally, the
usability under actual field conditions.
There is great variety in specifications and other con-
straints. Regarding geothermal effluents, for example, deter-
mining whether discharge from geothermal power plants meets re-
gulations requires that the analytical procedure be strictly
repeatable in the first place, in order to provide certitude in
judgment. That two observers come up with comparable results
is more important, from the standpoint of implementing regula-
tions, than having the obtained figures actually represent what
they are said to represent chemically. The legally recognized
procedure itself constitutes sort of an operational definition
of the measured quantity.
This is not to say that we are not interested in actual
composition, however, there is little doubt that this last know-
ledge does not need to be as uniaue and formally precise as the
knowledge of whether or not a specific effluent meets any spe-
cific regulation. Knowing where, when, to what extent, and un-
der which form some given chemical will interact with man is
actually more important than just knowing its concentration in
effluents. In other words, exact quantification of one figure
at one specific point is usually less meaningful than some broad
prediction of what the figure will be at a number of places and
under various circumstances. The same consideration applies to
predicting reservoir chemistry of the substances of interest
prior to development, which is the main subject of the present
expose'.
99
-------
In all these cases, accuracy is definitely more relevant
than sheer repeatability, behavior of the technique with dif-
ferent matrixes as opposed to behavior with one specific ef-
fluent is of concern, a capability to obtain more data at the
expenses of precision and accuracy may be worthwhile, in as
much as it upgrades the overall information obtained, and so
forth. Therefore, the required analytical specifications may
be quite different in connection with studying natural "systems"
from what they are in monitoring industrial effluents of one
specific nature.
Also, the constraints imposed by the working environment
during early stages of exploration and development may be quite
different from constraints effective after development of the
resource. To begin with, in the former case economic constraints
correspond to the high risk typical for early stages of develop-
ment, as opposed to the economics involved in developing an es-
tablished resource. Secondly, exploration is likely to take
place far away from laboratory and other facilities, and quite
possibly in rough country. By its very nature, exploration pro-
jects typically do not allow for long-lasting studies, and as a
result, there is little question of time-series studies and of
using permanent or semi-permanent field installations.
Of course, with progress, development conditions become
closer and closer to ordinary conditions in an industrial envi-
ronment. Here is a rough list of factors that have actually
played a part in developing the materials described in this
paper. The equipment ought to be operated by one man only, al-
though usually with one helper. It should be possible to operate
under various conditions, wind, dust and possibly rain. At the
same time, basically the same equipment should cover the needs
of different projects presenting very different field conditions.
In fact, even in the same project diverse occurrences of gas have
to be studied, which are usually hardly known at the onset.
Thus, a compromise must be reached between conflicting re-
quirements, the equipment has to be flexible in its application,
ensure reasonable accuracy, and remain small-sized, light and
rugged as far as possible. Basically, it appears that the pre-
sently described equipment as well as some of the equipment re-
ported in the literature has been designed to approach the charac-
teristics outlined above at the expense of making it fool-proof
and good for use by unskilled personnel. This is not a real loss
in the case of early studies, because a great deal of decision
making is left to field personnel, with regard to field analysis
and the detailed implementation of sampling programs so that
skilled personnel are needed anyhow.
100
-------
GAS SAMPLE CONTAINERS
DESCRIPTIONS
Two types of gas sample containers are currently utilized
with capacity of 15 ml and 50 to 100 ml respectively. The for-
mer type (Figure 1) has two silicone rubber stoppers, is U-shaped,
and includes a metal strap, spring and frame assembly to keep the
stoppers in proper place. The second type is a regular gas sam-
pling tube except for the valves, as 'torion' valves manufactured
by Sovirel are utilized instead of whole-glass stopcocks. (Figure 2)
The latter type looks very similar to that recommended by Giggen-
bach.1
Both tubes could be evacuated for gas collection, but prac-
tical use in exploration seems more dependable if they are pres-
surized above normal atmospheric pressure.2 The former type is
especially rugged and mailable, that is, it can reach the labora-
tory in shorter time than larger and less rugged items. Both
types are safer than most other types vis-a-vis accidental leaks
through valves.
DISCUSSION
Storage time is an important factor because the longer the
storage time, the more probable, i.e., statistically more fre-
quent become failures for a given container. There is no 'best'
container in the absolute sense, as odds for failure depend on
the way variable external action affects the container. More-
over, frequency of failures is only one factor, which must be
weighed against other factors such as cost, handiness in field
use and in transportation, and so forth. Neither the weight to
be attributed to each factor nor the associated odds for failure
with each factor are uniquely determined in the actuality.
The sample containers describe-;, in Figure 1 and 2 appear
to be suitable in the environment prevailing in exploration
projects and in the associated early evaluation of pollutants.
Giggenbach, W.F., 1976, A Simple Method for the Collection
and Analysis of Volcanic Gas SamplesTBull.Volean. Vol.39.
Elskens, I.; Tazieff, H., Tonani, F., 1964. A new method for
volcanic gas analyses in the field. Bull. Volcan.,Vol.27,
__-
101
-------
s\\\
^
\\SV"
ll 1
\W
~^j
\W
*«~_
\
—- -•
•
Figure 1.
102
-------
Figure 2.
GAS
Figure 3.
103
-------
In many cases one will have to collect more than one sample for
different determinations if a broad range of determinations is
required. This is just to cope with, (a) difficulties in col-
lection, (b) behavior in storage, and more generally (c) a de-
liberate overall policy to operate with smaller aliquots
rather than one sample. This will be discussed here below.
Let us consider some circumstances that point to the need
for collecting several aliquots.
1. Trivial contamination with air on sampling cannot be
afforded in cases, where knowledge of one or more of the
atmospheric component gases is or may be significant. The
required precautions on collection are usually such that
collecting small samples is much easier and in certain
cases the only feasible procedure. Trace amounts of non
atmospheric gases are not affected by contamination by air,
however, a larger volume of sample may be necessary. As a
consequence, both one small and one large sample are re-
quired in such cases.
2. Pyrex containers are not safe for samples where helium
must be determined. These samples are better stored in
metal containers, and the converse applies for hydrogen
sulfide. If the ratio between such sulfur gases as hy-
drogen sulfide and sulfur dioxide must be evaluated, or
if hydrogen sulfide must be determined in samples contain-
ing oxygen, separation on sampling is the only safe pro-
cedure. In addition to classical but cumbersome proce-
dures such as 'fixing' hydrogen sulfide in one sample,
and determining total sulfur in the other, separation on
molecular sieve 5 A proved effective for long term storage
prior to elution and determination (Elskens et al., 1964).
For such techniques which require packing the container
with some material, sample containers of the sort described
in Figure 1 are more convenient.
3. In a general way, using more aliquots has a number of
advantages in respect to factors which affect samples
erratically. Loss or breakage in transportation, mis-
labeling, heavy leak-in of air constitute one type of
erratic event that we may term 'failures'. On the other
hand trivial contamination in sampling or a basically
undetectable leak are the other type of erratic event
that we may term 'sampling errors''. Duplicate sampling
is an effective protection vis-a-vis erratic events pro-
vided the probability of occurrence of such an event is
not too great. One may require different levels of safety
depending on project features, but multiple sampling is
104
-------
cost-effective, provided the expected cost of failure ex-
ceeds the cost of additional sampling tubes and time.
Typically for the environment of exploration, i.e., for
early assessment of new areas, double-sampling is expected
to result in 2-10% increase in sampling cost. For the ex-
pected saving on failures to exceed the cost of double-
sampling the following conditions must be fulfilled;
P(l-P) greater or equal 0.1-0.02
respectively, where P is the probability of failures.
Double-sampling is seen to be cost-effective if P ex-
ceeds 10-27o respectively. Similarly, 20-470 failure
would cost-justify triple sampling.
It is worth drawing attention to the fact that the
reasoning above holds for a given equipment, that is, for a
given P value, whereas it is of no help by itself in selecting
the most appropriate container. However, containers for small
smples and an adequate sampling outfit reduce the average cost of
multiple sampling and are preferable in this respect. Basically,
the overall cost of extensive sampling in new and/or remote areas
is comparatively large, and both costly equipment of the type
described here and multiple sampling are cost-justifiable in
projects dealing with the assessment of the natural discharge.
FIELD GAS ANALYSIS
INTRODUCTION
Gas analysis in the field serves four main purposes:
1. It is possible to make relatively extensive gas surveys,
that offer a picture of the overall pattern in scattered
gas flow. In addition to directly supplying data, on-
line field analysis allows for adjustments in the sam-
pling pattern that result in better evaluation of the gas
stream.
2. Time-series of analyses are more easily obtained by field
analysis than by repeated sampling. Investigation of de-
pendence on time is part of overall flow estimates, but
also supplies information on different hydrodynamic com-
ponents of the same stream. Figure 3 points to a typical
situation where two different gas streams undergo mixing.
The mixing ratio is affected by minute changes in external
conditions. Time-series over a short time span are apt to
105
-------
detect the corresponding changes in composition; they^also
reveal the existence of and assess the two hydrodynamic
components.
3. Field determination offers a safe way around problems aris-
ing from sample storage. Sulfur gases are specially re-
active and may be heavily affected by storage prior to anal-
ysis. Field determination may be preferable to 'fixing'
hydrogen sulfide.
4. One problem with gas sampling is that you do not really
see what you are handling, it is as if you were blind-
folded. There are several ways of monitoring gas sam-
pling by on-line analysis; this substantially improves
the average quality of the samples obtained.
A great number of apparatus is available nowadays for field
work in the broad sense, including gas-chromatographs, as well
as infra-red, electrochemical and other analyzers. In connec-
tion with the early prediction of possible pollutants and the
assessment of natural discharge prior to the development of geo-
thermal resources, however, hydrogen sulfide and volatiles such
as ammonia, mercury and arsenic are the major concern. Under
the conditions prevailing before industrial development, pre-
liminary evaluation and prediction of the discharge of such chem-
icals depends to a sizeable extent on water analyses. Surveying
the mercury content in air might deserve consideration, however,
it is not discussed here, as mercury determination in soil is
likely to be a valid substitute for it.
The present expose' shall therefore concentrate on equip-
ment related to the assessment of the major constituents of the
gaseous discharge and hydrogen sulfide. However, the design of
the equipment resulted also from concern with constituents sig-
nificant in exploration, such as methane, hydrogen,etc. More
specifically, some pocket-sized items portable by an individual
even in rough country will be described. Such equipment is
viewed essentially as an aid in sampling.
DETERMINATION OF CARBON DIOXIDE BY ABSORPTION ON NaOH
High concentrations of carbon dioxide occur frequently in
geothermal manifestations and reservoirs. Direct measuring of
in excess of 90% carbon dioxide requires a precision on reading
of, say, better than 270 if significant data must be obtained.
In the laboratory such high concentrations were determined by
absorption on NaOH, e.g., by the Or sat's method. The idea is
106
-------
simply to measure the volume of the initial sample and then the
volume of the residual, unabsorbed gas. In terms of percent
error for carbon dioxide, if the residual volume is very small,
the precision of the estimate of the absorbed volume is accord-
ingly very good.
Actually, the sum of acid gases, C02, H2S, and S02, is so
determined. Typically in geothermal gases hydrogen sulfide is
a few percent, or less, of the carbon dioxide content. There-
fore, determining it separately and subtracting the result from
the data obtained by absorption still leaves us with excellent
precision on the estimate of carbon dioxide concentration.
Figure 3 shows a small absorption apparatus for field use. Some
10 ml of 4 N NaOH are prepared in the field to fill up the
measuring pipet, in which a ballast of air is left in order to
keep the increase in pressure within 10% of ambient pressure.
The sample is injected with a syringe, e.g., a plastic disposable
syringe. Typically one or more shots of 10-ml sample can be
made, and reading is from 0 to 0.50 ml with a precision better
than 0.05, making it possible to check for purity of CC>2 + H2S
gases up to in excess of 99.9970, in principals.
As a reasonable precision of 10% can be maintained in sam-
ple additions down to 1 ml by using appropriate syringes, the
lowest concentration of carbon dioxide still measurable with
this device is as much as 50%. Thus, the pocket-sized absorbing
pipet depicted in Figure 4 covers the concentration range of
carbon dioxide from 50 to 99.9970. It is of great use in check-
ing for proper sampling conditions on sampling high C02 gases,
where even limited contamination of the sample by air is bound
to completely distort the chemical composition of the gas.
DETECTOR TUBES
The principle of detector tubes goes back in time to the
lead acetate paper for hydrogen sulfide, the Marsh method for
arsenic, and in more recent times, the 'tupfel1 techniques
worked out by Feigl and others. Use of this technique in vol-
canic and geothermal environment has been discussed and de-
scribed by Elskens, et al. (1964) and F. Tonani (1972). It con-
sists of having a known amount of gas sample flow through the
DT, so that the constituent to be determined in the sample re-
acts with the solid stationary phase, resulting in discoloration
of the packing. The amount of discolored packing is proportional
to the amount determined, and a visual reading can be made if
some conditions are fulfilled. With uniform, plug-wise flow of
107
-------
Figure 4.
O
-------
the gas, as prerequisite to uniform, sharply outlined discolored
zone, precision may be very good.
Techniques described by the manufacturers of DT are devised
for monitoring air in regard to explosive or toxic gases and
vapors. Therefore, they are designed for the low range of con-
centration and/or for using comparatively large volumes of sam-
ple. In analyzing geothermal gases, especially in doing it with
a view to assessing significant amounts of such gases as carbon
dioxide and hydrogen sulfide, the concentrations encountered are
in the high range, and using small volumes of sample is often
required. Therefore, a number of variations of the original
technique had to be devised with the aim of obtaining accurate
results under such diverse conditions.
Experience has indicated that detector tubes for water
vapor, carbon dioxide, and hydrogen sulfide as well as HoS +
SC>2, are dependable for field use. Hydrogen sulfide DT are
very handy and rugged. Sensitivity is sufficient for most
practical purposes, being about 2 ppm for a 100-ml sample.
For determination in air, e.g., on assessing total flow patterns,
sensitivity is even better. The precision, 5-10%, is sufficient
for the purpose indicated here, and accuracy is better than that
obtained by analyzing samples after transportation and storage.
Carried or condensed water in the gas stream jeopardizes
the DT and must be removed prior to analysis. The DT in itself,
however, can work under high temperature and therefore could be
used for superheated steam. For example hydrogen sulfide DT
have been used at extremely high temperature at a volcanic vent
on Mt. Etna, Sicily. Later cross-checks in the laboratory have
indicated that the data are still dependable under such extreme
conditions. The DT still works after complete dehydration, and
in fact detects 82 vapor in addition to l^S under high tempera-
ture conditions.
HoS + SC>2 tubes have iodine as a reagent and therefore can-
not be operated at high temperature. They are used only for con-
centrations in the order of several percent units, by volume,
which do not occur very often in samples from occurrences of
mixed origin.
Carbon dioxide detecting tubes are now manufactured for a
broad concentration range, including the 60% range. However,
the more sensitive tubes in the series can be used up to the
same high concentration level by just using lesser volume of
sample and concurrently applying appropriate changes in the
determination technique. These DT's are especially sensitive
109
-------
to heat and light. They 'age' faster than other DT's, therefore,
the number of different types kept in stock should be as small
as possible.
The reaction in the C02 DT is reversible and the discolored
layer may 'travel' and spread on flushing the DT with a large
volume of gas, resulting in erroneous readings. Thus, flushing
a small sample into a C02 DT for determination requires some
additional precautions besides the quantitative handling of the
sample volume through the tube. Precision is 10% of read value
and may be better under favorable conditions. Sulfur gases, and
more generally reducing gases, interfere with this reaction, re-
sulting in the 'fading away1 of the discoloration. Fade color
in C02 DT points to either the presence of some amount of sul-
fur gases, or over-aged DT. Hydrogen sulfide and sulfur dioxide
can be removed in different ways, e.g., a short segment of an io-
dine DT (one DT can be cut into a number of such scrubbing-
analyzing devices). A cotton-wool plug impregnated with iodine
from a regular alcoholic iodine solution can be inserted into
the inlet end of the DT and scrub all sulfur gases from the
sample. Thus, DT's and the absorption device described earlier
cover the range of concentration from less than atmospheric
(300 ppm) to very pure carbon dioxide.
Water vapor is not of concern as a contaminant to the at-
mosphere, and its determination in naturally occurring geo-
thermal gases serves different purposes. For example, the
amount of water vapor in gas streaming out of natural steam
vents indicates whether there was superheating of the gas stream
by hot country rock. Mixing with air is also better understood
on the basis of such data. Referring such gases as carbon di-
oxide, hydrogen sulfide, methane etc. to the companion amount
of steam is of much use in interpreting the data in terms of
vaporization from the water table, mixing with air in the over-
lying unsaturated zone and so forth.
As water vapor is quite often the overwhelming constituent
in fumarole discharge, its determination in the surrounding air
is very suitable in view of determining the total discharge into
the atmosphere based on average composition and wind velocity.
DT's in general are very suitable as integrating devices in such
cases where the average composition must be determined. In the
case of water and carbon dioxide, however, one has to be wary of
the fact that both reactions are reversible and the discolored
zone 'travels' along the DT. In the water tube, the absorption
of water is reversible,, but the discoloration is irreversible,
therefore, much attention must be paid to controlling the gas'
flow through the DT. Expected precision is also about 5-10% if
the DT is properly handled, e.g., to avoid condensation of water
110
-------
SAMPLING OUTFIT
The sampling outfit proper consists of devices which trans-
fer the gas sample from the sampling point to a sample container
or to a detecting device, e.g.,detector tubes. A number of de-
vices are required to collect transferable samples from dis-
solved, bubbling or streaming gases (see for example Figure 6
showing an extraction device for dissolved gases). They are
not discussed here; only the sample transfer outfit is presented.
Under the constraints described earlier, the requirements put on
such sample transfer outfits are pretty strict, as trouble in
the middle of sampling under rough field conditions, in hardly
accessible areas, may be very costly.
All in all, syringes fitted with hand operated valves are
suitable and safe. For easier operation without giving up much
in the way of safe results, a manifold consisting of two three-
way valves such as depicted in Figure 1 has proved useful. Al-
though similar 'plumbing' may be designed for use with other
gas sampling tubes, it is apparent from Figure 1 that the U-
shaped container makes an especially compact assembly possible,
Figure 5 shows how a similar outfit can be operated by one per-
son only, even in a difficult environment.
Syringes with automatic valves for operation as hand pumps
are marketed by several manufacturers. Automatic valves, how-
ever, are subjected to getting stuck. For example, valves that
operate satisfactorily with water or gas alone are troublesome
when moisture or carried water reaches them and is retained by
capillarity at the narrow sections of the valves.
Draeger supplies a hand pump for sampling through the de-
tector tubes. One stroke is 100 ml and smaller amounts cannot
be handled, which is very inconvenient in the analysis of geo-
thermal gases by DT's. A hand pump marketed by Nalgene has fa-
vorable features. One stroke is only 14 ml, and it is fitted
with a vacuum gauge. This allows for checking that the sample
flows smoothly and completely into the sampling or analysis line.
Moreover, a fraction of a stroke can be measured down to one
milliliter by sticking a scale on the plunger shaft. Admixtures
of water, however, appear to result in frequent breakdowns in
that the pump is no longer tight. While battery powered pumps
are of use when heavier and more sophisticated equipment is al-
lowed, e.g., when sampling man made wells and other easily ac-
cessible settings, their use is not described in connection with
the specific subject of this paper.
Ill
-------
SAMPLE OF
GAS
WATER
OUT
~I WATER IN
Figure 6. Circulation by gravity, hand pump, or power pump
112
-------
ABSTRACT
SAMPLING FOR RESERVOIR FLUID RECONSTRUCTION
M, J. Reed
U.S. Geological Survey, Menlo Park, California
Much can be learned about the geochemistry of a geotherraal sys-
tem if the chemical composition of the reservoir fluid is known.
Downhole sampling methods used extensively by the New Zealanders
provide the most accurate compositions of reservoir fluids. In
the absence of a downhole sample, the fluid composition can be
mathematically reconstructed from the chemical analyses and mass
flow rates of separated gas and liquid phases sampled at the sur-
face.
