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

-------