The major drawback of fluid reconstruction from surface samples
occurs when several permeable strata or several fractures con-
taining fluids of different compositions contribute to the well
flow. In these cases the flow from the wellhead is a mixture of
the reservoir fluids from the producing zone, and the reconstruc-
tion gives an average fluid composition. If a well produces from
a single reservoir fluid, the reconstructed composition will be
independent of the flow conditions; but, if a well produces a mix-
ture of fluids, the composition will change as the flow conditions
change.
At Cerro Prieto, Mexico, several wells produce from liquid-filled
permeable strata or fractures over a large range of depth. Li-
quids of distinct chemical composition enter a well when the pres-
sure in the well drops below the pressure in the permeable zone.
At high flow rates, when the pressure in the well is low, all
the permeable zones are able to contribute to the flow. Sampl-
ing a well at different flow conditions provides a general in-
dication of the composition range in the reservoir fluids.
At The Geysers, California, some wells produce steam from frac-
tures distributed over a large depth interval„ Changes occur
in the gas composition of the steam. In a single well, the con-
centrations of hydrogen sulfide and boric acid tend to be higher
in shallow fractures than in deeper fractures. In this low-pres-
sure system, the deliverability of the fractures appears to con-
trol the change in composition with changing flow conditions.
Sampling at the wellhead under a variety of flow conditions will
supply data on the chemical range of fluid in the reservoir frac-
tures and allow the prediction of composition for any flow con-
ditions.
113
-------
Techniques of Pressure and Temperature Measurement
and Sampling in Geothemnal Wells
by Laurence W. Ross, Larry L, Brown and Ralph E. Williams
Denver Research institute
University of Denver
Denver, Colorado 80208
Abstract
The Denver Research Institute, in the course of a program for
establishing a manual for the design of wells for exploitation
of hydrothermal sources of geothermal energy, developed a need
for highly accurate temperature and pressure measurements, and
for determining the chemical composition of the fluid deep with-
in the well, before flashing took place. Therefore DRI reviewed
the scientific principles available for performing such measure-
ments, and designed a probe for T and P measurements in real time,
This probe was tested for the first time in well 8-1 in the East
Mesa field in September 1976. The sampler selected for high tem-
peratures is based on a design that was developed at the Los Ala-
mos Laboratories, but commercially available samplers can be
used at temperatures up to about 400°F. Laboratory and field
results, and a comparative review of measurement tools available
to the geothermal industry, are reported. Experiences of other
investigators, and current industry trends, are also summarized
114
-------
I. Introduction
The Denver Research Institute has the mission, under Contract EY-76-S-
02-2729.*000 from the U.S. Energy Research and Development Administration
(ERDA), to perform high-accuracy measurements of temperature, pressure and
chemical composition deep within geothermal wells that are flowing in the
two-phase mode. Such wells are often called "flashing" wells.
DRI has developed a tool for measuring pressure and temperature up to
2000 psia (135 atm) and 500°F (260°C) , which should be useful in most of the
two-phase wells and liquid dominated wells now in existence or contemplated.
Sampling devices are available for extracting fluid samples up to 1 liter
in volume from geothermal wells at these levels of pressure and temperature,
but a simplified design involving the actuation of a solenoid valve is pro-
posed.
11. Principles of Pressure and Temperature Measurement in Geothermal Wells
a) General considerations
Measurement of pressure or temperature involves the creation of an
electrical or mechanical signal that is a known function of the measured
effect.
Mechanical measurements within a geothermal well may be stored
inside the measuring probe for later interpretation, but electrical measure-
ments are relayed to the surface for "real-time" observation. Therefore
electrical measurements depend on a transducer to transmit the fundamental
electrical signal to signal-conditioning apparatus either within the probe
or at the surface, to permit a reading at the surface. Common principles
employed in transducers are listed in Table I, and discussion may be found
115
-------
Table I
Types of Transducers
for Measurement of Pressure and Temperature
Electrical Parameter
and Transducer Type Typical Application
a) Passive Transducers
1) Res istance
Potentiometric type Pressure, displacement
Strain gauge Force, displacement
Hot wire meter Gas pressure
Resistance thermometer Temperature
2) Capaci tance
Variable capacitance gauge Pressure, displacement
3) Inductance
Magnetic circuit transducer Pressure, displacement
Reluctance pickup Pressure, displacement
Differential transformer Pressure, displacement, force
Eddy current gauge Displacement
Magnetostriction gauge Pressure, force
b) Self-Generat?ng Transducers
Thermocouple, thermopile Temperature
Piezoelectric pickup Pressure changes
116
-------
In modern textbooks, for example, that of Cooper.1
The majority of electrical measurement principles generate rela-
tively weak signals which suffer in accuracy when transmitted over the
length of a submerged cable, which may extend for three miles below the
surface. This circumstance has led to a need for electronic assemblies that
can be inserted into geothermal wells in order to achieve signal condition-
ing, or at least signal amplification, within the well. Progress in such
"downhole electronics" is only now beginning to be observable, as discussed
below.
b) Pressure measurements
The absolute pressure within geothermal wells may be sensed by one
of three basic principles: piezometry, strain gauges, or bourdon techniques.
All three are available in commercial packages, but they are subject to
various limitations.
The most accurate method is probably the use of piezoelectric
crystals. The Hewlett-Packard Company of Palo Alto, California, manufactures
a pressure gauge for which a resolution of 0.01 psi is claimed up to a total
pressure of 12,000 psi with no hysteresis and with high stability. The
quartz crystal transducer is fabricated such that its resonant frequency
shifts according to the applied pressure. The gauge requires the use of a
single conductor cable. It has no moving parts, and it is fairly compact:
1-7/16 in. 00, 39-3/8 in.long, weight 11 Ib. The most serious limitation
of this gauge is the maximum temperature limit of 150°C imposed by the
electronic elements inside the probe.2
Another very accurate system is the Sperry-Sun downhole concept,
in which a gas (e.g., nitrogen) is supplied through a small tube to a cham-
ber deep within the well, and the downhole pressure is balanced at the surface.3
117
-------
The tube is 3/32 in, 00, stainless steel. This system is fundamentally the
same as that reported by Fournier and Truesdel1 of the U.S. Geological Sur-
vey,1* but the Sperry-Sun system incorporates an electronic system for data
readout. Reports from field applications have indicated that the Sperry-
Sun system is accurate to within 2 psi, and that it is limited to a depth
of about ^,000 ft. because of the adverse effects of heat and salinity on
the stainless steel tube. Another potential limitation is the chamber dia-
meter of about 3i in. OD which may present a significant obstacle to flow
within the we!1.
Strain gauges are widely used in the geothermal industry because of
their simplicity and flexibility, but their low electrical output (i.e.,
their low change in resistance as a function of applied pressure) generally
limits their use to shallow applications where temperatures are sufficiently
low to permit downhole electronics for amplification of the signal for
transmission to the surface.
A purely mechanical device for pressure measurement is marketed by the
Kuster Company of Long Beach, California. It features a bourdon element
that drives a scribe, creating a trace on a brass foil, using a mechanical
clock to move the scribe at a constant rate, thus creating a record of
pressure vs. time. The foil is interpreted later under a magnifying
assembly. The gauge is compact, but is suspended on a wire line that pre-
sents some risk of kinking or breaking when the flow is fairly rigorous,
according to field experience.
The Denver Research Institute has a need for relatively high accuracy
up to 250°C and 2000 psia, and this accuracy cannot be achieved with any of
the principles discussed above, at the depths of interest (up to 2 miles).
To achieve the desired accuracy, DRI has adopted the potentiometric type of
118
-------
transducer, because it is the only type that will provide a relatively high
electrical voltage output without the necessity of downhole electronics.
Four transducers were ordered from Gulton Industries in Costa Mesa, Calif-
ornia, including two specially built transducers for measurement to 2000
psia at 250°C, and two stock transducers for measurement to 500 psia at
180°C. The transducers depend on a bourdon tube to drive a slide across
a resistance winding; the winding is excited by a constant current, and
the measurement is taken as the potential at the resistance according to
the slide's position. The temperature characteristics of the Gulton trans-
ducers are fairly linear except at the most extreme conditions (Fig. 1).
The ORI transducers feature resistance windings with 330 turns of wire,
so that the 2000 psia transducers will record pressure changes in steps
of about
ps,.
This is not necessarily a disadvantage for absolute pressure measurements,
since any measured "step" must occur at a pressure level where the slide
has just encountered the next turn of the winding, and the value of the
resistance (and therefore the pressure) at this particular point is known
within fairly high precision, e.g., 2 to 3 psi . However, the usefulness
of these transducers for recording pressure fluctuations is still not well
understood, and field tests in geothermal wells are needed to resolve this
question.
c) Temperature measurements
Measurement of temperature in geothermal wells is simpler than
measurement of pressure, because no mechanical force or displacement is
Involved,
Thermistors are widely used for temperature measurement in geothermal
119
-------
2000
200°£ 400°F
75° f *3X)°F/5000F
1500
A/
RN
E-
AA/V
vv\
M^
BLK
E +
1 +
YSI TYPE 139
TEMPERATURE PROBE
BLK
TO SHIELD
(ARMOR)_[_
-GULTON PRESSURE
POTENTIOMETER
Figure 2.
Wiring diagram of the pressure and temperature
sensing elements used in the DRI probe
120
-------
wells, but DRI has elected to use a platinum resistance thermometer because
it can be purchased off-the-shelf at a guaranteed accuracy well within 1°F.
The electrical circuitry associated with the platinum resistance thermometer
is integrated with the wiring for the pressure transducer as shown in Fig. 2,
such that both employ the armor of a seven-conductor cable as ground.
Thermocouples can be used for downhole temperature measurement, if
a thermopile can be assembled that is sufficient to create an electrical
signal that can be measured accurately at the surface, and if its reference
junction can be held at a cold temperature for a sufficient time. This is
the technique used by the Los Alamos Laboratory.
A mechanical gauge for temperature measurement is available from
the Kuster Company (see ll.b above), and it can be used in combination with
the Kuster pressure gauge to obtain a P and T log simultaneously.
III. Sampling of Fluids in Geothermal Wells
a) General considerations
Sampling of geothermal fluids is generally achieved at the well-
head. In the case of two-phase flow (i.e., flashing flow), which will be
the most prevalent case in wells intended for electrical power production,
this means that both a vapor phase and a liquid phase must be sampled and
analyzed. Furthermore, both flow rates must be known accurately.
Sampling at the wellhead introduces errors due to uncertainties
inchemical analysis and flow rates, and this is especially true for gases
such as C02, H2S and NHj which are distributed between the vapor and liquid
phases of the fluid,
This situation has led the Denver Research Institute to seek a
121
-------
sampling concept which could be employed deep within producing geothermal
wells, at depths where the vapor phase has not yet developed in the flowing
fluid.
Several sampling devices suitable for use in geothermal wells are
available, including at least two that are commercially manufactured.
Sampling devices that operate on mechanical principles have been
employed for many years, but only the Kuster design (Kuster Company, Long
Beach, California) and the Klyen design (Forgan Jones Limited, Auckland,
New Zealand)5 are presently available for purchase. All mechanically actu-
ated samplers rely upon a force applied at a given time in order to open
(or close) a valve port. Either a check valve prevents the incoming sample
from reversing direction or, in the case of the "flow-through" principle of
sampling, closing of a valve trips the sample in a chamber.
Electrically actuated samplers are available from the oil well
service industries, but they feature a maximum temperature limit of about
300°F.
The Los Alamos Laboratory has developed a double-chamber sampler
for extraction of liquid samples within geothermal wells at temperatures
up to 200°C. Two chambers, each approximately 250 ml in volume, can be
opened sequentially (before-and-after) by actuation of a drive motor.
The Denver Research Institute has proposed a sampling concept
based upon actuation of a solenoid valve, for use of temperatures up to
250°C. A representative design is shown in Fig. 3. The solenoid valve is
actuated at the surface at the desired time, admitting fluid into the evacu-
ated chamber at left; a check valve retains the sample in the chamber, as
in other sampler designs (see above). Extraction of the sample for analysis
Involves a certain degree of design and technique, depending on the measured
122
-------
CO
HIGH-TEMP SOLENOID VALVE -
CHECK VALVE -
SHUT-OFF VALVE -
EVACUATED
SAMPLING CYLINDER?
Figure 3. Proposed high-temperature solenoid valve-actuated
sampling system
-DRILLS C'SINK FOR NO 10-32 S.H.CS
12 PLACES
WELD-' ^CIRCLE SEAL FILTER
MATERIAL, TUBE 2"O.D. > 1.5"I.D.
COLD DRAWN, STEEL IOIB
Figure 4. Schematic diagram of the DRI probe for measuring pressure and
temperature in geothermal wells
-------
quantities of interest.and their relative importance.
IV. Field Experiences in Measurement
The Denver Research Institute visited the East Mesa geothermal field
in the Imperial Valley in September 1976 in order to perform the first in
a series of three tests in geothermal wells as required by Contract EY-76-S-
02-2729.»000 with the U.S. Energy Research and Development Administration
(ERDA)
DR1 has constructed a probe to house the pressure and temperature
measuring capabilities described in Section II above. The probe is designed
to contain the measuring elements in a body of 2 in OD. The probe is shown
schematically in Fig. 4. It is about 30 in. long, including the cable head
connector for connection to a seven-conductor armored cable. The probe is
shown ready for application in Fig. 5-
DRI has also constructed a calibration assembly for testing of the
pressure and temperature sensors under laboratory or field conditions. The
calibrator is designed for testing up to 3000 psia and 600°F, and it measures
3 in. ID and kS in. interior length. A dead weight test assembly provides
the calibration pressure, and temperature adjustment is provided by electri-
cal heater windings and a West model 400 controller. The f.>.''..?-ready cali-
brator assembly is shown in Fig. 6.
The U.S. Geological Survey provided the field support for application
of the DRI probe at the East Mesa field in September 1976. Measurements were
performed in well 8-1 of the East Mesa field. Figure 7 shows the operation
of erection of the riser pipe (6 in. diameter) which is designed to contain
the probe and the attached sinker bars, and it also shows the U.S.G.S. logging
truck which provides the cable, winch and electronic measurement capability
124
-------
-
Ln
Figure 5. The DRI probe, assembled and ready for use
-------
Figure 6. Calibrator assembly for pressure and temperature
calibration of the DRI probe
Figure 7. Erection of riser pipe (6 in. diameter) to hold
the probe and attached sinker bars before
insertion down the well. This well is no. 8-1
of the East Mesa geothermal field
126
-------
for the downhole test. Figure 8 shows a closer view of the wellhead at well
8-1, illustrating the mounting of the sheave pulley that directs the cable
from ground level to the top of the riser pipe.
The cable enters the riser pipe over an upper sheave pulley and through
a "lubricator" as shown in Fig. 9- The splashing that is visible in Fig. 9
is necessary until the probe and cable are sufficiently submerged to provide
a significant weight downward into the well, after which the rubber collar
of the lubricator is tightened around the cable, and splashing ceases. The
clamping force of the rubber collar is adjustable, and it is sufficient to
lift the entire riser assembly (three 8^-foot pipe sections, plus lubricator)
from the wellhead.
V. Downhole Measurement Tools of the Future
a) Downhole sensors
Sensors for use at geothermal temperatures and pressures are probably
adequate for the immediate needs of the geothermal industry. Temperature
measurement, in particular, is not limited by the severity of geothermal
conditions. Pressure sensors are probably adequate up to 500 F, when selected
among the alternatives presenraci by DRI, Sperry-Sun, and present-day strain
gauge technology. Above 500°FS however, the limitations imposed by insulat-
ing materials are critical, and improvement must come from the employment
of inorganic materials .
b) Downhole electronic assemblies
The development of simple amplification circuitry by use of diodes
and triodes fabricated from metal-ceramic configurations seems certain to
extend the capabilities of downhole measurement, in the near future. However,
127
-------
CO
Figure 8. Close-up view of well
no. 8-1, showing method
of attaching lower sheave
pulley for the 7-con-
ductor cable.
Figure 9. Top of riser pipe, show-
ing cable entering the
riser through the lubri-
cator. The blade-
shaped device is a cable
rest. Splashing occurs
when lubricator is not
yet clamped ti
-------
ceramic materials are fundamentally vulnerable to moisture and gas intrusion,
and this is a significant development problem in their application as insu-
lators under geothermal conditions.
For the near term, cooling of electronic components by Dewar flasks
or other devices offers an ingenious means of obtaining sensitive downhole
measurements that are not obtainable by any other means, as demonstrated by
the Los Alamos Laboratory.
c) Summary
The present generation of tools for measurement of physical condi-
tions within geothermal wells is subject to an upper temperature limit of
about 250°C, mainly due to the limitations imposed by synthetic resin insu-
lating materials. It is reasonable to expect that electronic assemblies
with insulation provided by ceramic materials will permit this temperature
limitation to be raised. Sampling of fluid within geothermal wells will
probably continue to be achieved mainly with mechanical devices, because
electrical components such as motors and solenoid valves probably cannot
operate at temperatures in excess of 250°C.
References
1. Cooper, W.D., Electronic Instrumentation and Measurement Techniques.
Englewood Cliffs, N,J. : Prentice-Hall, Inc. (1970). Chapter 14.
2. Miller, G.B., R.W.S. Seeds and H.W. Shira, Paper no, SPE 4125 presented
to the Society of Petroleum Engineers of AIME (1972).
3. Weeks,S.G. and G.F. Farris, Oil Gas J. (Jan. 5, 1976).
4. Fournier, R.O. and A.H. Truesdell, U.S. Geol. Soc. Surv., Prof. Paper
750-C, pp. 146-150 (1971).
5. Klyen, L.E., Geothermics 2, 57-60 (1973).
129
-------
Collection of Chemical, Isotope, and
Gas Samples from Geothermal Wells
by
Nancy L. Nehring and Alfred H. Truesdell
U.S. Geological Survey, Menlo Park, Caliornia 94025
ABSTRACT
Reconstruction of the downhole composition in a geothermal well
requires field measurements of total fluid enthalpy and separator
pressure along with chemical, isotope, and gas analyses of separated
water, steam condensate and gas phases. Analyses of each phase for
components distributed between phases are necessary because the fluids
may not be in equilibrium at the sampling points. Cooling after
separation must be sufficient to prevent loss of volatiles and water
vapor. The collection and analysis schemes presented are designed to
allow complete chemical and gas analyses and isotopic analyses of major
carbon, oxygen, hydrogen, and sulfur containing species.
INTRODUCTION
Steam Wells
Steam wells such as those at The Geysers do not present major
collection problems in obtaining a representative sample. It is usually
sufficient to collect a sample from a bypass on the wellhead, taking
precautions to avoid condensation in the bypass. This can be
accomplished either by flowing sufficient steam to insure the temperature
of the collected steam is the same as that in the well or by insulating
a large diameter bypass. A probe Inserted to the center of a large
diameter .flowing steam line has been found to provide representative
samples— .
Hot Water Wells
Sampling hot water wells is more difficult. Small amounts of fluid
must be removed from a rapidly flowing, usually two-phase, fluid in a
controlled manner so the samples are representative of the whole fluid.
Simply connecting a condenser to a pipe carrying a water-steam mixture
does not produce the correct proportion of the two phases. It is
possible to collect good samples of the water and steam phases after
separation of the two-phase fluid is completed in a cyclone separator
and use enthalpy measurements to calculate the total compositiqn.
130
-------
SEPARATION OF HOT WATER AND STEAM
Understanding the operation and limitations of separators is
necessary before reliable samples can be collected. A centrifugal Webre
separator (Fig. 1) is provided on most production wells. It consists of
a vertical drum with a tangential inlet at the center, an outlet for
water at the bottom, and a central outlet tube open at the top which
collects the steam. The two-phase fluid enters tangentially at the
center, spins rapidly around the outer wall of the drum with the water
collecting on the outside and falling to the bottom and the steam moving
to the inside and flowing down the central tube. The steam-water level
is set by adjustment of a valve in the water flow line. Since water in
the steam line to the turbines would be highly undesirable, it is usual
to stabilize the water level by allowing a small flow of steam into the
water line. If sampling is then done from the water line, apparent
nonequilibrium fractionation of gases between steam and water will be
observed. Allowing some steam in the water line also would produce
slightly lower apparent enthalpy but this can be avoided by careful
adjustment of separator water level during enthalpy measurements. Water
samples should be taken, if possible, from a separate tap below the
water outlet of the separator where the water is less likely to be
contaminated with steam (Fig. 1).
If a large separator is not available, the sample can be taken
with a mini cyclone separator (Figs. 2 and 3)< With this separator the
water and steam flows are adjusted to allow some water to issue from the
steam outlet while collecting from the water outlet and vice versa. The
separator and the tube from the bypass must be well insulated and the
flow limited to maintain the pressure in the separator during sampling
nearly equal to the pressure in the well. These measures are necessary
to insure that condensation does not occur before or in the separator
and that the water and steam collected from the separator have the same
compositions as the phases in the well.
Enthalpy Measurements
Enthalpy measurements are routinely made on production wells by
measuring the flows of steam and water from the large separator. The
measurement of total enthalpy without a large separator is difficult and
approximate but1 may be done by critical lip pressure measurements or by
measuring the steam/gas ratio at two different pressures^- . If it is
known that boiling in the formation does not occur, the bottom hole
flowing temperature can be used to obtain a value for the enthalpy.
131
-------
INLET
U)
WATER SAMPLING POINT
WATER OUTLET
=® STEAM SAMPLING POINT
STEAM OUTLET
2 FEET
Figure 1. Cross Section of Webre
Separator showing best
sampling points
c
o o
INLET
O
WATER OUTLET, VALVE,
AND PRESSURE GAUGE
O O
U
STEAM OUTLET
1 inch
Figure 2.
Mini Cyclone Separator
of New Zealand design
-------
OJ
WATER OUTLET
INLET
Figure 3.
Top view of Mini Cyclone
Separator with pressure
gauge attached
10
• 20
-30
-40
80
Figure 4.
Steam-gas Separator modified
from a 100 ml graduated
cylinder
-------
CONDENSATION AND COOLING
Steam and water exit from the separator at temperatures
considerably above surface boiling and must be cooled for safe handling
and to avoid the loss of volatile substances. This is most easily
accomplished with a double coil condenser of V stainless steel tubing
connected to the separator (or bypass) by a regulating valve. Larger
diameter tubing may allow inconveniently large slugs of gas and
condensate to form. The first coil is immersed in water which is allowed
to boil, effectively reducing the temperature of the fluid sample to
100°C or less. The second coil is held in an ice or water bath or in
the air according to the desired temperature of collection. The water
and steam phases can be cooled by serial passage through the same
condenser but are treated differently when they emerge.
SEPARATION OF STEAM CONDENSATE AND GAS
Collection of steam condensate and gas from a steam well is the
same as from a water-steam separator but the methods used may vary
because of different steam/gas ratios and because it may be necessary
to check for incomplete steam separation in the separator. The steam
condensate and gas issue from the condenser in discontinuous slugs. The
use of V tubing produces relatively small, uniform slugs and the
average composition of issuing fluid is uniform over the period of
collection of any single sample.
A sample of gas without the steam condensate is best for gas
analyses and for the study of isotopes of carbon, sulfur, and hydrogen
in gas components. A steam condensate-gas separator (Fig. 4) was
constructed from a graduated cylinder by attaching two tubulations at
the top and one at the bottom. The flow from the condenser enters at
the top (vinyl tubing is used for all connections) and the flow of
condensate from the bottom is restricted with a clamp so that gas alone
issues from the remaining top tubulation. The tubing on the bottom is
inserted in a large bottle of steam condensate to prevent back flow of
air into the separator during gas collection.
FIELD MEASUREMENTS
Steam/Gas Ratios
The steam/gas ratio may be measured by timing the alternate
displacement of gas by condensate and vice versa in a bottle with one
tube at each end (the least accurate method because during the
condensate collection part of the volume is occupied by rising gas
bubbles) , by measuring the flow of water with .a graduated cylinder and
the flow of gas with a soap film flow-meter from the gas-water separator
134
-------
(more accurate but requiring exact balancing of flows), or by allowing
gas to displace condensate in a tube or separator of known volume and
measuring the quantity of condensate displaced (the most accurate field
method). None of these field methods measure the quantity of gas
dissolved in the condensate which may be a large part of the total when
the steam/gas ratio is high. More accurate steam/gas ratios may be
obtained from laboratory analyses of steam condensate and gas collected
in a single bottle (see below).
pH Measurements
During separation and cooling the pressure of CC>2 is above or near
one atmosphere and C02~supercharged water or steam condensate issues
from the condenser. Part of the CC>2 is rapidly lost and the pH and
t^COs, HCC>3 and C0% concentrations change until equilibrium with
atmospheric C02 is established. For this reason pH measurements and
field alkalinity titrations are not very useful. It is more useful to
collect all the C02 by SrCOs precipitation (described later) and to
measure the pH and alkalinity on air-equilibrated samples in the
laboratory. This pH measurement can be related to the state of ionization
of all weak acids and bases.
Pressure and Temperature Measurements
Separator pressure is needed to calculate the water/steam ratio.
Temperature measurements of the water and steam condensate issuing
from the condenser are needed to calculate the distribution of gases
between phases and are desirable to determine if isotopic fractionation
of D or 180 occurred.
COLLECTION OF WATER
Figure 5 is a schematic diagram of the collection procedure
discussed in the paper.
Dissolved Salts
Water samples for dissolved salts are filtered through a 0.45 y pore
membrane filter (or if suspended clay is suspected, through a 0.1 y pore
membrane) and part is acidified to pH 2 with HCl to prevent precipitation
of Ca or Mg carbonates or adsorption of cations on the walls of the
bottle. The dissolved salts are analysed by conventional water chemistry
methods. Silica is usually present in amounts exceeding saturation with
amorphous silica at room temperature so a separate sample is diluted
1:10 (or 1:20 for fluids originally over 300°C) with silica-free water to
prevent precipitation and preserve monomeric silica necessary for the
molybdate method. Other preservation methods are required for trace
elemental analysis.
135
-------
Well
water
condenser
CO
temperature
pH
chemistry
silica dilution
D, 0-18 water
C02 (+SrCl2'NHifOH)
H2S (+CdCl2)
NH3 (+acid)
Oit (+formaldehyde)
(C-14)
water/steam
separator
steam
pressure
enthalpy
condenser
steam condensate
steam condensate/gas
separator
total gas
+ condensate
(w/ NaOH)
gas
steam/gas ratio
temperature
pH
chemistry
D, 0-18 water
(dissolved C02s H2S)
NH3 (+acid)
dissolved gases (w/o NaOH)
tritium
(C-14)
gases (w/ NaOH)
C-14
Figure 5. Schematic diagram of sampling procedure
-------
Isotopes
Water samples for isotopic analysis (180 and D) are collected in
glass bottles with polyseal caps without filtration (which might allow
evaporation). A sample for isotopes in dissolved sulfate (180 and 3I+S)
is collected in one or two liter bottles according to concentration (at
least 20-30 mg SO^ is needed for analysis) and preserved with 4 ml/A of
formalin solution to prevent bacterial oxidation of H2S to SO^. Collec-
tion of dissolved C02 for 13C and H2S for 3ifS will be discussed under
dissolved gases.
Dissolved Gases
Although most gas partitions into the steam phase, enough remains
in the water to make analysis of dissolved gases important. The most
effective way of preserving C02 dissolved in water (not only C02 but all
carbonate species) is to precipitate SrCOs in a glass bottle by adding
concentrated NHi^OH saturated with SrCl2. The ammonia buffers the
solution so that all forms of C02 are converted to C0$ and precipitated.
Precipitation must be done immediately because C02 is rapidly lost to
the atmosphere. Later in the laboratory the sample is filtered, weighed,
and saved for isotopic analysis.
Hydrogen sulfide is similarly precipitated as CdS with
solution. The addition of buffer is unnecessary because the natural
bicarbonate - carbon dioxide buffer maintains the solution at a high
enough pH (>3) to prevent loss of H2S with an excess of Cd present.
Ammonia should be preserved in a sealed bottle, acidified if the
discharge is alkaline and analysed as soon as possible. Other dissolved
gases may be collected by allowing the water to almost fill a weighed,
evacuated gas bottle (Fig. 6) through a vinyl tubing attachment to the
condenser. The head space of this bottle is analysed for gases in
equilibrium with the water.
COLLECTION OF STEAM CONDENSATE AND GAS
Dissolved Salts
Sampling of condensate may be done directly from the condenser
or through the bottom tubulation of the steam condensate-gas separator
and preserved as for water samples. Chemical analysis may be limited
to boron, ammonia, bicarbonate and other substances that have volatile
forms that may be carried in steam. On samples from a separator it is
necessary to determine the completeness of the separation by analysis of
the condensate for Cl, Na, or another nonvolatile substance.
137
-------
0.7 by 10 cm glass tube
with closed end and
0.03" hole
Threaded nylon
bushing
008 viton o-ring
Threaded glass
adapter
Ace
glass
part
5027-20
300 or 500 ml gas bulb
Figure 6. Exploded view of gas sample bo.ttle. The bottle is
opened by sliding the hole inside the o-ring. The
tube may be evacuated with the bottle closed by
clamping a second o-ring and washer to the top of
the adapter. Gas bottles with NaOH solution are
evacuated by a water aspirator with gentle
boiling of the solution.
138
-------
Isotopes
Condensate samples for 180 and D are collected as described above.
Sulfate is only contained in steam condensate if produced by the
oxidation of t^S and is not collected for isotopic analysis. CC>2 and
H2S may be precipitated as described above from the condensate or
separated from the caustic in the gas bottle (described later).
Fractionations resulting from solution of gases in water are small and
either of these separates may be analysed depending on which contains
the major amount of the gas.
Tritium and carbon-14 may allow limits to be placed on the age of
geothermal fluids or on the amount of dilution by near surface water.
The analyses are difficult and costly and geothermal fluids contain
very little of these isotopes so the collection should be carefully made
to avoid contamination. A steam condensate sample should be collected
for tritium, as a brine sample requires distillation to remove the salts.
Dry, clean bottles should be well flushed by flowing the sample through
vinyl tubing to the bottom of the bottle and allowing the bottle to
overflow for some time. The bottle should then be sealed with a polyseal
cap and taped with vinyl electrical tape applied in the clockwise
direction.
Carbon-14 can be collected from steam condensate by precipitation
of SrCC>3, but the amount of sample (>3 gms of carbon usually needed for
analysis) requires long settling times. The collection of steam phase
CC>2 from the gas-water separator into gas bubblers containing C02~free
ammonia may be quicker and is less prone to contamination. If the
steam/gas ratio is very high this method may require inconveniently
long collection times, and SrCOs precipitation from steam condensate
may be preferred.
Gases
The total cooled steam sample (gas and condensate) is collected into
an evacuated 300 or 500 ml gas bottle (Fig. 6) containing 50 to 100 ml
of 4N NaOH (prepared as carbonate-free as possible). The flow from the
condenser should be reduced by regulating the valve on the separator or
steam bypass. A pressure-relief valve may be used to prevent over-
pressuring the gas bottle. This sample allows the most accurate
measurement of steam, C02 and H2S but the quantity of residual gases
may be too small for some determinations. In this case a gas-only
sample (without condensate) may be collected from the top of the gas-
water separator. A larger quantity of caustic (100 to 150 ml) may be
required because a larger quantity of C02 and H2S will be collected.
139
-------
The flow of gas into the bottle may at first exceed the flow into the
gas-water separator and the gas flow into the bottle should be
restricted by pinching the vinyl tubing until a balance is achieved.
Shaking the gas bottle increases the rate of CC>2 and H2S absorption and
may be necessary at the end, although allowing the gas to bubble through
the caustic produces adequate absorption during most of the collection.
The C02 and H2S analyses are obtained by wet chemical analysis of NaOH
solution. The other gases are analysed by gas chromatography. Dissolved
NH3, C02, and H2S in the steam condensate may be collected and analysed
by the same methods described above for water.
— J. Farison, Oral commun., 1977.
2/
— James, Russel, 1964, Alternate methods of determining enthalpy and
mass flow: Proc. U.N. Conf. on New Sources of Energy, Rome, 1961,
v. 2, p. 265-267.
140
-------
GEOTHERMAL DOWNHOLE SAMPLING INSTRUMENTATION
R. Fournier and J. M. Thompson
U. S. Geological Survey, Menlo Park, California
A sampling device 5 cm in outside diameter,
with an internal volume of about 500 ml has been designed
to collect liquid and gas samples in wells where both
steam and water are present at temperatures up to about
280°C. Detailed plans for construction of the sampler
have been published by Fournier and Morganstern (1971) .
A long flexible stainless steel tube serves as the support
cable. The sample device is lowered in the open position
and fluid flows through it during its descent. This pro-
cedure circumvents the flashing of water to steam and pre-
ferential partitioning of volatiles into the sampler that
may occur when an evacuated chamber is opened at the point
of collection. Closure is accomplished by gas pressure
applied from the surface through the flexible tube to a
piston and plunger within the sample chamber. Any non-
reactive gas is a suitable pressure medium. Continued
application of gas pressure during withdrawal of the device
prevents leakage, either into or out of the sample chamber
that otherwise may be caused by changing conditions of
temperature and pressure. The sampling device has been used
successfully to collect water and gas samples from research
holes drilled in hot-springs areas of Yellowstone National
141
-------
Park, Wyoming. The main limitation to the use of the device
is the need for using a flexible stainless steel tube as long
as the well is deep. At least one commercially available down-
hole pressure measuring instrument uses a long, flexible stain-
less steel tube for its operation, and it may be possible to
use the same tube in combination with the sampler described
here to collect down-hole water and gas samples.
A second down-hole sampler is now being tested that
is similar in basic design to the first sampler, but is sus-
pended on a wireline and is closed by a spring-loaded mechan-
ism triggered by an inertial mass. The mass operates in a
similar manner to the one described by Klyen (1973) : at the
depth where a water and gas sample is to be collected the
sampler is allowed to drop as quickly as possible for about
3 to 5 meters and then is brought to a sudden stop. An in-
ertial mass, suspended on a string, continues its downward
movement and triggers the closure mechanism.
BIBLIOGRAPHY
Fournier, R. O., and Morganstern, J. C. 1971, A device for
collecting down-hole water and gas samples in geothermal
wells: U.S. Geol. Survey Prof. Paper 750-C, C151-155.
Klyen, L. E., 1973, A vessel for collecting subsurface water
samples from geothermal drillholes: Geothermics, v.2,
p. 57-60.
142
-------
ABSTRACT
CASE HISTORY OF GEOTHERMAL EFFLUENTS SAMPLING
FOR FOUR GEOTHERMAL WELLS
G. R. Conner and P. B. Needham, Jr.
College Park Metallurgy Research Center
Bureau of Mines, U. S. Department of the Interior
College Park, Maryland
Detailed studies of the chemical nature of flowing geothermal
brines have been conducted for the past 2 years at geothermal
wells located in the Imperial Valley of California. These con-
tinous onsite brine analyses are being carried out in support
of a Bureau of Mines' geothermal test facility designed to
study the serious corrosion and scaling processes associated
with resource recovery from these brines.
Corrosion studies have been conducted on two geothermal reser-
voirs of the Imperial Valley; (1) the Salton Sea geothermal
field (geothermal wells Magmamax #1 and Woolsey #1) , with brines
ranging from 24 to 32 wt-pct total dissolved solids (TDS), pres-
sures ranging from 250 to 500 psig, and temperatures ranging
from 1800 to 305° C; and (2) the East Mesa geothermal field (geo-
thermal wells Mesa 6-1 and Mesa 6-2), with brines ranging from
1 to 7 wt-pct TDS, pressures ranging from 100 to 200 psig, and
temperatures ranging from 100° to 200° C.. An important step to
understanding the observed corrosion and scaling phenomena is
the development of the proper model of the1 chemical environment
for a flowing geothermal brine. To develop such a model, accu-
rate elemental determinations, a determination of the major
chemical ions, pH determinations, and accurate gas component
determinations must be obtained.
Analytical procedures have been developed to determine the major
anions in solution (chloride and carbonate species) by titration
techniques. Determinations of pH have been made by usual methods.
Analytical procedures are described for the determination of 10
major brine constituents (Na, Ca, K, Li, Sr, Mn, Fe, Pb, Cu, and
Ag) by atomic absorption spectroscopy (AAS) and for the determina-
tion of gaseous components in the brine (02, N2, H2, CH^, and H2S)
by gas chromatography (GC).
143
-------
Procedures for extracting the brine samples from operating
geothermal systems were developed for the wet chemical, AAS,
and GC analysis. Studies were made of various types of sample-
condensing coils, the location of sampling ports with respect
to flow, various brine-flow patterns, and the effect on the chem-
ical analyses of pipe, valves, and elbow geometries. The use
of chemical additives to the brine samples in order to preserve
species in solution for AAS and carbonate analysis is also des-
cribed.
Results will be discussed for tests on the East Mesa geothermal
field during 1975 and on the Salton Sea geothermal field in 1976.
These results have shown that the dynamic nature of the chemistry
of brines flowing from geothermal wells requires constant onsite
monitoring of the major brine constituents. In particular, the
onsite chemical studies have clearly shown the dependence of the
observed brine chemistry on (1) the brine flowrate from the well,
(2) reservoir and/or well design engineering changes, and (3)
the total operating time of the geothermal well. In addition
substantial differences were also noted in the chemical makeup
of individual geothermal wells located on the same geothermal
field. These results, therefore, show that each of these condi-
tions is en important parameter that must be reported with any
reliable brine chemical analysis.
144
-------
ABSTRACT
ESTIMATION OF POLLUTANT CHARACTERISTICS FROM GEOCHEMICAL
SURFACE INVESTIGATIONS
F. B. Tonani, Geochemex, and
H. T. Meidav, Geonomics, Inc., Berkeley, Calif.
Estimation of reservoir chemistry is ipso facto an estimation
of pollution potential due to leakage of effluents during ex-
ploration or production from that reservoir. The estimation
of reservoir chemistry is one of the important goals of field
chemis try.
Reservoir chemistry may be estimated under a variety of condi-
tions and in many different ways:
1. When the water issuing in a hot spring is proven to be a
direct leakage of reservoir water (and hence chemical ther-
mometry may be applied in its simplest form).
2. When the spring or well water is a mixture of reservoir water
and, for example, shallow groundwater, mixing models are
utilized to determine geochemical temperatures, and con-
currently resulting in estimation of true reservoir chemistry.
3. When a mixing model cannot be determined, based upon the
chemical analysis itself, external constraints can be used
to determine the mixing models. Such constraints may be
other types of data, such as electrical resistivity. When
a reasonable estimate of reservoir temperature may be made,
a correlation of temperature vs. salinity may result in a
similar constraint.
4. In absence of specific information about the reservoir water,
statistics of chemical composition of underground or surface
water in the same general area usually provides a reasonable
prediction of reservoir chemistry.
5. When a fumarole occurs in the area, the gas chemistry can be
employed to infer the composition of the liquid phase, through
the computation of chemical equilibrium. This applies strictly
to volatile substances, but previous knowledge of correlation
between the volatile substances and other dissolved consti-
tuents may result in estimates of the composition of reservoir
water.
145
-------
ESTIMATION OF POLLUTANT CHARACTERISTICS FROM
GEOCHEMICAL SURFACE INVESTIGATIONS
F. Tonani and H. T. Meidav
Knowledge of the chemical composition of the water and gas
phases in fluids produced from a geothermal reservoir is part
of a series of iterations whereby an increasing refinement of
the information is possible with increasing details of explora-
tion. In the initial phase of exploration, only broad knowledge
of the geothermal system can be expected from surface reconnais-
sance. As the knowledge of the reservoir improves because of
increased exploration activity, the drilling of some exploration
holes, and the production of those holes, a finer and finer
understanding and knowledge of reservoir chemistry becomes
possible.
Geochemical exploration, like all other surface-types of
exploration, is a process of establishing a geological-chemical
model of the system. If that model turns out to be promising
from a geothermal exploration point of view (especially when
the chemical data are supported by other types of data, such as
geophysical data), drilling activity will probably follow, to
test the viability of assumptions regarding the reservoir. How-
ever, a great deal may be said about the geochemical nature of
reservoir fluids from surface investigations. For the future
146
-------
developer of the geothermal reservoir, it is important to know
whether considerable problems due to unique chemistry or unique
pollution or corrosion problems are likely to occur, because
such potential problems may affect the economics of the entire
operation. In our discussion, we wish to provide some insights
on how, in general, we can assess the gross chemical character-
istics of a reservoir from the data of water and gas analyses
that are obtainable on the surface.
It is important to keep in mind that in many classical geo-
thermal areas, a considerable natural flow of water and gases
has been going on over extended periods of time. Under such cir-
cumstances, by studying the chemical composition of the fluids
coming to the surface of the earth and discharging to the at-
mosphere, it might be possible to estimate the change that occurs
in natural emission as the result of intensive production from
the reservoir. Under very large scale production conditions,
it is permissible to think about environmental improvement which
might result from a controlled geothermal production. For
example, the amount of energy that is continuously being released
by a large power plant, such as The Geysers geothermal field
power plant complex, is equivalent, over a period of several
years of production, to the total amount of energy released by
a large volcanic eruption. Hence, it is conceivable that under
the geological conditions in some geothermal areas, besides the
reduction of efflux to the environment through hot springs and
147
-------
fumaroles in the area, the extraction of energy may result in a
measure of control over potential volcanic eruptions which
would have brought to the surface a large amount of solid,
liquid, and gaseous materials. This is especially true when the
volcanic activity may be shown to be related to hydrothermal
activity within the upper part of the crust such as at Pozzuoli
near Naples, Italy. An underlying axiom is that the release of
energy is directly related to the release of chemicals to the
atmosphere and vice versa. However, the study of the potential
change in the environment from the point of view of effluent
discharge reduction is not the main thrust of this paper.
There are a number of ways by which the reservoir chemistry
may be estimated. A variety of geological and chemical field
conditions affect the chemistry of the fluid which issues at
the surface, and hence, a number of different approaches must
be employed to estimate reservoir chemistry.
Case 1: Direct Sampling of Reservoir Water (Figure 1)
Under the simplest conditions, the water issuing in a hot
spring at the surface is an actual sample of direct leakage from
the reservoir at depth. In such a case, chemical analysis of
surface water would provide an estimate of reservoir chemistry.
The assumptions made in assessing reservoir chemistry in this
simplest case is that : (1) little or no re-equilibration or
change in composition occurs at lower temperatures as the water
flows from the reservoir to the surface, and (2) the hot water
148
-------
HOT SPRING
SELF SEALED
FRACTURE
METEORIC WATER
OR IMPERVIOUS CAP ROCK
RESERVOIR
Figure 1. Direct sampling of reservoir water—some hot
springs result from uncontaminated reservoir
water.
30m M/ It
Si 02
-20
Cr» -
>•
t-
-»• —
*.
10
Figure 2. Simple mixing model. The mixing problem is
solved by using chemical data and internal
data (in this case the solubility of silica).
A: salinity of surface waters in the area
B: salinity of mixed sample of reservoir and
surface water
C: salinity of reservoir water.
The mixing model was introduced to calculate
reservoir temperature (vertical arrow).
149
-------
coming from great depth does not mix with cooler, shallow ground
water. The test to establish whether the simplest case of direct
sampling holds is provided, firstly, by the internal consistency
between the different chemical thermometers; secondly, from
other lines of evidence that the fluid does indeed represent
reservoir water (for example, by virtue of its similarity or
difference from other types of water in the area), and thirdly,
by evidence from totally independent data such as geophysical
data which suggests that a high temperature system does occur in
the same area.
Case 2: A Simple Mixing Model (Figures 2 through 4)
Fournier et al. (1974) have suggested a methodology to cal-
culate reservoir temperature by means of measuring the chemistry
of both hot and cold springs in the area. By setting up two
equations in two unknowns, they have shown that it is possible
to estimate the fraction of original reservoir fluid
that has risen upward and has mixed with shallower ground water.
If the fundamental assumptions made by Fournier et al. hold
for a. certain set of geological conditions, it is possible to
estimate the original chemistry of the reservoir by repeating
the same calculations that they have made for the purpose
of reconstituting the original reservoir silica content or reser-
voir temperature.
A large number of chemical constituents or chemical ratios
may be employed in order to estimate original reservoir chemistry
150
-------
1.0e p m No
0. I e p m K
Figure 3. Simple mixing model. Sodium to potassium
ratio is not affected by dilution by surface
waters. The mixed sample of reservoir and
surface water does not carry relevant
information about reservoir water.
151
-------
Figure 4. Simple mixing model. One more non-linear
relationship of chemical equilibrium.
A: salinity of surface waters in the area
B: salinity of mixed sample of reservoir
and surface water
C: salinity of reservoir water.
As opposed to the silica graph, the reservoir
temperature is not estimated from this graph.
152
-------
or the amount of mixing of two different water systems. The
sodium-potassium thermometer does not allow the calculation of
the amount of dilution by mixing with fresh ground water, be-
cause the ratio does not change when dilution with pure or
nearly pure water takes place. The silica method does allow
the calculation of the amount of dilution, because the solubil-
ity curve of silica is non-linear. On the other hand, if we
have been successful at estimating the original temperature res-
ervoir and, additionally, we have information about the calcium-
potassium ratio, we are in the position of establishing the
degree of dilution because the Ca-K ratio is the ratio of the
square root of calcium over potassium and as such is affected
by dilution or mixing.
Case 3: The Complex Mixing Model (Figures 5 through 7)
In many situations, the conditions which are required by
the simple mixing model described above are not fulfilled. In
such cases, we have a complex mixing model and external con-
straints would be necessary to solve the mixing problem. Cases
where the simple mixing model may not apply are: (1) when the
spread of silica content in surface water is exceedingly large
or when the percentage of reservoir water mixing with surface
water is very small; (2) a decision cannot be made as to
which of the different silica solubility curves applies in the
case under consideration; (3) the use of the Ca-K relationship
is inapplicable because of precipitation of the calcium as
153
-------
Ul
LONGITUDINAL RESISTIVITY TO 1000 ft. SUBSURFACE
/~9 BASEMENT *'3*X RESIST I VI T Y F • FX TRAPOI A TED
S/ OUTCROP T.S -' """ ""roe E - EXTRAPOLAItU
DATA COLLECTED AND INTERPRETED BY
TSVI MEIDAV AND ROBERT FURGERSON
UNIVERSITY OF CALIFORNIA, RIVERSIDE
MARCH, 1971 REVISF.D SEPTEMBER, 19 ?l
Figure 5.
-------
WEST BRA
14 1
0
h-
LLJ
UJ 500
u.
UJ
^ 1000
O
CO
x 1500
h-
Q.
g 2000
2500
*
-
"
.
? '
i'e'
IT
07
WLEY CALIPATRIA FAULT E/^T
3 4
i BRAWLEY FAULT ,
* I * '
J1-M&) i!!! il
I 0.7
i: :|©iS:
,X:y
-••. •. '7^*;
':-••- •••.'•'. •/•'.V-.':"^
HfcS
.
'••- .•'.••'.-'i
•x^XvX-:-:;:-
!*!'I'XvX*X*X
;:::::*.^ .9 iv!':'
I'XvXvXvX
. !"I'X'X'X"X'X
^v^^'V;;-K;'^v^o''-;^X;XvXvXv
; '••^•^ • ' W\;-;-;';; ^-- ':- '.^XvXv-X'X';
liiil
1- < 1 5
5 6 16 I/
1 J. . . •»•!'* • 1 l~l
"fop
a»»4::f
•.•.".'."."•"•"•'•^».'>;vx /v 3 0 '••"•
: ::::::::;;;
5.fr..9.;;i
: :::::::::::
i;^:j:^::;:;;;;;ii;
::::::*.:::::::::::!::::
*•*»*«
8.0.
.......
_
~
-
-
Figure 6.
-2
RESISTIVITIES IN OHM-METERS
24 6 8 10 12 14
16
l
18
DISTANCE FROM BRAWLEY, WILES—*
Geoelectric cross-section A-A*, Brawley Traverse,
r
ESTIMATED SESERVOIR TEMPERATURE
FROM GEOTHERMAL GRADIENT, GEOCHEMISTRY
RESULTING ESTIMATE
OF RESERVOIR WATER
SALINITY.
SALINITY
Figure 7.
Complex mixing model. Correlation between salinity
and temperature due to broad mixing process plus
information on reservoir temperature result in
a constraint on possible reservoir water salinity.
155
-------
calcium carbonate, which violates one of the assumptions of the
simple mixing model. Under such conditions, external constraints
would be valuable in determining the mixing models. Such con-
straints may come, for example, from electrical resistivity data.
In many geothermal surveys, an estimate has been made of reser-
voir temperature. If a correlation between salinity and tempera-
ture can be shown to exist in the given prospect area, it is then
possible to estimate the salinity of the reservoir at depth from
the salinity of the fluids which have intermixed with the
shallower reservoir.
Case 4: Statistical Approach (Figure 8)
In the absence of specific information about the reservoir
water, and where either simple mixing models are not considered
valid, or complex mixing models do not provide a straightforward
solution, a statistical analysis approach may provide a gross
estimate of reservoir chemistry.
One objective of any geochemical sampling program is to dis-
tinguish the different types of reservoir water that are preva-
lent within a given area and the boundaries of their distribution.
The results and analysis of the different types of water, combined
with other types of information such as geophysical data, thermo-
metric data, etc., are aimed at determining which of the different
types of water is representative of reservoir water. In the Lar-
derello case, the sampling of water issuing from a formation in
the region provides a statistical sampling of the composition
of the reservoir water. In
156
-------
GEOTHERMAL AREA
Figure 8.
Samples from the same reservoir formation, having
essentially the same composition as reservoir water,
occur over a vast region wherever the reservoir
formation outcrops or comes close to surface. This
forms the basis for a statistical approach to
estimating reservoir water chemistry.
300
Figure 9.
Concentration of boric acid in reservoir water can
be estimated from temperature and from concen-
tration in dry steam—graph shows available
information to date. Coarsely hatched band
may be considered consensus correlation.
157
-------
other cases, such as in The Geysers, California, the reservoir
is not well defined. There, we make inferences about the chemi-
cal composition of the reservoir water which underlies the steam
reservoir, based upon analyses of both hot and cold springs over
a very large area. Even when the reservoir itself is not well-
defined from a stratigraphic, lithological point of view, the
reservoir may be defined as an aquifer, by virtue of a constant
or uniformly varying composition of reservoir water. Systematic
sampling of deeo ground water over a large area would provide an
inference regarding the original composition of the geothermal
system before various other factors such as outflow, concentra-
tion, and heat have changed it from its original state. For
example, The Geysers field is surrounded by sodium-chloride-type
water which would give rise to the inference that the deeper
water in The Geysers reservoir is also of the NaCl tvr»e. In the
case of a vapor-dominated system such as The Geysers, uncertainty
remains as to the actual composition of the liquid phase of the
reservoir because of possible changes in composition that may
have taken place as a result of evaporation and heat effects,
Some of these uncertainties may be resolved by other methods
such as those described below under Case 5.
Case 5: Inferences from Physical-Chemical Relations (Figure 9)
The content of different substances in dry steam depends
primarily on the chemical constitution of the reservoir water.
We may cite an example of this type: the content of boric
158
-------
acid in the steam is directly proportional to the content of
boric acid in the reservoir fluids, where the proportionality
constant is dependent upon temperature. By corollary, if we
have sufficient statistics of the boron/chloride ratio in the
characteristic reservoir water of the particular geothermal
system in the area, we can then estimate the chloride content
of the geothermal brine, and therefore of the sodium chloride
of the brine. The same reasoning may be applied to the relation-
ship between boron and other chemical constitutents. It might be
difficult to employ boric acid as an indicator under conditions
where condensation of steam may take place before the steam has
reached the ground surface. In such a case, other indicators,
such as ammonia content, may be valuable for determining reservoir
composition, utilizing the same reasoning as above.
In summary, we have attempted to indicate a number of tech-
niques by which the pollutant characteristics of geothermal
systems may be estimated from geochemical surface investigations
of both liquids and gases. In any of the above cases, the as-
sumptions and calculations that may be made are based upon con-
ceptual models. The veracity of estimates depends to a large
extent on the accuracy of any one model. Quite naturally,
initial estimates on reservoir composition would improve with
the continued exploration in a given area, where more and more
representative samples of reservoir fluids would be obtained
in the course of increasingly deeper drilling.
159
-------
ABSTRACT
SAMPLING AND ANALYSIS OF GEOTHERMAL BRINES
FROM NILAND FIELD, CALIFORNIA
H. K. Bishop
San Diego Gas & Electric Company, San Diego, California
The Salton Sea Geothermal Field which lies just south of the
Salton Sea in the Imperial Valley of Southern California produ-
ces a highly saline brine Development of this resource requires
that the properties of the geothermal fluids be fully understood.
Currently, San Diego Gas & Electric (SDG&E) in cooperation with
the U.S. Energy Research & Development Administration has con-
structed and is operating a 10-MW Geothermal Loop Test Facility
(GLEF) which utilizes geothermal fluid from the Salton Sea Field.
This paper discusses SDG&E's experience in sampling and analyzing
the geothermal fluid flowing through the GLEF.
160
-------
SAMPLING & ANALYSIS OF GEOTHERMAL BRINES
FROM NILAND FIELD, CALIFORNIA
Introduction
The San Diego Gas & Electric Company/U.S. Energy
Research and Development Administration Geothermal Loop
Experimental Facility (GLEF) located near Niland in the
Imperial Valley, California, has operated on a high salinity
(200,000 ppm) geothermal fluid from the Salton Sea field
since May, 1976. The conversion process being tested is a
four-stage/binary system, Figure 1. The geothermal fluid is
flashed at successively lower pressures in open drum-type
separators. The steam generated passes through a scrubber
and into a tube and shell heat exchanger where the heat is
transferred to a secondary, working fluid. The working
fluid is condensed after expansion across a throttling valve
which simulates a turbine. Waste heat is removed via a
conventional spray pond. With a turbine generator, to be
installed later, this facility would be a 10 MW power plant.
Water has been used as the working (binary) fluid during
start up and initial operation. Replacement with isobutane,
which is expected to be the preferred working fluid for a
binary power plant, is planned.
The GLEF has been operated for approximately 2,000
hours with frequent interruption for inspection and facility
modification. During this period the facility has been
operated on one production well (Magmamax #1). Flow rates
were 400,000 pounds per hour, approximately half the design
capacity of the GLEF. The temperature and pressure of the
fluid entering the first stage is 370°F and 165 psia. Brine
exit temperature is 200°F. All unflashed brine is reinjected
into the reservoir.
Brine Chemistry
The geothermal fluid available from this reservoir
is a hypersaline brine containing approximately 200,000 ppm
total dissolved solids (TDS) mostly in the chloride form,
Table 1. These chlorides remain in solution during the heat
extraction process and are subsequently injected back into
the reservoir. Certain minor species, however, such as
silica, lead and iron have limited solubility and, as the
brine is cooled during the heat extraction process, they
precipitate from solution and deposit on pipe and vessel
surfaces.
161
-------
GEOTHERMAL TEST FACILITY
BY-PASS VALVE
CONDENSERS
2nd STAGE 3rd STAGE
CONDENSATE
PUMP CONDENSATE
PUMP
REINJECTION PUMP
BRINE TO
REINJECTION
(I D-
BRINE FLASH
DRUM
1=STEAM SCRUBBER
= STEAM- ISOBUTANE
EXCHANGER
q= FLOW DIRECTION
Figure 1. Geothermal test facility.
-------
TABLE I
GEOTHERMAL FLUID COMPOSITION
NILAND RESERVOIR
(MAGMA MAX NO. 1)
ELEMENT
mg/l
Sodium 40,600
Potassium 11,000
Calcium 21 400
Chlorides 128^500
Iron 315
Manganese 681
Zinc 244
Silicon 246
Barium 142
Lead 52
Strontium 440
Lithium 180
Magnesium 105
Copper 3
Ammonia 360
Total Sol ids 219,000
pH 5.3
Oxidation Reduction +25
Potential
GAS ANALYSIS
ELEMENT PERCENT
Carbon Dioxide 9&.14
Methane 0-68
Nitrogen 0.02
Oxygen N-D-
Hydrogen N-D-
Hydrogen SuIfide 0-18
*N.D. ~ Not Detected
163
-------
The principal noncondensable gas is carbon dioxide.
Small amounts (up to 30 ppm) of hydrogen sulfide are also
found in the geothermal brine. Ammonia is also present in
the geothermal brine and has a significant effect on the
brine chemistry.
Sampling
In order to document the operation of the GLEP and
to understand the complex chemistry of the geothermal brines,
the brine and steam flows throughout the GLEF are sampled
routinely. Sampling is usually accomplished using a glass
bottle and a sample cooler. The sample cooler consists of a
bundle of 20 stainless steels tubes approximately 4 inches
long inside a 2-inch steel pipe. The geothermal fluids pass
through the tubes and the cooling water around the outside
of the tubes. Generally the samples are taken from near the
pipe wall. However, for two phase flow and if suspended
solids are to be measured a probe is inserted into the pipe
to provide a more representative sample.
Noncondensable gas samples are taken from either
the well head or from the first stage heat exchanger, where
all steam from the first stage is condensed, utilizing a
stainless steel sample bomb. A very simple procedure is
used in which the bomb is vented to the atmosphere for
approximately 5 minutes to insure equilibrium conditions -are
established, then the exit valve is closed and followed by
the inlet valve trapping a sample of geothermal fluid in the
bomb.
Suspended solids develop when the solubility of
varied species within the geothermal brine is exceeded
during the heat extraction process. These are normally
determined using a sample probe and a filter. A known
amount of geothermal fluid is withdrawn from the flow with
the sample probe and passed through a 10y filter. Suspended
solids present are collected on the filter and are subsequently
weighed and analyzed. The sample procedures and results are
described elsewhere in this symposium by J. H. Hill, Lawrence
Livermore Laboratory.
Analysis
No particularly unusual mineral species appear to
be present in the geothermal brines. However, the number of
species is very large so that identification can be difficult
and interference is a real problem. Presently, two major
instruments, an atomic absorption spectrophotometer (AA) and
a gas chromatograph have been used to analyze the liquid and
164
-------
vapor phase respectively. Any solids present were put into
the solution and analyzed by AA. The method of standard
additions is normally used in the AA analysis. Other
instruments that have been used successfully include a
conductivity bridge, a pH meter and a specific ion electrode
for ammonia. Both the conductivity and pH of the fluids
were measured at room temperature.
Gas samples were analyzed almost immediately after
sampling to reduce any possible reactions in the fluid or
with the bomb surfaces. The temperature of the gas sample
was also raised to match that of the fluid during sampling.
This is particularly important in a liquid-dominated geo-
thermal reservoir where a representative sample of the
geothermal fluid will include considerable water. Thus, at
room temperature the sample bomb will be almost filled with
water which can absorb some of the gases. The noncondensable
gases measured can be different then at the temperatures of
the geothermal fluid. Returning the temperature of the bomb
to that of the geothermal fluid should provide a gas sample
representative of the operating pressure and temperature
conditions in the facility.
Certain species such as silica although difficult
to measure are important because of their contribution to
scaling in the geothermal facility. We have found
digestions in a Parr bomb followed by AA analysis to be the
most accurate method of measuring the silica present in the
geothermal brines.
Conclusions
Sampling aid analysis of geothermal fluids, parti-
cularly the hypersaline brines from the Niland reservoir is
difficult. Representative samples must be taken of hot
fluids and two phase flow. Sampling procedures must also
allow for cooling the fluids in order to conduct analysis
without markedly changing fluid properties and to allow for
deposition of scale on the sample ports and line. The
chemical specie present must be determined in the presence
of other species with potential interferring properties. All
of this effort must take place under field conditions nor-
mally some considerable distance from any well established
laboratory. Either a field laboratory must be provided or
specimens must be treated so they can be transported without
degradation due to storage to such a laboratory.
165
-------
INTERPRETATION OF ANALYTICAL RESULTS/ THERMAL AND
NONTHERMAL WATERS/ LAVA PLATEAUS REGION OF
NORTHEASTERN CALIFORNIA AND SOUTHERN OREGON
C. W. Klein and J. B. Koenig
GeothermEx, Inc.
901 Mendocino Ave.
Berkeley, CA 9^707
Abstract
Nonthermal spring and well waters of the Lava Plateaus region (60,000
km2) are characterized by low IDS (average < 150 ymho/cm) and the weight
abundance relationships Ca > Na L Mg > K and HC03 + C03 >» S04 = Cl.
Si02 varies widely, averaging about 25 to 30 mg/1 but reaching 50 to 70
mg/1 in waters derived from lacustrine sediment or glassy tuff. Evolu-
tion of C02 gas becomes more notable with increased proximity to the
volcanically active High Cascades.
Lakes occupying closed basins in this region show greatly increased TDS.
Within the anion group, HC03 + C03 » S04 = Cl is still common, but
certain waters contain SO^ and sometimes Cl in greater abundance than
the carbonate-bicarbonate ions. Major cation concentrations are highly
variable; either Na or Ca predominates, and K is usually the least
abundant.
166
-------
Thermal springs and wells show: (1) increased IDS, usually about 500
to 2,000 mg/1; (2) Si02 usually less than 100 mg/1 but up to 180 mg/1-
(3) relative cation weight abundances of Na » Ca 1 K » Mg- and
(4) relative anion abundances S04 > Cl > HC03. These relationships are
quite typical of nearly all higher temperature waters in the region.
In waters sampled at about 50°C or less, the mixes of cations and
anions tend to vary, either due to mixing of deep and hot Na-SO^-Cl
waters with shallow, cool Ca-HC03 waters or perhaps because of rapid
heating and circulation of meteoric waters along fault planes without
coming into chemical equilibrium in a deep reservoir. Thermal waters in
which chloride is the dominant anion are uncommon throughout the Lava
Plateaus region.
Maximum reservoir temperatures indicated by Si02 content and an assump-
tion of quartz equilibrium reach about l40°C (= 100 mg/1 Si02) to 160°C.
Na-K-Ca temperatures tend to be lower, perhaps due to continued rock-
water interactions during ascent of fluids or because Si02 reflects
equilibrium with chalcedony or amorphous Si02. Na-K-Ca temperatures of
mildly thermal and cool waters of the region occasionally are higher
than those calculated for the thermal waters.
The uniformity of results across this wide region is notable, as is the
relationship of most thermal waters to Quaternary (or Late Tertiary)
faults. The data suggest that heating of the waters is most typically
caused by circulation to several kilometers along faults in a geothermal
gradient of 50° to 60°C/km, with attendant rock-water interaction. We
believe that little or no mixing occurs in the region's hotter thermal
waters as they rise to the surface, and reservoir temperatures greater
than 160°C seem uncommon. If waters exist stored at deeper levels and
higher temperatures, they have not been sampled and their composition
remains unknown, but saline, Cl-rich compositions may be most likely.
Introduct ion
Since 1970 GeothermEx, Inc., of Berkeley, California, has carried out on
the behalf of numerous clients reconnaissance level and detailed explora-
tion programs in what we will here name the Lava Plateaus region of
northeastern California and southern Oregon. This work .has included
sampling of over AOO hot and cold springs and wells. The results of
these analyses, along with some 200 data points published mostly by the
USGS, have given us a fairly thorough picture of the major element
chemistry of the region's thermal and nonthermal ground waters. It is
our intention here to outline some of these observations. Hopefully,^
they will be useful to others working in the area and to persons seeking
167
-------
to define the possible effect of its geothermal development or regional
trends in the character of thermal waters.
The region comprises 60,000 km2 of volcanic terrane (Figure 1). It is
mostly a portion of the Basin and Range structural and geomorphic
province, characterized by north- to northwest-trending tilted fault
blocks, bounded by steeply-dipping normal faults. The exposed strati-
graphic section consists of Tertiary and Quaternary volcanic and
sedimentary rocks. The volcanic rocks are basaltic to rhyolitic in
composition. Basement, lying at probably 7,000 to over 12,000 feet, is
doubtlessly of Mesozoic age. At the west fringe of the area lies the
Cascade Range, including in California the areas of Mount Lassen and the
Medicine Lake Highland, which have had silicic eruptions in the last
10,000 years.
In the east and northeast, surface hydrology is characterized by internal
drainage to shallow, ephemeral lakes. Rivers drain portions of the west.
However, the lack of surface waters in some areas is notable. Over many
hundreds of square kilometers north of Mount Lassen, for example, there
is nearly no surface water to be found. All precipitation percolates
down through highly permeable basalt flows, issuing as springs at flow
fronts and along fault scarps, sometimes after many miles of travel and
in truly astounding volumes.
Average precipitation on the Plateaus region ranges from as little as 8
inches in basins to over 30 inches on higher mountains. The Cascade
Range catches most moisture which might otherwise fall on the plateaus,
with 50 to 70 inches of annual precipitation falling at higher altitudes.
Ground water is abundant at shallow levels beneath basins, range front
fans and landslides, and at variable levels beneath highlands composed
of PIio-Pleistocene, highly permeable basalts. These rubbly rocks can
act as unconfined aquifers, beneath which the water table is nearly its
level in adjacent lake basins. However, sections of older,
more weathered basalt and tuff and diatom!te are, of course, relatively
impermeable. In some basin areas, artesian conditions are common.
Principal thermal springs and wells typically issue along range-front
faults or are believed to be associated with major faults, on the basis
of geologic mapping and geophysical studies. Deep circulation along
faults is thus the probable mechanism of heating except at Mount Lassen
and the Medicine Lake Highland, where magmatic heat sources undoubtedly
exist. No evidence for shallow magmatic sources elsewhere has thus far
been encountered.
168
-------
LIMITOF CENO-
ZOIC VOLCANIC
ROCKS
KLAMATH FALLS /
MEDICINE LAK
CALDERA —/
^
\
;
/
NEVADA
IDAHO
MAJOR FAULT
YOUNG CASCADE VOLCANO
Figure 1. Sketch map showing the lava plateaus region,
169
-------
Hydrochemistry
In discussing the ground waters, chemical types will here be illustrated
according to relative concentrations on a weight basis in a system
employed by Truesdell in the Proceedings of the 1975 UN Geothermal Sym-
pos i urn:
Water Type
A = B — A approximately equals B in weight concentration
A > B --- A is 1 to 1.2 times B
A > B --- A is 1.2 to 3 times B
A » B -- A is 3 to 10 times B
A >» B - A is more than 10 times B
Cold springs, upland well waters, and some basin wells of the region are
characterized by low total dissolved solids (IDS) and specific conduct-
ance (SpC). IDS averages less than 150 mg/1, specific conductance less
than 150 micromhos. Weight abundance relationships are typically:
Ca > Na 1 Mg > K and HC03 + C03 >» S04 = Cl
C03 is rarely present as pH is usually near 7. Cl and SO^ are each less
than 2 mg/1, and HC03 is about 30 to 100 mg. With rare exceptions Ca is
the most and K the least concentrated cation. Na is usually more con-
centrated than Mg, but there are exceptions where these two ions are
about equally concentrated, and occasionally Mg is the more abundant.
Si02 varies widely, averaging about 25 to 30 mg/1 but reaching 50 to 70.
These higher levels are sometimes seen in nonthermal well waters derived
from lacustrine sediments or glassy tuff, where silica solubility may be
controlled by amorphous silica more than quartz. Very rarely, we have
seen Si02 as high as about 120 mg/1 in cold springs far from known
thermal areas and which otherwise display no traces of possible thermal
effects or other anomalies. Where we have applied the Na-K-Ca geother-
mometer to the cold waters, the resulting temperatures calculated are
usually low, as would be expected, hbwever, about cnewater sample in 15 or
20 gives a temperature which is abnormally high. We suspect that this
is caused by circulation of such waters through rocks or mineral zones
from which potassium is easily leached.
170
-------
Ground waters in the lake basins commonly show slightly to greatly
increased IDS, to over 1,000 mg/1. The types of water become somewhat
variable; either Na or Ca predominates the cations, and K is usually
the least abundant. Within the anion group, bicarbonate is still the
most abundant species, and rough equivalence of S04 and Cl is still
common. However, certain waters contain SO^ and sometimes Cl in greater
abundance than the carbonate-bicarbonate ions. These variable composi-
tions doubtlessly reflect mixing of such factors as changing sediment
composition, local concentrations of somewhat saline pore fluids, and
lateral migration of meteoric waters into the basin aquifers.
Thermal springs and wells are of fairly uniform character throughout
the region. Typically:
Na » to >» Ca ^_ K » Mg and
SO^ > to » Cl > to » HC03
TDS are usually about 500 to 2,000 mg/1. Amongst the cations, Na is
characteristically dominant. Ca is greater than K and Mg the least
concentrated. Although usually Ca ^_ K, the relative abundancies vary
from ;L to », and between K and Mg we have seen K > Mg to K >» Mg.
Anion analyses most frequently show that SO^ is more concentrated than
Cl, and bicarbonate is the least abundant. These relationships are
quite typical of nearly all the higher temperature waters. At about
50°C or less, the mixes of cations and anions tend to vary. TDS are
low, and we occasionally see Ca > Na and HC03 as the dominant anion.
This is perhaps due to mixing of deep and hot Na-SO^-Cl waters with
shallow, cool Ca-HC03 waters or because of rapid heating and circulation
of meteoric waters along fault planes without coming into equilibrium
in a deep reservoir. Thermal waters in which chloride is the most
abundant anion are very uncommon throughout the region.
Si02 in the thermal waters is usually less than 100 mg/1 but occasion-
ally is seen to 180 mg in springs above about 80°C. The 100-mg level
indicates a quartz equilibrium temperature of 140°C. Na-K-Ca tempera-
tures tend to fall below the quartz temperatures by about 10° or 20°C,
perhaps due to reequi1ibration upon cooling; some cation temperatures
are higher and some are a good deal lower than the quartz values. This
suggests possible equilibrium with Cristobalite or chalcedony, as has
been pointed out also by Mariner and others and by Reed in studies of
parts of the same region.
171
-------
Interpretations
The uniformity of hot spring compositions across this wide region is
notable, we think; in particular, as they are usually characterized by
relatively low total dissolved solids compared with hot springs world-
wide and dominance of the anions by S04 instead of Cl or HC03. If we
move westward into the Cascades either in Oregon or at Mount Lassen in
California, higher IDS to over ^,000 mg/1 are common, and Cl is very
dominant. In the Basin and Range province in Nevada, reported IDS tend
to be near or above 2,000 mg/1, with HC03 or Cl the most abundant anion.
Relative absence of chloride and low IDS in the hot springs may be caused
by several conditions, none of which are certain. These include: low
aquifer temperatures, short residence time, the effects of rock compo-
sition, and dilution of deep, very hot chlorided waters by meteoric
waters during travel to the surface.
Although the presence of deep, hot, chloride-rich aquifers can be imag-
ined, evidence of their existence has not been seen, either in surface
waters or in the few deep tests which have been drilled.
Rock composition is a factor which is hard to evaluate without petro-
chemical studies. We don't know the relative amounts of SO^ and Cl
which hot waters can leach from igneous rocks of the area. Some SOit may
originate in basin sediments, with which most of the thermal springs are
associated.
By comparison, note the correlation between relatively low chloride and
the absence of either basement rocks or andesitic volcanics from the
region. Chloride is the characteristic anion of hot springs not only in
the Cascades but also in the andesitic chains of Central America. Base-
ment rocks are usually in the vicinity of hot springs in Nevada, in
which SO^ is rarely the most abundant anionic species.
Some workers have suggested that where aquifer rocks do not include
sources of a particular anion in abundance, such as evaporites, the
dominant anion of a hot spring may be related to its relative maturity
or residence time. Thus, bicarbonate waters are youthful while sulfate
waters are of intermediate and chloride waters of most advanced age.
The Plateaus hot springs would thus be classified as intermediate. The
relatively low apparent aquifer temperatures would be consistent with
this, and both could be controlled by relatively rapid circulation of
the waters to a limited depth down and back up along fault zones. Water
could heat to a reservoir temperature of 130°C by circulation to only
172
-------
^,300 m down a 30°C/km gradient. This depth is very roughly where we
expect the top of Mesozoic basement, so the waters may experience at
most a brief residence in basement rocks. Exploration experience at
some springs suggests localization of heated waters along fault planes
or at least an absence of extensive hot aquifers easily tapped by deep
drilling. However, the thermal aquifer at Klamath Falls is apparently
quite large and may store water for a considerable period of time.
We are, by the way, gratified to see that Ed Sammel and co-workers at
the USGS have reached conclusions about the Klamath Falls geothermal
system which are essentially similar to our own.
In conclusion, the uniformity of composition across this wide region
and generally good agreement between the silica and cation geothermom-
eters suggest that little or no mixing occurs In its hotter thermal
waters as they rise to the surface. Reservoir temperatures greater
than 130° to 160°C seem uncommon. We seem to be seeing in the plateaus
region an example of an extensive geothermal province of uniform
chemical character.
It may be possible to similarly outline other geothermal provinces
within which expectable ranges of effluent composition can be constrained
to limits useful in modeling the possible environmental effects of
development.
References
Mariner, R. H., et al.,1974. The chemical composition and estimated
minimum thermal reservoir temperatures of selected hot springs
in Oregon. USGS open-file report.
Reed, M., 1975. Chemistry of thermal water in selected geothermal areas
of California. Calif. Div. Oil and Gas, Report No. TR15-
Sammel, E. A., 1976. Hydrologic reconnaissance of the geothermal area
near Klamath Falls, Oregon. USGS Water-Resources Investigation
open-file report WRI76-127-
173
-------
ABSTRACT
GEOCHEMICAL ANALYSIS OF FLUIDS CIRCULATED THROUGH A
GRANITIC HOT DRY ROCK GEOTHERMAL SYSTEM
J. W. Tester, C. Grigsby, C. Holley, and L. Blatz
Los Alamos Scientific Laboratory
University of California
Los Alamos, New Mexico
Results of preliminary field tests involving circulating filtered
water through a hydraulically fractured geothermal reservoir com-
posed of granite at 200°C are reported. The reservoir is connected
to the surface by two approximately 10,000 ft (3 km)_holes, used
for injection and recovery fluid. Chemical composition of the
fluid samples was monitored at the site using spectrophotometric
and specific ion electrode techniques. Supplemental measurements
were made at our central laboratory facility using neutron acti-
vation, emission spectroscopy, and atomic absorption. The major
cations and anions in a fluid specimen that had circulated through
the fractured granitic system for a mean residence time of 10 hrs
would include:
Si02 (180-260 ppm), Na (400-750 ppm), K (400-650 ppm), Ca (2-
10 ppm), Mg (1-3 ppm), Al (<1 ppm), Fe (2-3 ppm), Mn (>0.01 ppm),
Cl (400-600 ppm), F (12-14 ppm), HCO, (150-200 ppm), PO,
(0.1-0.2 ppm), and pH (6.8-7.0). J
With fresh water injection a steady state exiting fluid composition
is reached within 10 hrs indicating a rapid rate of dissolution
mass transport between the rock surface and injected fluid. Of
the major components only the fluoride levels of 12-14 ppm present
a possible environmental hazard. Trace element analysis indicated
that levels of other toxic elements were below EPA recommended
levels for continuous irrigation or livestock feeding. Because com-
plete recirculation (or reinjection) of fluid will be primarily
used in the present hot dry rock demonstration system after the
temperature has been reduced with an air-cooled heat exchanger,
only periodic treatment of a part of the water may be required to
remove fluoride. Organic dyes have been periodically injected into
the system for tracer analysis of residence time distribution. Both
activated charcoal and activated alumina have been investigated in
field and laboratory experiments to determine their fluoride and
dye removal efficiencies. Experiments showed that fluoride levels
would be reduced to <0.1 ppm with a properly activated alumina bed.
Silica was also removed quite effectively. As expected, the acti-
vated charcoal removed up to 20 ppm of dye to below the detection
level of our spectrophotometric technique (0.05 ppm).
174
-------
ABSTRACT
CHEMICAL PROFILE OF THE EAST MESA GEOTHERMAL
FIELD,IMPERIAL COUHTY. CALIFORNIA
R. T. Littleton and,E. Burnett
U.S. Bureau of Reclamation, Boulder City, Nevada
The East Mesa Geothermal Field, about 18 miles east of El Centro,
California, is a thermal knot in saturated deltaic and lacus-
trine sediments. A deep convective system drives geothennal
fluids through a reservoir of interconnected lenses of sand and
sandstone throughout a depth interval of about 4,500 feet be-
tween about 7,500 and 3,000 feet.
A gross evaluation of the geochemical profile of the reservoir
is attempted by interpretation of the resistivity and self-
potential curves of the dual induction log. The geochemical
profile shows a typical pattern throughout the reservoir. Typ-
ically, salinity increases with depth reaching a maximum be-
tween 2,000 and 3,000 feet, then decreasing to about 6,000 feet,
and then increasing sharply to a depth of about 7,500 feet. Be-
low 7,500 feet we are unable to compute salinities with confi-.
dence and our reasons are discussed.
The shape and dimensions of the geothermal field in relation to
depth are discussed. Chemical profiles are computed from dual
induction logs of 10 geothernal wells in and around the East Mesa
Geothermal Field and related to laboratory analysis of fluids re-
covered from drill stem tests and produced at the wellhead. The
relation of the geochemical profiles to the movement of geothermal
fluids in the reservoir is discussed.
175
-------
CHEMICAL PROFILE
OF THE EAST MESA FIELD
IMPERIAL COUNTY, CALIFORNIA
by
Robert T. Littleton and Earl E. Burnett
U. S. Bureau of Reclamation
Lower Colorado Region
February 1977
INTRODUCTION
The East Mesa Geothermal Field is about 18 miles east of
El Centre in Imperial Valley, California. Exploration and
evaluation of temperature gradients, gravity responses, and
resistivity measurements by Robert W. Rex and his colleagues
at the University of California at Riverside, largely supported
by Reclamation funds, led to the discovery of the field. The
dimensions of the field are beginning to emerge as the result
of drilling by Reclamation, Magma Energy Company, and Republic
Geothermal Incorporated.
This paper was prepared by the authors largely on the basis
of interpretations from the dual induction resistivity logs
of 10 geothermal deep wells and laboratory analyses of water
samples of 5 geothermal deep wells.
The authors express their appreciation to Reclamation's
Regional Director, Manuel Lopez, Jr., and Regional Planning
Officer, M. K. Fulcher for support in these studies. The
authors also appreciate the generosity of Mr. W. C. McCabe of
Magma Energy Company and Robert W. Rex of Republic Geothermal
Incorporated for releasing geophysical logs of geothermal deep
wells.
THERMAL DIMENSIONS OF THE GEOTHERMAL FIELD
The East Mesa Geothermal Field apparently is a thermal knot
in saturated deltaic and lacustrine sediments. A deep
convective system drives geothermal fluids through a
reservoir of interconnected and faulted lenses of sand and
sandstone throughout a depth interval between 3,000 and
7,500 feet. About 70 percent of this interval is sand and
sandstone creating a reservoir of roughly 3,150 feet of
176
-------
lenticular permeable beds interconnected to various degrees
due partly to stratigraphic interfingering but largely by '
transecting faults.
Isothermal contouring based upon equilibrium temperature
profiles in 10 geothermal deep wells (Figures I through 4 )
suggest the dimensions of the geothermal reservoir. When
considered together, the four different isothermal maps
indicate the shape of a geothermal plume rising into a cold
ground water reservoir. The plume is bent north-northwest
from an apparent geothermal center in about the NEJ of sec.
7, T. 16 S., R. 17 E., SBM, following avenues of permeability
established by the stratigraphic makeup of the reservoir but
also by transecting faults.
GEOCHEMICAL PROFILE OF THE RESERVOIR
General Considerations
A gross evaluation of the chemical profile of the reservoir
is made by interpretation of the resistivity and self-potential
curves of the dual induction logs. The chemical profiles
are shown as total dissolved solids which is estimated from
the computed electrical conductivity. Chemical profiles were
computed for 10 geothermal deep wells (see Figures 5 through 7)
The method used to calculate the electrical conductivity of
fluid in the reservoir is fairly well known. It is based
upon formulae and methodology discussed in text books such
as Sylvain T. Pirson's Handbook of Well Log Analysis (Prentice
Hall, 1963) and Service Company Manuals by Schlumberger Well
Surveying Corporation. The chemical profile computed is for
salinity only, as indicated by the concentrations of sodium
and chloride ions.
Computer Program
The computer program has been developed to calculate the
electrical conductivity of formation water from a suitable
set of resistivity and self-potential curves. The methodology
requires an accurate measure of the mud filtrate resistivity
at 77°F. To convert the mud filtrate resistivity (Rmft-^ at
some temperature t± to 77°, we use the following formula:
Rmf77 = (Rnift1) (t±/77)
where t1 = fahrenheit temperature
177
-------
EXPLANATION
y Geothermal W^ll
3- Wildcat Well For Oil & Gas Or Geothermal
XW— Elevation Of Isotherm (All Elevation Values
and Contours Shown Are Minus Sea Level)
NOTE Compilation and Interpretation
by Robert T Littleton & Earl E
U S Bureau of Reclamation
Lower Colorado Region
EAST MESA GEOTHERMAL FIELD
Imperial County, California
FEBRUARY 1977
Figure 1. 150°F isotherm subsurface contour map.
178
-------
EXPLANATION
127
O Heat Flow Hole
® Geothermal Well
I
-O- Wildcat Wei! For Oil & Gas Or Gsotheimal
-l-SMat~ Elevation Of Isotherm (All Elevation Values
and Contours Shown Are Minus Sea Level'
NOTE. Compilation and Interpretation
by Robert T Littleton & Earl E I
U S. Bureau of Reclamation
Lower Colorado Region
EAST MESA GEOTHERMAL FIELD
Imperial County, California
FEBRUARY 1977
Figure 2. 250°F isotherm subsurface contour map.
179
-------
•PSTKOFtMA
EXPLANATION
Geothermal Well
-O- Wildcat Well For Oil & Gas Or Geothermal
••fffd'^ Elevation Of Isotherm {All Elevatioc Values
and Contours Shown Are Minus Sea Level ?
NOTE Compilation and Interpretation
by Robert T Littleton & Earl E I
U S Bureau o! Reclamation
Lower Colorado Region
EAST MESA GEOTHERMAL FIELD
Imperial County, California
FEBRUARY 1977
Figure 3. 300°F isotherm subsurface contour map.
180
-------
EXPLANATION
ffl
O Heat Flow Hole
® Geotherrr.aJ Well
-O- Wildcat Well For Oil & Gas Or Geotherma]
I
-l-tUCI-~ Elevation Of Isotherm (All Elevation Values
and Contours Shown Are Minus Sea Lavsl )
NOTE ComFilation and Interpretation
by Robert T Littleton & Earl E Burnett
U S Bureau of Reclamation
Lower Colorado Region
EAST MESA GEOTHERMAL FIELD
Imperial County, California
FEBRUARY 1977
Figure 4. 350°F isotherm subsurface contour map.
181
-------
CO
to
PARTS PER MILLION
WELL NO. 44-7
I*
£4
ui
0
0 1000 2000 3000 4000 SOOO
PARTS PER MILLION
WELL NO. 8-1
1000 2000 3000 4000 5000 60DD 7000 GOOD 9DOO 1OOOO
1000 2000 3000 4000 5001
10000 11000 12000
PARTS PER MILLION
WELL NO. 48-7
NOTE: Compilation and Interpretation by:
Robert T. Littleton and Earl E. Burnett
U.S. Bureau of Reclamation
Lower Colorado Region
Figure 5. Mesa anomaly. Chemical profile of geothermal
wells, (continued)
-------
00
U)
1000 2000 3000 4000 50OO 6OOO 7OOO qOOO 9OOO
PARTS PER MILLION
WELL NO. 6-8
NOTE:
1000 2000 3000 4000 5000 6OOO 7000 80OO
PARTS PER MILLION
WELL NO. 6-1
12000 13000 14000 15DOO 1600O 17OOO 13000
1000 2OOO 3DOO 4ODO SOOO 6QOQ 700O
PARTS PER MILLION
WELL NO. 31-1
Compilation and Interpretation by:
Robert T. Littleton and Earl E. Burnett
U.S. Bureau of Reclamation
Lower Colorado Region
Figure 6. Mesa anomaly. Chemical profile of geothermal
wells, (continued)
-------
CO
U.
Z
m 5000
Q
J.LL
1000 2000 3000
5000 6000 7000 8000 9OQO
PARTS PER MILLION
WELL NO. 38-30
1000 2000 3000
saao eooo 7000
PARTS PER MILLION
WELL NO. 16-29
I 3000 -
3500 -
H..OOL
5000 9000 7000
PARTS PER MILLION
WELL NO. 5-1
1000 2000 3DOO 4000 5000
TOaa 3000 9000
PARTS PER MILLION
WELL NO. 18-28
NOTE: Compilation and Interpretation by:
Robert T. Littleton and Earl E. Burnett
U.S. Bureau of Reclamation
Lower Colorado Region
Figure 7. Mesa anomaly. Chemical profile of geothermal wells.
-------
Formation temperature at a given depth _is required to compute
salinity from the SP curve but is not required to compute
salinity from resistivity curves. A measure of geothermal
gradient existing during logging is used to estimate formation
temperature at any given depth. The following formula is used:
tf = tg + (G) (D)
where t„ = formation temperature in F
t = temperature of drilling fluid near
s surface in °F
G =
(BET - SURF. Temp)
(TD - SURF. Depth)
Geothermal gradient as
indicated by the bottom
hole temperature (BHT)
when logged and the surface temperature.
Surface temperature should be measured about
50 feet below land surface to avoid surface
cooling.
D = depth of formation in feet.
TD = total depth of hole.
Values of resistivity are taken from two curves representing
the mud-invaded zone resistivity (shallow induction curve)
and the uninvaded zone resistivity (deep induction curve).
SP values are a. measure of static SP deflection from the
shale line. The computer program described in this paper
was devised for use in a Texas Instruments SR-52 card pro-
grammable calculator. The program can be adjusted to other
programmable calculators. A FORTRAN version also exists.
The SR-52 program is given in Appendix 1.
The methodology used here can give misleading results due to
invasion of mud filtrate water into the permeable formation.
The authors believe that such invasion is minimal in saturated
formations having solely or primarily intergranular per-
meability. Laboratory analysis of fluid recovered on drill
stem tests independently corroborate the mineral concentration
obtained from the resistivity curves of the log. The depth
and total dissolved solids of the five Bureau of Reclamation
wells are shown by a "+" on Figures 5 and 6 . Two lines
of evidence support the contention that formation water is
recovered. Near equilibrium formation temperature is
achieved during a short test indicating considerable
flushing. Further corroboration is by several thousand
185
-------
feet of water column trapped in the drill pipe within a
matter of 15 minutes. After about 15 minutes the tool
plugs with sand and the test is essentially over.
Conversely reliable results are not obtained in the reservoir
with substantial or total fracture permeability. Mud invasion
is apparently excessive and deep and the program measures the
salinity of the mud primarily.
Characteristics of the Chemical Profile
The general shape of the chemical profile is depicted in
Figures 5 and 6 by drawing a heavy line more or less along
the trend of resistivity values for each permeable bed. Wells
in the central part of the field, i.e., 6-1, 6-2, 44-7, and
48-7 show a broad buildup of chemical concentration between
depths of 1,500 and 3,500 feet. The critical reversal points
are usually between 2,000 and 3,000 feet depth. A reversal
from lower to higher mineralization occurs in most of the
wells at depths between 5,700 and 6,900 feet except in
Well 38-30 where the depth is about 4,700 feet.
CONCLUSIONS
With adequate dual induction logs, it is possible to estimate
the chemical concentration in the intergranular reservoir
of the East Mesa Geothermal Field. The chemical concentration
in the fractured reservoir cannot be reliably determined
because of deep invasion by the drilling mud.
REFERENCES
Pirson, S.T., 1963, Handbook of Well Log Analysis. Prentice
Hall, New York, New York.
Schlumberger Well Surveying Corporation, 1972, Log Interpretation
Volume I - Principles. Schlumberger Limited, 277 Park
Avenue, New York, New York.
Schlumberger Well Surveying Corporation, 1974, Log Interpretation
Volume II - Application. Schlumberger Limited, 277 Park
Avenue, New York, New York.
186
-------
APPENDIX 1
An SR-52 Program for Computing Water Salinity
Concentrations from Resistivity and Self-Potential Logs
Memory Register Initialization
R01 Mud Filtrate Resistivity at 77°F (ohm-meters)
R02 6000, a constant
R05 2.15, a constant
R06 Drilling mud temperature near land surface, °F
R08 Geothermal Gradient, °F/foot
R09 0.62, empirical constant to convert electrical
conductivity to ppm.
R13 70.7, a constant
R14 11764.74, a constant; may be adjusted for a particular
ground water reservoir
Other Memory Used
ROO F, formation factor
RIO 0, Humble formula porosity
Rll m, cementation factor
R12 ppm
Instructions
Initialize memory registers, press CLR, press A, key in depth
in feet, press RUN, key in the invaded zo?e re!"*i;ltJj.eS '
for that depth, press RUN, read F, press RUN, read 0 PJ?"
RUN, read m" press RUN, key in the uninvaded zone resistivity,
Ho, for that depth, press RUN, read resistivity log ppm.
187
-------
For the SP log water quality, press B, key in depth, press RUN,
key in static SP deflection, SSP, in negative millivolts, press
RUN, read the sp log ppm. For the next sand body press A and
repeat the process. If no SP log is available, do not press B
but proceed to a new depth by first pressing A and continue as
above. If only an SP log is available, begin by first pressing
CLR, then press B and continue as above for the SP log and
proceed to a new depth by again pressing B, etc. In other words,
LABEL A is for resistivity logs and LABEL B is for SP logs and
work independently of one another. Of course, the pertinent
memory initialization (for RMFyy and so forth) must be made for
either kind of log.
LISTING:
000
001
002
010
020
020
030
*LBL
A
RC1
0
8
X
HLT key in depth
+
RC1
0
6
= formation temp,
STO 040
0
4
X
HLT key in Rxo
7
7
RC1
0
1
STO
0
0
050
Entry
Step
HALT display F
v 050
RC1
0
9
*log 060
RC1
0
5
INV
*log
HALT display 0
STO
1 070
0
*log
*l/x
X
RC1
0
0
*log
Entry
HALT
display M
STO
1
1
RC1
0
4
X
HALT key
in Ro
7
7
RC1
0
0
*l/x
X
RC1
0
2
STO
188
-------
090
100
110
Entry Step
1 120
2
HALT display ppm
*LBL
B
HLT key in depth
X
RC1
0
8
+ 130
RC1
0
6
4
6
0
X
RC1
1
3
*l/x
X
5
3
7
X
HALT key in SSP,
INV
*log
X
RG1
0
1
*l/x
HLT display SP log ppm
Example: Rmfyy =1.0 Ohm-meter )
Surf. temp. = 90° F )
G = .01 F°/foot )
D = 4460 feet )
Rxo =9.0 ohm-meter )
Ro = 6.0 ohm-meter )
INPUT
SSP
F
0
m
ppm
= -15.0 millivolts)
= 15.73 )
= 0.22 )
= 1.8 )
= 9000, resistivity) OUTPUT
ppm =
log )
11339, SP log )
a neg.
#
189
-------
ABSTRACT
STATUS REPORT, RAFT RIVER PROJECT
SAMPLING, ANALYSIS, AND ENVIRONMENTAL EFFECTS STUDIES
A. C. Allen, J. M. Baldwin, R. E. McAtee
Allied Chemical Corporation, Idaho Falls, Idaho
This report describes sampling methods presently being used for
the geothermal wells and associated effluents at the Raft River
Geothermal site near Malta, Idaho. Analytical techniques used
for species difficult to accurately determine are discussed.
Also reported are the plans for the on-site laboratory under
construction; including the analytical support it will provide.
The primary criteria for development of a sampling system was
preservation of the environment of the sample source. Well
samples were maintained at the well temperature and pressure
until time of analysis or until the sample was cooled, depres-
surized, and diluted. Pressure cylinders are used to collect
samples from sample sources with temperatures above the boiling
point. Plastic bottles are used to collect from sample sources
below the boiling point.
Chemical analysis problems mostly occur in sample preparation.
Problem species discussed are S~ and F~. Sulfide ion requires
special sampling and sample preservation techniques to obtain
reliable analysis. Fluoride determinations are usually requested
for the surface and the bulk of agricultural samples. The stan-
dard bulk analysis technique is not useful for grain samples, but
ignition techniques appear promising. Chelating agents were used
to remove surface F~.
The laboratory being fabricated at the Raft River geothermal site
will furnish limited analytical chemistry support. Analytical
procedures will be limited to determining the concentration of
major components since skilled chemical technicians will not be
available. Trace element analysis and the large variety of en-
vironmental analysis will continue to be analyzed in the larger
laboratories.
190
-------
Introduction
Several techniques have evolved for sampling geothermal fluids. This
is natural since the requirements vary from site to site. For instance,
Union Oil's problem of sampling steam at the Geysers^1 ^ is significantly-
different from the USGS problem of sampling springs(2) or of sampling hy-
drothermal resources, such as found at Raft River^.
This report describes the sampling, analysis, and preservation tech-
niques used by Allied Chemical Corporation at the Idaho National Engineer-
ing Laboratory (INEL) for the geothermal wells and associated effluents
at the Raft River Geothermal Site near Malta, Idaho. It will describe
some chemical studies performed at the site. The analytical support for
experiments being performed at the geothermal site will be discussed.
Also, this report will describe the analytical facility being constructed
at the Raft River Geothermal Well #1 site, its present analytical support,
and future direction.
The Raft River Geothermal Site is located in south central Idaho near
the city of Malta. There are a number of hot water wells in the area in
addition to the three geothermal wells drilled by Aerojet Nuclear Corpora-
tion. The three wells have been drilled to depths of 4989, 5988, and
5816 ft with bottom of the hole measured temperatures of 144 , 145 , and
144°C, respectively^. Wells #1 and #2 had artesian flows of 600 and
600 gpm, respectively. Well #3 has not been tested. Also, Well #1 was
drilled at the junction of two minor fault zones. Well #2 was drilled
3000 ft from Well #1 on one of the fault zones. Well #3 was drilled about
5000 ft away from the fault zone. The purpose of locating the wells this
way was to assist in the geothermal water reservoir study. These wells
are along the border of a large irrigation district. This would place a
large cold water reservoir in the same area as the geothermal wells. If
191
-------
this cold water reservoir is the water source for the geothermal water,
it could mean an almost unlimited source of geothermal water. It is
this set of circumstances that appears to exist in many areas of Idaho's
Snake River Plain that has caused such interest by geothermal developers.
It would appear that this could be one of Idaho's major energy sources.
Sampling
The sampling technique used for hot pressurized fluid systems, such
as the geothermal wells, was developed to preserve the sample environment
as nearly as possible. Temperature change and interraction between the
sample and the sample vessel walls are two variables encountered in stream,
spring, and hot spring sampling. Pressure changes can cause chemical
changes in samples from below the surface or from wells held under pressure.
In order to minimize these problems, a sampling system was developed that
would maintain pressure and temperature when the water sample was trans-
ferred to the sampler and transported to the place of chemical analysis.
The system also includes a means to quench the sample by simultaneous
cooling and diluting. This should be carried out just before analysis,
but can be done prior to transit to an analytical laboratory.
The sampling system consists of three parts; the sample vessels, the
storage oven, and the desampler. The sampler is a 304 stainless steel
sample cylinder equipped with stainless steel valves and purge line, as
shown in Figures 1 and 2. The equipment is similar to the evacuated cylin-
ders described by A. Stoker^ '. Some samplers are equipped with a stain-
less steel in-line filter holder. This unit used a 25mm diameter, 5-micron
Teflon filter. Gas samples are collected in these samplers, but are not
kept at well temperature. The oven is constructed for rugged use, as
shown in Figures 3 and 4. It is a stainless steel box with three inches
192
-------
Figure 1. Geothermal high pressure sampler.
Figure 2. Geothermal high pressure sampler with temperature
readout.
193
-------
Figure 3. Sample storage oven,
Figure 4. Sample storage oven,
194
-------
of Marinite insulation, a stainless steel liner, and a 1100-watt heater.
The desampler, shown in Figure 5, is used to cool and dilute hot pressur-
ized geothermal fluid while keeping it at a pressure that will not allow
phase separation. A nonreactive gas, such as helium or nitrogen, is used
to pressurize the system. It can also be used to purge air or other reac-
tive gases from the sample.
A study was made to determine the best approach to sampling and sample
care in a pressurized fluid system. An evaluation of the following methods
was made:
1) A pressurized sample kept at wellhead temperature and pressure
until analysis time;
2) A pressurized sample treated with HCL and kept at wellhead tempera-
ture and pressure until analysis time;
3) A pressurized sample treated with HF and kept at wellhead tempera-
ture and pressure until analysis time;
4) A pressurized sample treated with Na.OH and kept at wellhead
temperature and pressure until analysis time;
5) A pressurized sample simultaneously cooled and diluted under pres-
sure as quickly as possible after collection and stored in a
plastic bottle until analysis time.
The results indicated that the reagent-treated samples caused inter-
ference problems and the analysis dataware inconsistent.Whereby the
samples kept at wellhead temperature and pressure and the samples cooled
and diluted under pressure created few analysis problems and the data were
consistent. However, it was noted that the sample stored at wellhead
temperature until analysis did have more precipitate than the sample that
was cooled and di-uted under pressure and stored in a plastic bottle.
Qualitative analysis of the filter solids collected, when the samples were
195
-------
OILUTION
COOLIMfl
VESSEL
vo
THERMOCOUPLE
-D—
HELIUM SOUftCE
©
Figure 5. Desampling apparatus.
RRGE I HISTORY
900
700
900
900
140
3-100
leo
1-1-76
TIME
Figure 6. Raft River geothermal
well # 1 chemical
history.
-------
filtered at the wellhead, shows Fe to be the major constituent. When the
sample was not filtered and stored at wellhead temperature and then fil-
tered, the major filter solids were Fe and Si02. Because of possible prob-
lems with Si02 losses, the method requiring immediate cooling and diluting
under pressure was adopted.
Geothermal Site
Samples from sources with temperatures below the boiling point are
collected in plastic and glass bottles. The sample bottle material is
selected according to the chemical species to be analyzed. Samples col-
lected from water sources other than geothermal sources are treated ac-
cording to the procedures described by E. B. Brown, M. W. Skougstad, and
M. J. Fishman in Book 5, Chapter A-l of the test, "Techniques of Water
Resources Investigations of the United States Geologic Survey"^ .
Analysis
The problems encountered in the analysis of this geothermal water are
typical of conventional water analysis. The geothermal water from the
wells near Malta, Idaho is very low'in dissolved solids compared to geo-
thermal water from other areas. This results in reduced interferences
during analysis. Consequently, for most chemical species, only small
changes in existing methods of analysis were necessary. Table I lists
the species, the method of analysis, and the method sensitivities.
Table II lists some of the purposes for the analysis of a particular
species. This is only a partial listing showing some of the uses the chemi-
cal analysis, but it is apparent different activities require different
sets of chemical knowledge. The most comprehensive demand for chemical
information comes from environmental sources. Not shown in Table II, are
197
-------
TABLE I. ANALYTICAL METHODS USED FOR GEOTHERMAL FLUIDS,
ANALYTICAL METHODS USED FOR GEOTHERMAL FLUIDS
SPECIES
HCO3~
cos-
TOTAL P
Si(OH)4
I-
cr
F-
so4-
No+
SI
LI+
Sr»
K*
H,2 +
S2*
NH4*
GASES
ANALYTICAL- METHOD
TITRATION (AUTO)
TITRATION (AUTO)
COLORIMETRY
COLORIMETRY
ION SENSITIVE ELECTRODE
ION SENSITIVE ELECTRODE
ION SENSITIVE ELECTRODE
COLORIMETRY
EMISSION SPECTROGRAPHY
ATOMIC ABSORPTION SPECTROSCOPY
EMISSION SPECTRO6RAPHY
ATOMIC ABSORPTION SPECTROSCOPY
ATOMIC ABSORPTION SPECTROSCOPY
ATOMIC ABSORPTION SPECTROSCOPY
ION SENSITIVE ELECTRODE
ION SENSITIVE ELECTRODE
ION SENSITIVE ELECTRODE
MASS SPECTROMETRY
SENSITIVITY
GRAMS
-7
IO"7
id"9
IO-6
IO8
__
IO'7
IO'8
I07
IO'9
IO'7
IO9
IO9
IO9
IO'9
IO'9
IO7
IOIZ
TABLE II. CHEMICAL INFORMATION REQUIRED BY DIFFERENT
GEOTHERMAL DEVELOPMENT ACTIVITIES.
Li Na K Mg Ca Sr Hg As SiO^ F- Cl
HCUa= SOU= CO? 07
Exploration x x
Drilling x x
Reservoir Eng. x x x x
Equip. Design
Environmental x x x x
Reinjection
Research x x x x
x x
x
x x
x
XXX
x
XXX
x
x
x
x
x x
x
x x
XX X
X
XX X
X X X X
X X X X
X
X X X X
XXX
XXX
X
XXX
the frequency, speed of analysis, and accuracy required. These also vary
from activity to activity. For instance, during well drilling only a few
species need to be analyzed, but requests are likely to be frequent and
rapid analyses are essential. Results for environmental purposes usually
are not required quickly, but must be accurate with reliable statistics.
198
-------
Sample Preservation Techniques
Sulfide ion determination is a problem analysis requiring special
sample preservation techniques. The sulfide losses appeared to result
from the air oxidation of sulfide ion to sulfate ion. Zinc acetate and
sodium carbonate were added to co-precipitate the sulfide ion. The sul-
fide ion was redissolved and the concentration determined by the ion selec-
tive electrode method. Samples were stored by this method of preservation
for up to 7 days without appreciable sulfide losses.
Mercury ion is preserved by adding enough nitric acid to adjust the
sample solution pH to 1. Sodium dichromate is added to eliminate any
volatile mercury compounds.
For the remainder of the species present in the geothermal fluids,
the only preservation technique used was sample dilution. In a study,
using various concentrations of various acids to preserve certain species
in the sample fluid, it was found that sample dilution resulted in data
with better precision.
Special Chemical Studies
An experiment using the geothermal water from the Raft River Geother-
mal Well #7 for irrigation of a variety of grains and potatoes was per-
formed. The analysis consisted of determining fluoride concentrations on
the surface and within the grain and potatoes. Table III shows the results
of the analysis.
199
-------
TABLE III. FLUORIDE CONCENTRATIONS ON THE SURFACE
OF AGRICULTURAL SAMPLES.
Fluoride,Cone.
u9/ml
Triangle Dairy Potable Water Well 0.81
Triangle Dairy Warm Water Well 14.2
Erwln Potable Water Well 8.4
Erwin Stock Hater Well 9.7
Erwin Irrigation Well 11.5
Ucly Irrigation Well 2.0
BLM Hot Water Well 5.6
Cranks Hot Water Well 4.1
haft River Seothermal Well SI 8.0
Raft River Goothermal Hell #2 9.5
Raft River Geothermal Well #3 5.0
Cane Canyon Spring < 0.4
Redrock Noll Spring < 0.4
Cedar Noll Spring < 0.4
Irrigation Hell NE of Malta, Idaho < 0.4
Raft River 0.8
Wytes Well 7.6
TABLE IV. FLUORIDE DATA FROM REGIONAL WELLS AND SPRINGS.
Fluoride Concentration (ug/g)
Sprinkled With
Geothermal Water
Barley 8.7
Wheat 32.
Oats 54.
Potatoes
Flooded With
Geothermal
Water
0.7
4.3
0.1
Sprinkled With
Fresh Water
0.7
4.2
8.9
0.07
Total Fluoride Concentrations for
Agricultural Samples
Barley
Wheat
Oats
Potatoes
14
33
65
16
23
0.:
f~ concentration in geothermal water is ^9.5 ug/tnl
F" concentration in fresh water is ->-2.0 yg/nl
200
-------
The method for fluoride, suggested by Orion Laboratories^ for total
grain analysis, resulted in the grains forming a jelly-like substance that
resisted analysis. A method involving ignition of the grain powder in a
Ca20 slurry was used. The method used to analyze the surface fluorides
involved in the leaching of the samples with a complexer buffer solution
consisting of citrate, triethylzmine, and sodium chloride. The fluoride
concentration was determined by anion selective electrode. The standard
addition method was used to check for interferences from the sample matrix.
Fluoride Concentrations of Regional Effluents
Although fluoride analysis of water samples required no special prepara-
tive or analytical methods, data from regional wells and springs have been
tabulated in Table IV. This study did prove to be of interest due to the
variation in fluoride concentration of regional effluent sources.
201
-------
Ferric-Ferrous Ion Study
A study was made on the geothermal water taken from Raft River Geother-
mal Well #1 to determine the ferrous-ferric ion ratio. This study was
necessary due to the lack of down hole sampling equipment and consequently,
the ability to sample at the source of the geothermal fluid. It is be-
lieved that ferric ion comes from the fluid source, whereby, ferrous ion
results from fluid reaction with the well casing and other down hole
equipment.
The Hach DR-2 spectrophotometer at the well site was used for the
analysis. The ferrous ion concentration was determined colorimetrically
by developing the characteristic yellow color of ferrous ion complexed
with 0-phenanthroline. Then, hydroxylamine was added to reduce ferric ion
to ferrous ion and the total ferrous ion concentration determihed. The
difference represents the ferric ion concentration. The ferrous ion con-
centration in the geothermal fluid was determined to be 0.100 ± 0.004 ppm.
Ferric ion content was below the limit of detectability. It is probable
that most of the iron in the fluid originates from the casing and other
ferrous materials dropped into the well- during drilling and Togging
'••aerations.
New Temperature Predictions at Raft River
Geothermometry calculations from Raft River Geothermal Well samples
predict aquifer temperatures up to 188°C. Maximum measured bottom hole
temperatures have been about 146°C which probably indicates the major geo-
thermal aquifer has not yet been tapped. Both geothermometers employed
show higher aquifer temperatures for RRGE2 than for RRGE1. See Table V.
The Si02 and Na-K-Ca- geothermometers remain the most reliable indi-
cators of aquifer temperature^8'9'. The Si02 thermometer depends on the
202
-------
TABLE V. ANALYTICAL DATA AND PREDICTED TEMPERATURE FROM RAFT RIVER WELLS.
Sample #
Well -Pres.(psi)
RRGEl
RRGEl
RRGEl
RRGEl
RRGE2
RRGE2
RRGE2
g RRGE2
RRGE 2+
RRGE2"1"
RRGE2+
RRGE2+
RRGE 2
RRGE 2
Crank
BLM
1.27f*
2-160
2-125
3-125
2-125
5-115f
6-110f
10-108
1 -b 120f
2 -v. 120f
5 -v. 120
6 -v. 120
4-115
5-110
Date
Sampled SiO?fppm)
3-23-75
3-23-75
4- 4-75
4- 4-75
7- 9-75
7- 9-75
7- 9-75
7- 9-75
7-17-75
7-17-75
7-17-75
7-17-75
7-24-75
7-24-75
10-18-74
10-18-74
94
88
73
66
148
169
165
143
175
165
128
148
152
152
88
83
Ca (ppm)
50
52
56
56
36
38
37
38
40
40
37
41
38
40
125
54
Ca(mol)
1.25x10-3
1.3 xlO-3
1.4 xlO~3
1.4 xlO~3
9 xlO"4
9.5 xlO-4
9.25xlO-4
9.5 xlO"4
1 xlO-3
1 xlO^
9.25x10-4
1.025x10-3
9.5 x!0"4
1 xlO"3
3.12x10-3
1.35x10-3
Na(ppm)
350
368
307
305
355
368
373
387
391
397
390
394
385
1082
534
Na(mol)
.0152
.016
.0132
.0133
.0154
.016
.0162
.0168
.017
.0173
.017
.0171
.0167
.0470
.0232
K (ppm)
25
25
21
21
35
43
38
33
37
33
35
34
34
32
20
K(mol)
6.39xlO-4
6.39xlO-4
5.37xlO~4
5.37x10-4
8". 95xlO"4
1.09x10-3
9.72x10-4
8.43xlO-4
9.46x10-4
8.43x10-4
8.95xlO"4
8.69xlO"4
8.69xlO"4
8.18x10-4
S.llxlO"4
SiO
Temp. °C
125
125
100
100
165
170
170
160
175
170
132
165
165
165
125
125
Na-K-Ca
Temp. °
169
167
164
164
190
199
193
182
188
189
185
184
184
138
144
*Samples filtered before collection
+Samples maintained at collection temperature till just before analysis
-------
solubility of quartz as a function of temperature. The Na-K-Ca thermo-
meter is an empirical calculation based on exchange reactions of the feld-
spars. Raft River solutions chemically align closely with other geo-
thermal water of rhyolitic origin. Since rhyolites are composed of pri-
marily orthoclase (K-feldspars), quartz, and plagioclase (Na-Ca feldspar
solid solution) both geothermometers should apply. Application of this
thermometer to Raft River Geothermal Wells predicted higher than measured
temperatures.
The Na-K-Ca thermometer predicts substantially higher temperatures
than the Si02 thermometer. Except for samples collected on 7/17/75, the
samples were cooled to ambient temperature following collection. This
may be sufficient to precipitate some amorphous silica, thus reducing the
silica content in solution. Filtered samples show higher Si02 content
than the unfiltered samples and this probably reflects reaction with iron
oxides entrained from the casing. Qual itative analysis of f I Iter samples shows
Fe to be the major constituent and solids filtered just prior to emmission
spectroscopy show both Fe and Si to be major components. Because of the
likely problems with Si02 precipitation, the temperature predicted by the
Na-K-Ca geothermometer is like1'/ to be more accurate. The possible excep-
tion is the filtered samples collected on 7/17/75 which predict temperatures
of 170° and 175°C.
These data indicate higher aquifer temperatures than those thus far
encountered. The temperatures predicted by the shallow Crank and BLM Wells
are lower due to continued reaction during transit in the rock . For
the same reason, the higher temperatures predicted by the geothermometers
indicate the wells may not yet be near the maximum aquifer temperature.
204
-------
TABLE VI. AVERAGE CONCENTRATION VALUES FOR SPECIES
ANALYZED AT THE RAFT RIVER GEOTHERMAL WELLS.
WeVI History of the Raft River Gepthermal Wel^s
The most important study of a geothermal well is its chemical history
from the time of completion. The routine analysis of dominant species
over this period develops a record concentration. Once the magnitude is
known and trends established, it can assist in predicting more accurately
source temperature, corrosion, and deposition rates in well and power pro-
ducing equipment, the rate of well depletion, and many other determinations
of importance in the geothermal well development. Figures 6, 7, and 8
show the variation of Ca++, CaC03, Si02, Cl~, F", and Na+ concentrations
over the sampling period of each well.
Table VI shows the average value for the concentration of the listed
chemical species.
WELL RRGE-2
SPECIES
HC03~
TOTAL P
Si(OH)k
Cl
F~
Na+
TOTAL Si
sr
+
WELL RRGE-1
CONCENTRATION
(PPM)
77
2.5
0.012
162
0-036
717
5.6
61
56
403
44
1.3
1.5
27
WELL RRGE-3
CONCENTRATION
(PPM)
40.9
0.51
0.02
212
0.021
682
8.6
55
38
383
70
1.04
1.04
36
CONCENTRATION
(PPM)
53
0.47
0-25
193
0.24
2001
4.3
32
193
1156
67
3.1
6.4
91
TABLE VI.
AVERAGE
ANALYZED AT THE
205
-------
RRGE 2 HISTORY
01
c 4)
o x
C —
(D C
U =5
C
O
o
Figure 7. Raft River geothermal well # 2 chemical history
RRGE 3 HISTORY
2SOO
2000
1500
_IOOO
E
100
50
HC05 as Ca
A Si02
Figure 8. Raft River geothermal well # 3 well history,
206
-------
Research Programs at the Raft River Geothermal Site
A number of experiments have been, or are being performed at the Raft
River Geothermal Site. These include the following:
1) Fluidized bed heat exchanger study for the control of scale on
heat exchange surfaces.
2) Direct contact heat transfer experiments.
3} Agricultural experiments using geothermal water to irrigate a
variety of grains and vegetables.
4) Corrosion study of materials for power plant components.
5) Experiments to determine the feasability of using treated geo-
thermal water in cooling tower applications.
6) Fish farming study to determine growth and effects of geothermal
water on fish.
On-Site Analytical Chemistry Laboratory
The experimental use of the Raft River Geothermal Site has been men-
tioned to show the need for many on-site chemical analysis. On-site chemi-
cal analysis are required for:
1) monitoring of selected chemical species required for well histories,
2) environmental monitoring and studies,
3) reservoir and well evaluation,
4) experimental support, and
5) analysis of unstable chemical species that cannot be stored for
transport to off-site laboratories.
This facility is being assembled at the Raft River Geothermal Site. The
laboratory will contain an atomic absorption spectrophotometer, a colori-
metric spectrophotometer, an automatic titrator, a gas partitioner, a
specific ion meter, electrodes, and a number of pH, salinity, and conductivity
207
-------
meters. The greatest limitation is the lack of trained personnel. Once
they are available, the laboratory will be capable of a broad spectrum
of analyses.
In-line Monitoring Programs
Plans have been initiated for the installation of monitoring equip-
ment for in-line measurement of pH, oxidation-reduction potential, and
conductivity at the geothermal wellheads. Equipment has been ordered and
installation will take place upon shipment of this equipment. The moni-
toring will be possible because pH, ORP, and conductivity probes designed
for 125°C temperatures and 150-psi pressure are avaliable. This equipment
is designed and manufactured by Balsbaugh Inc. This system, we hope, will
signal changes in well composition. Researchers at Battelle Northwest in
Richland, Washington are working to develop probes for conductivity and
oxidation-reduction potential that will withstand higher temperatures and
pressures so monitoring systems of this type can be extended to other geo-
thermal sites.
Conclusions
1. After testing several sampling techniques we adopted collection under
pressure followed by simultaneous cooling and dilution under pressure.
This technique prevents errors due to phase separation and minimizes
the hazard of deposition on walls of the sample container. Samples
kept hot until analysis time sometimes contained precipitates high
in Si02 and iron.
2. An ignition technique for fluoride in grain was developed.
3. A sulfide ion preservation technique was developed so samples could
be analyzed several days following collection.
208
-------
4. Well composition for the three geothermal wells were plotted as a
function of age. Most components gyrated wildly during the first
few months following drilling then settled into a concentration range
with much smaller deviations.
5. The need for on-site chemical analysis was documented. A laboratory
is being constructed. Typical of remote geothermal locations the
limit on laboratory capabilities will be imposed by the lack of
trained,personnel.
209
-------
References
1. Christoffersen, R. N., Wheatley, R. N., and Baur, J. A. "Union Oil
Company of California's Geothermal Sampling Techniques", Proceedings
of the First Workshop on Sampling Geothermal Effluents, October 20,
21, 1975.
2. Ball, J. W., Jenne, E. A., Burchard, J. M., and Truesdell, A. H.
"Sampling and Preservation Techniques for Waters in Geysers and Hot
Springs", Proceedings of the First Workshop on Sampling Geothermal
Effluents, October 20, 21, 1975.
3. Kunze, J. F., et.al., Geothermal R&D Project Report for Period July 1,
1975 to September 30, 1975. Prepared under contract by Aerojet Nu-
clear Company to ERDA, ANCR-1281, December, 1975.
4. Kunze, J. F., editor, Geothermal R&D project Report for Period April 1,
1976 to June 30, 1976. Prepared under contract by EG&G for ERDA,
TREE-1008, October, 1976.
5. Stoker, A. K. and Purtymun, W. D. "Some Problems Involved with Sampling
Geothermal Sources.", Proceedings of the First Workshop on Sampling
Geothermal Effluents, October 20, 21, 1975.
6. Brown, E. B., Skougstad, M. W., Fishman, M. J. Techniques of Water
Resources Investigations of The United States Geologic Survey V-5.
7. Gallaway, H., Shoat, R. E., Skaggs, C. H., American Industrial Hygene
Association Journal, October, 1975.
8. Fournier, R. 0. and Rowe, J. J. "Estimation of Underground Temperatures
from the Silica Content of Water from Hot Springs and Wet Steam Wells",
Amer. J. Sci. 264, 1966.
9. Fournier, R. 0. and Truesdell, A. H. "An Empirical Na-K-Ca Geothermo-
meter for Natural Waters", Geochimica et Cosmochimica Actu 37, 1973.
210
-------
BOREHOLE GEOPHYSICAL LOGGING AS COMPLEMENT TO
GEOTHERMAL WELL EFFLUENT SAMPLING
S. K. Sanyal and R. B. Weiss
Geonomics, Inc., Berkeley, California
ABSTRACT
Sampling and analysis of well effluents is an important means
of detecting ground water pollution. There are, however, some
drawbacks in this technique. If water is sampled from an uncased
(open hole) well, the water represents a mixture of water from
various permeable layers, rather than from an individual layer.
This is an undesirable situation if water quality varies signifi-
cantly from layer to layer. If the water is sampled from a cased
well, the sample represents only the layers at the depths of well
completion. One way of obtaining a continuous vertical profile
of water quality with depth is to apply geophysical borehole
logging to complement well effluent sampling and analysis, in
both open and cased holes. Both real and hypothetical examples
are given to illustrate the potential role of borehole geophysical
logging in pollution monitoring.
Geophysical Well Logs in Water Quality Studies
In order to evaluate the possibility of ground water pollu-
tion, one has to assess the qualities of both the shallow ground
water and the geothermal water. The. most direct means of assess-
ing water quality is by chemical analysis of the1 water sample.
However, water sampling from wells can be complicated by several
factors as discussed below.
Both shallow ground water and geothermal water occur in a
sequence of porous and permeable rock layers alternating with
impermeable layers (Figure 1). If water is sampled from an
uncased (open-hole) well, the water represents a mixture of
water from various permeable layers, rather than from an
211
-------
SELF POTENTIAL—» RESISTIVITY
M
M
to
WATER SALINITY
==^J SHALE (IMPERMEABLE)
SAND (PERMEABLE)
PROFILES
BASELINE
I YEAR
2 YEARS
c
Figure 1. Vertical profile of water
quality from well logs.
Figure 2. Temporal changes in log
response indicating change
in water quality profile.
-------
individual layer This is an undesirable situation if water
quality varies significantly from layer to layer If wate?
is sampled from a cased well, sampled water represents only the
layers at the depths of casing perforations. One way of obtain-
ing a continuous vertical profile of water quality with depth
is to use geophysical borehole logging techniques. For example
after a well is drilled and before it is cased, electrical and
self -potential logs are normally run (simultaneously) in the
well. A careful analysis of these logs permits calculation of
water conductivity versus depth. If sodium chloride is the pre-
dominant dissolved constituent, the parts per million of the
total dissolved solids can be calculated from the conductivity.
If large amounts of other dissolved compounds are present, one
can calculate an "equivalent" Nad concentration in the water.
If the chemical analysis of a water is known, an equivalent
NaCl concentration can be calculated by using standard charts.
The primary importance of using well logs is to obtain a vert-
ical profile of the relative amounts of total dissolved solids
without the need for sampling. Such a profile serves as the
baseline data against which temporal changes in water quality
can be evaluated. Figure 1 shows a case where water salinity
increases with depth as inferred from the well logs. By repeated
logging of a well at regular intervals , any temporal change in
water quality can be detected.
Figure 2 illustrates a temporal change in water quality as
monitored by well logging. Here, after one year, layer 3 shows
evidence of contamination by saline water, while the other four
layers preserve their original quality. After 2 years, the third
layer shows further deterioration in quality, while layers 2 and
4 have begun to be contaminated. Thus, unlike monitoring by
sampling and chemical analysis, the proposed technique not only
indicates changes in water quality but also identifies the exact
depths of contaminated layers and the degree of contamination in
each layer.
If the wells are uncased, electrical resistivity and SP
logs can be used for such monitoring. If the wells are
cased, these logs cannot be used. In such situations we
recommend monitoring by pulsed neutron logging. If the
porosity of a formation is known (from core data or other
logs) , salinity of the formation water can be calculated
from a pulsed neutron log. Unlike electric logs this log
is not affected by the presence of casing around the borehole.
Both these logs can be used to monitor water quality more
effectively if used in conjunction with water sa nplj-ng-
Chemical analysis of a few water samples can be used t
"calibrate" the proposed monitoring technique. Pulsed
neutron logging can be used to detect changes in the
to
213
-------
concentration of certain special elements (e.g., boron, lithium)
in formation water. These elements have very large neutron
capture cross sections and consequently are easily detected by
this log. Temperature logs may be used to monitor the movement
of a contaminant, which is either hotter or cooler than^the
aquifer. Disposal of geothermal waste water can be monitored
by this method.
Monitoring of the Contaminant Front
It is important to monitor the entire profile of water
quality in a well particularly if there is a large variation
in permeability of the various layers. Waste water disposed
of in a well will travel faster through the more permeable
layers (e.g., layer 3 in Figure 2). Thus, the surface of con-
tact, between contaminated and uncontaminated water will not be
uniform (by any reckoning). Reservoir engineering calculations
can be applied to monitor and predict movement of this complex
boundary. Molecular diffusion and mechanical dispersion make
this boundary even more complex. Reservoir engineering tech-
niques can be applied to evaluate diffusion and dispersion.
Figure 3 shows a hypothetical disposal problem. By monitoring
the water quality profile (calculated from logs) in the
observation well, one can infer the movement of the contaminated
water front. Such a monitoring technique can provide sufficient
forewarning so that water wells can be saved from contamination,
The disposal activity may be stopped or the producing wells may
be selectively re-completed in layers safe from the advancing
contaminated water.
Similar non-uniformity in the advance of contaminated
water is also seen in the areal sense. This non-uniformity
is caused primarily by the areal variation in formation
permeability as well as the production characteristics of
other wells in the aquifer. Reservoir engineering calculation
techniques can be applied to monitor and predict the areal
advance of the contaminated water. Figure 4 illustrates
such a situation. The observation wells not only provide
vertical definition of the advancing front but also the areal
definition.
Thus, a judicious combination of well-logging, water
sampling and concurrent reservoir analysis will be the most
effective water quality monitoring scheme. Log monitoring
can be done in observation wells (such wells will not be
produced) as well as producing or injection wells, If the
well is not cased, an electrical log can be used for monitor-
ing. For steel-cased wells pulsed neutron logging should be
used. However, a well can be cased with fiber glass casing
214
-------
DISPOSAL OBSERVATION WATER
WELL WELL WELL
CONTAMINATED WATER
AFTER A FEW YEARS
Figure 3. Vertical conformance of the contaminated water front.
WATER WELLS
O
2 .-OBSERVATION
O WELL
^-DISPOSAL
WELL
Figure 4. Areal spread of contaminated water,
215
-------
to allow for monitoring by electric logs. Comparison with
chemical analysis of selected water samples will improve log
monitoring. If there is no salinity contrast between the
contaminated and the uncontaminated waters, electrical logging
will not be an effective monitoring tool. In such cases the
salinity of the disposal water may be increased by adding a
suitable inorganic salt or the neutron capture cross section of
the disposal water may be increased by adding a high neutron
capture cross-section element (provided water does not already
have a significant amount of that element). Electrical logs
can be used in the former case, while pulsed neutron log can
be used in the latter.
Contamination Problems in Cased Holes
Aquifers may also be contaminated by the movement of con-
taminated water behind the casing. This is caused by differences
in piezometric levels between various aquifer layers. Fluid
migrates from a higher pressure to a lower pressure one through
cracks or channels in the cement between the casing and the
formation. Such flow behind casing in a disposal well may
contaminate a shallow aquifer without being detected. Some-
times such flow causes deposition of radioactive salts behind
casing after several years of flow. A gamma ray log run in the
well will detect this anomalous radioactivity. By comparing
gamma ray logs run in the same well at various times, one can
often detect flow behind casing. In the extreme case, flow be-
hind casing may create holes in the casing due to corrosion.
Injected disposal water may escape into shallow ground water
aquifer and contaminate it. A hole in the casing can sometimes
be inferred from an anomalous increase in gamma ray intensity
at a certain depth. However, this technique may not be de-
finitive in all areas. On the other hand, SP log inside a cased
hole is sensitive primarily to corrosion effects in the casing
itself. Hence, changes over time of the SP response inside
cased holes are likely to reveal zones of active corrosion.
Gamma ray logging is inexpensive and is always run together
with pulsed neutron logs. If pulsed neutron is used for
monitoring, it may be monitored for any anomalous change in
gamma ray intensity. Temperature logs (regular or differential)
can be used to detect a variety of flow-behind-pipe problems.
216
-------
ABSTRACT
USE OF RADIOACTIVE TRACERS IN GEOTHERMAL OPERATIONS
0. Vetter, Vetter Assoc., Laguna Beach, California
Tritium in form of tritiated water (THO) is the most suitable
isotope for tracing the liquid water or steam in the field. It
will follow the water (in the brine) through all phase changes
as opposed to any other tracer. Some possibilities to use THO
in reservoir evaluations in steam, - or liquid - dominated re-
servoirs are outlined. Of particular interest is the precise
determination of the reservoir recharge through injection wells.
Qualitative and quantitative data can be obtained. Conclusions
can be drawn as to the numbers, size, direction and conductivity
of fractures and/or high permeability streaks. Pros and cons
for the use of THO are given.
Other radioactive isotopes can be used to supplement conventional
analytical measurements to solve some of the unique problems in
geothermal operations. Pb-210; Zn-65; Ag-110 and etc., are per-
fectly suited to determine the behavior of the equivalent non-
radioactive elements with an accuracy not possible by using con-
ventional analytical techniques. Measurements of scaling ten-
dencies and material balances are a few examples.
Radioactive isotopes can also be used to determine chemical or
hydrodynamic data through the steel of pipes, vessels, etc.
Velocities and flow rates of steam vs. brine, scale formation
and flash percent are described as examples.
217
-------
THE LBL GEOTHERMAL BRINE DATA COMPILATION PROJECT
Steven R. Cosner and John A. Apps
Lawrence Berkeley Laboratory
University of California
Berkeley, California 94720
ABSTRACT
Data on geothermal fluids from the principal geothermal resources
in the United States are being compiled at the Lawrence Berkeley Laboratory.
The data collected are stored on a computer system to facilitate selective
retrieval and manipulation. Printed copies of the compilation will be
available with documentation from the laboratory as report LBL 5936.
218
-------
a projectVuwS ™?*™<» •». funded
geothermal fluids for the principal qeotherl 5 I * 9nd COITlptle data on
States. The project will contr bute tn . SV -°urces 1n the United
designed to overcome major " n9ineer^ ror
of geothermal resources for
Lawrence Brrtre^Sr W ^™ USed
statistical evaluations of a^nth^mT? ^lon Wl11 be used to
s
steam utili^fLnu .Priv?*e industry has proved the success of dry
rpoili^ i I - "2* Spring data are be1n9 collected by the U S
Geological Survey in Reston, Virginia, using the 6EOTHERM computer system.
in Tab?paihaVl??f hC°11KCted,a"?1TP.iled for the 9eothermal areas listed
chPmi,lrv H»*! 1 ^e b6?? dnlled 1n a number of other are«, but fluid
cnemistry data from the wells are not available.
Table 1. Geothermal areas from which fluid chemistry
data have been compiled.
California Nevada
Sal ton Sea Beowawe
East Mesa
Heber .
Mono-Long Valley New Mexico
Baca Location No. 1*
Idaho
Raft River ^*-
Roosevelt Hot Springs
Baja California, Mexico
Cerro Prieto
* Baca Location No. 1 KGRA is same area as Valles Caldera.
219
-------
The sources of most of the data are the published literature, and
reports published by government agencies and national laboratories.
Other sources are private well operators. It has been difficult and many
times impossible to obtain comprehensive data from larger companies who
own geothermal wells.
The data are stored on the Berkeley Database Management System, or
BDMS. With BDMS, databases may be built easily using data elements or
categories of storage tailored to the data being compiled. The computer
system facilitates selective data retrieval and manipulation, editing
and updating, and addition of new data. The system automatically main-
tains indices to the information to aid in retrieval and manipulation
once the data are compiled.
The compilation of geothermal data has over fifty storage categories
or data elements as shown in Table 2.
Figure 1 is a computer printout of a typical record of compiled data.
The computer system is designed to print only the data present, with no
extra space left for absent data.
For a few geothermal wells there are a number of fluid analyses
published. In these cases all the data are compiled with multiple record
listings for each well. There is usually a wide variation of constituent
concentrations among samples from a single well, so the analysis consid-
ered best is given as the first listing for the well.
It has been observed during the compilation project that the units
of concentrations in fluids are treated carelessly. The compilation
gives the units used by the source, whether they are parts per million
by weight, milligrams per liter, or something else.
A bibliography of all the sources of information used in the fluid
data file has been maintained. This bibliography will also be stored
on the BDMS computer system. Computer storage on BDMS allows computer
searches for articles which discuss specific topics. To aid in the com-
puter search, descriptors or key words are assigned to each article.
A bibliographic printout is shown in Figure 2. The entry at the
upper right of the printout denotes the principal author and year of
publication of the report. This entry corresponds to the bibliographic
source given in the fluid data printout. The bibliographic printout
also lists the descriptors to help a reader determine the subject matter
of the report.
Once all the available data are compiled, evaluations of geothermal
fluids will be made. Where sufficient data are available, the analyses
will be checked for accuracy using thermodynamic equilibrium models.
Temperature vs. concentration diagrams for the common elements can be
made to help characterize the geothermal effluents. Evaluation of the
220
-------
Table 2. Data elements, geothermal fluid data file.
to
Basic Information
Well name
KGRA or geothermal field
Location
County
State
Country
Well Information
Well owner
Lessee
Drilling company
Date drilled
Well Data
Well depth
Temperatures
Depth of temperature readings
Pressure, shut-in
Flow rates, pressures
Production interval
Reservoir lithology
Production interval
Brine Information
Sample type
Sample date
Sample number, laboratory
Sample location
Sampling method
Condition of Sample
Condition of well during sampling
Physical Data
PH
Range of readings
Temperature of pH reading
Electric potential
Temperature of reading
Specific gravity
Temperature of reading
Specific conductance
Temperature of reading
Viscosity
Total dissolved solids
Residue on evaporation or sum?
Total alkalinity
Other measurements
Brine Data
Method of analysis
Error limits
Units
Constituent name
Concentration
Other note (trace, errors, etc.)
Bibliographic Data
Sources
-------
RECORD 56
CODE NAME"SINCLAIR 4A
SAMPLE TYPE-WATER
VELL SINCLAIR 4
SALTON SEA KGRA
LOCATION" T12S, R13E* SEC. 4, 400FT N, 200FT £. FROM S QTR.COR.
IMPERIAL COUNTY, CA., USA
VELL INFORMATION
CVKCR-- GEOTHERMAL ENERGY AMD MINERAL CORP.
DATE DRILLED— 25 APR 64-4 JUN 64
VELL DATA
DEPTH 1617 METERS
TEMPERATURE 255 C AT WELLHEAD
PRESSURE* SHUT-IN-- >445 PSIG.
SAMPLING INFORMATION
DATE— 3 APR 75
SAMPLE NUMBER, LABORATORY-- SAMPLE 6THS, LAWRENCE LIVERMORE LAB.
SAMPLE LOCATION-- 25FT FROM WELLHEAD
SAMPLING METHOD— SAMPLING PROBE INSERTED INTO TOP PORTION OF 6
IN. PIPE, COLLECTED UNDER PRESSURE INTO TEFLON-LINED
STAINLESS STEEL BOTTLE.
CONDITION OF SAMPLE" SAMPLE TEMP»210 C, SAMPLE PRESSURE-23Q
PSIG. SAMPLE NOT COOLED DURING SAMPLING.
CONDITION OF WELL DURING SAMPLING— FULL FLOW THROUGH 6 IN. PIPE.
PHYSICAL DATA
PH« 5.10 .-...-. TEMP DURING READING" AMBIENT
ELECTRIC POTENTIAL" .ISO VOLT. TEMP DURING READING* AMBIENT
TOT DISS SOLIDS- 291.00 G/L, SUM
300.1)0 G/L* RESIDUE ON EVAPORATION
OTHER DATA—
DENSITY" 1.22 AT 25 C.
ENTHALPY- 210 CAL/G.
COMMENT-- PH AND EH VALUES TAKEN ON SAMPLE PRESSURIZED BY GAS
COLLECTED WITH SAMPLE.
BRINE DATA
METHOD OF ANALYSIS-- RECONSTRUCTED ANALYSES--CALCULATED FROM
LIQUID AND SOLID PHASES IN SAMPLE.
ACCURACY OF ANALYSIS— 5 PERCENT, EXCEPT NA AND AG--10 PERCENT;
AL--I PPMJ S POSSIBLY LOW BY A FACTOR OF 2 OR 3.
UNITS— MG/L
CONSTI-
TUENT
SI
NA
X
CA
KG
CL
AS
AL
CU
FE
MN
PB
S
CONCEN-
TRATION COMMENT
278
64000
14700
30400
76
179000
0.7
5
3.3
1600
1310
111
14
CONCENTRATIONS NOT CORRECTED FOR STEAM LOSS.
BIBLIOGRAPHIC DATA
SOURCES--
HILL 75
W1THAM 76
HELGESON 68
MUFFLER 69
Figure 1. Computer printed output of compiled data,
222
-------
abundance or scarcity of information may lead to suggestions for further
data collection efforts.
HILL 75
TITLE- SAMPLING A TWO-PHASE GEOTHERMAL BRINE FLOW FOR
CHEMICAL ANLAYS1S.
AUTHOR- HILL* J.H.;MORRIS* C.J. CCALIFORNIA UNIV.*
LIVERMORE (USA). LAWRENCE LIVERMORE LAB.3.
REFERENCE- SAMPLING A TWO-PHASE GEOTHERMAL BRINE FLOW
FOR CHEMICAL ANALYSIS. UCRL-77544* CALIF.
UNIV.* LLL* LIVERMORE* CALIF.* (5 DEC !975>*
27P.
DESCRIPTORS- SALTON SEA KGRA; GEOTHERMAL BRINES;
SAMPLING METHODS; CHEMICAL ANALYSIS; GEOTHERMAL
WELLS* TWO-PHASE FLOW.
Figure 2. A computer-produced bibliographic listing.
It is the intent of this project to make the compilation as complete
as possible Therefore, anyone who has information or data on fluids
from geothermal wells is sincerely requested to send the data to the
authors. Contributions of current data are necessary to maintain the
compilation as a useful tool aiding in geothermal energy development.
At this point the fluid data compilation is available as an easily
readable computer printout. It will be published later by Lawrence
Berkeley Laboratory as report number
223
-------
ABSTRACT
ANALYSIS OF TRACE CONTAMINANTS
IN LOW TO MEDIUM SALINITIES GEOTHERMAL FLUIDS
K. Y. Chen
Environmental Engineering Program
University of Southern California
Los Angeles, California
Water samples from low to medium salinities KGRA were collected
and analyzed for complete chemical composition, with special
attention given to trace contaminants. The analytical deter-
minations including speciation of redox species, for example,
As(III) and As(V); partition between soluble and particulate
fraction; as well as cations and anions balance.
Among a total of 49 geothermal water samples collected from
California, Nevada, and Mexico, 56% of the samples exceed the
maximum permissible concentration of 50 ppb arsenic for drinking
water, while 83% of the samples exceed the recommended limits of
10 ppb. The concentration range for total arsenic is 0.5 to
3,006 ppb. Concentrations of As(V) range from 5 to 95% of total
arsenic. 80% of the samples exceed the maximum permissible
fluoride concentration of 1.3 ppm (79 to 90°F) for drinking
water, with a range of 0.13 to 15.8 ppm. 98% of the samples
exceed the recommended limiting concentration of 0,5 ppm boron
for agricultural usage. The range was from 0.14 to 4.56 ppm.
Ammonia and sulfide are significant contaminants in vapor
dominant systems. In some cases, concentration of chloride,
sulfate, and total dissolved solids may also present potential
water quality problems. Trace metal concentrations are gener-
ally low and below the recommended drinking water standards.
224
-------
AUTHOR INDEX
Name
Allen A. C.
Altshuler, S. L.
Apps, J. A.
Asaro, F.
Baldwin, J. M.
Bishop, H. K.
Blatz, L.
Bowman, H. R.
Brown, L. L.
Burnett, E.
Cavanagh, L. A.
Chen, K. Y.
Conner, G. R.
Cosner, S. R.
Crecelius, E. A.
Fournier, R.
Fruchter, J. S.
Frye, G. A.
Grigsby, C.
Hamersma, W.
Hebert, A.
Hill, J. H.
Holley, C.
Jepsen, A. F.
Juprasert, S.
Klein, C. W.
Koenig, J. B.
Kruger, P.
Paqe
190
16
218
55
190
160
174
55
114
175
42
224
143
218
2
141
2
42
174
69
55
46
174
3
84
166
166
14
Langan, L.
Littleton, R. T.
Ludwick, J. D.
McAtee, R. E.
McCurdy, R. A.
Meidav, H. T.
Michels, D. E.
Needham, Jr., P. B.
Nehring, N. L.
O'Connell, M.
Otto, Jr., C. H.
Reed, M. J.
Robertson, D. E.
Ross, L. W.
Sanyal, S. K.
Soinski, A. J.
Subcasky, W. J.
Tester, J. W.
Thompson, J. M.
Tonani, F. B.
Truesdell, A. H.
Tsai, F.
Vetter, 0.
Weiss, R. B.
Wheeler, D. W.
Williams, R. E.
Wollenberg, H.
Woodruff, E. M.
3
175
2
190
16
145
70
143
130
15
46
113
2
114
84, 211
1
43
174
141
97, 145
130
84
217
211
42
114
55
44
«.S. COVERT PRINTING OFFICE: 1978 - 785-974/1231 Regie, No. 9-1
225
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
REPORT NO.
EPA-600/7-78-121
3. RECIPIENT'S ACCESSION NO.
TITLE AND SUBTITLE
PROCEEDINGS OF THE SECOND WORKSHOP ON
SAMPLING GEOTHERMAL EFFLUENTS
5. REPORT DATE
June 1978
6. PERFORMING ORGANIZATION CODE
. AUTHOR(S)
Subir Sanyal and Richard Weiss
8. PERFORMING ORGANIZATION REPORT NO,
PERFORMING ORGANIZATION NAME AND ADDRESS
Geonomics, Inc.
3165 Adeline Avenue
Berkeley, CA 94703
10. PROGRAM ELEMENT NO.
1NE624
11. CONTRACT/GRANT NO.
68-03-2468
12. SPONSORING AGENCY NAME AND ADDRESS
U.S. Environmental Protection Agency—Las Vegas, NV
Office of Research and Development
Environmental Monitoring and Support Laboratory
Las Vegas, NV 89114
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
EPA/600/07
15. SUPPLEMENTARY NOTES
For further information, contact Donald B. Gilmore, Project Officer, (702)736-2969,
ext. 241, in Las Vegas.
16. ABSTRACT
This is a compilation of papers presented at the second in a series of workshops on
sampling and analysis of geothermal effluents held February 15-17, 1977, at Las Vegas,
Nevada. The purpose of this workshop was to continue the exchange of ideas and knowl-
edge initiated in the first workshop of October 1975 with the intent of eventually
developing an acceptable set of standard geothermal effluent sampling and analysis
methods. Thirty-one papers were presented by representatives of industry, universities
and government. All of the abstracts and 17 of the papers are published in this docu-
ment.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI F'ield/Group
Monitors
Pollution
Sampling
Analysis
Geothermal energy
Methods development
13B
14A
14B
14C
14D
18. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (This Reportf
UNCLASSIFIED
21. NO. OF PAGES
256
20. SECURITY CLASS (This page I
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (Rev. 4-77) PREVIOUS EDITION is OBSOLETE
------- |