TN295 
 
 IE 
 
 No. 9216 
 


 
 A* *<^ 
 
 "W 
 ^ 
 
 ■> * 
 
 ;• / ^/-TW^^ ^ 
 
 F * <0, 
 
 
 
 <* *<T T** .6* ^ •• . » • A O 
 
 
 ^°^ 
 
 :- «b v* 
 
 
 
 
 
 4 0. 
 
 
 1 *1>*Z» 
 
 
 3,0-^ 
 
 
 
 
 
 
 ' • <$> <P\ o^ 
 
 ^<=>- 
 
 8 e, % V 
 0^ V> '">* A 
 
 
 <, 
 
 ^~ . « • 
 
 v > ..i^v <*i 
 
 -. o 
 
 ^° A 
 
 
 
 
 
 
 
 
 
 
 
 i • ^n A » J 
 
 ^ 
 
 
 
 *^v :m m.\ ^^ 
 
 •^ ^ 
 
 •^^ 
 
 
 
 
 .?.* A 
 
 
 
 A <^ — « G 
 
 q,_ ♦ . . o 
 
 -5°^ 
 
 ^ .,. 
 
 ** '*.'♦ ^ 
 
 o V 
 
 
 
 
 . . • a <* ''Tvi* .o^" ^ •• • » • A 
 
 
 
BUREAU OF MINES 
 INFORMATION CIRCULAR/ 1989 
 
 3<W 
 
 / 
 
 In Situ Leach Mining 
 
 Proceedings: Bureau of Mines Technology 
 Transfer Seminars, Phoenix, AZ, April 4, 
 and Salt Lake City, UT, April 6, 1989 
 
 UNITED STATES DEPARTMENT OF THE INTERIOR 
 
Information Circular 9216 
 
 In Situ Leach Mining 
 
 Proceedings: Bureau of Mines Technology 
 Transfer Seminars, Phoenix, AZ, April 4, 
 and Salt Lake City, UT, April 6, 1989 
 
 By Staff, Bureau of Mines 
 
 UNITED STATES DEPARTMENT OF THE INTERIOR 
 Manuel J. Lujan, Jr., Secretary 
 
 BUREAU OF MINES 
 T S Ary, Director 
 
1 
 
 
PREFACE 
 
 In April 1989, the U.S. Bureau of Mines held technology transfer seminars on in situ 
 leach mining at Phoenix, AZ, and Salt Lake City, UT. The papers presented at those 
 seminars, as well as three reports of related research, are contained in this Information Cir- 
 cular. The papers highlight various aspects of Bureau research on in situ leach mining, in- 
 cluding a field experiment in copper leaching in Arizona and the use of computer programs 
 to model physical and chemical parameters involved in in situ mining. 
 
 The technology transfer seminar used as a forum for the transfer of this research is one 
 of the many mechanisms used by the Bureau in its efforts to move research developments, 
 technology, and information resulting from its programs into industrial practice and use. To 
 learn more about the Bureau's technology transfer program and how it can be useful to you, 
 please write or telephone: 
 
 U.S. Bureau of Mines 
 
 Office of Technology Transfer 
 
 2401 E Street, NW. 
 
 Washington, DC 20041 
 
 202-634-1224 
 
CONTENTS 
 
 Page 
 
 Preface i 
 
 Abstract 1 
 
 Introduction 2 
 
 In Situ Copper Mining Field Research Project, by Jon K. Ahlness and Daniel J. Millenacker 4 
 
 Generic Design Manual: Cost Model for In Situ Copper Mining, by Joseph M. Pugliese 7 
 
 Introduction to the Environmental Permitting Process for In Situ Copper Mining, by Daniel J. Millenacker 14 
 
 Laboratory Core Leaching and Petrologic Studies To Evaluate Oxide Copper Ores for In Situ Mining, by Steven E. Paulson 
 
 and Harland L. Kuhlman 18 
 
 Methods for Determining the Geologic Structure of an Ore Body as It Relates to In Situ Mining, by Linda J. Dahl 37 
 
 Computer Modeling Applications in the Characterization of In Situ Leach Geochemistry, by Dianne C. Marozas 49 
 
 Modeling Infiltration to Underground Mine Workings During Block-Cave Leaching, by Robert D. Schmidt 58 
 
 HydroLogic: An Intelligent Interface for the MINEFLO Hydrology Model, by Michael E. Salovich 67 
 
 Predicting and Monitoring Leach Solution Flow With Geophysical Techniques, by Daryl R. Tweeton, Calvin L. 
 
 Cumerlato, Jay C. Hanson, and Harland L. Kuhlman 73 
 
 Enhanced Well-Drilling Performance With Chemical Drilling-Fluid Additives, by Patrick A. Tuzinski, John E. Pahlman, 
 
 Pamela J. Watson, and William H. Engelmann 86 
 

 UNIT OF MEASURE ABBREVIATIONS USED IN THIS REPORT 
 
 atm 
 
 atmosphere, standard 
 
 L/min 
 
 liter per minute 
 
 °C 
 
 degree Celsius 
 
 M 
 
 molar 
 
 c/lb 
 
 cent per pound 
 
 m 
 
 meter 
 
 cm 
 
 centimeter 
 
 mD 
 
 millidarcy 
 
 cm 2 
 
 square centimeter 
 
 MHz 
 
 megahertz 
 
 cm 3 
 
 cubic centimeter 
 
 min 
 
 minute 
 
 cP 
 
 centipoise 
 
 mL 
 
 milliliter 
 
 D 
 
 darcy 
 
 mL/min 
 
 milliliter per minute 
 
 ft 
 
 foot 
 
 mm 
 
 millimeter 
 
 ft 2 
 
 square foot 
 
 fim 
 
 micrometer 
 
 ft/ mi 
 
 foot per mile 
 
 /^mho/cm 
 
 micromho per centimeter 
 
 ft/min 
 
 foot per minute 
 
 mm/min 
 
 millimeter per minute 
 
 ft/s 
 
 foot per second 
 
 mol 
 
 mole 
 
 p 
 
 & 
 
 gram 
 
 mol/L 
 
 mole per liter 
 
 gal/d 
 
 gallon per day 
 
 MPa 
 
 megapascal 
 
 (gal/d)/ft 2 
 
 gallon per day per square foot 
 
 ms 
 
 millisecond 
 
 gal/min 
 
 gallon per minute 
 
 mt 
 
 metric ton 
 
 g/cm 3 
 
 gram per cubic centimeter 
 
 mt/yr 
 
 metric ton per year 
 
 g/L 
 
 gram per liter 
 
 mV 
 
 millivolt 
 
 h 
 
 hour 
 
 pet 
 
 percent 
 
 Hz 
 
 hertz 
 
 ppm 
 
 part per million 
 
 in 
 
 inch 
 
 psi 
 
 pound (force) per square inch 
 
 Kbyte 
 
 kilobyte (1,024 bytes) 
 
 rpm 
 
 revolution per minute 
 
 kg 
 
 kilogram 
 
 s 
 
 second 
 
 km/s 
 
 kilometer per second 
 
 st/d 
 
 short ton per day 
 
 kW 
 
 kilowatt 
 
 st/yr 
 
 short ton per year 
 
 L 
 
 liter 
 
 wt pet 
 
 weight percent 
 
 lb 
 
 pound 
 
 yr 
 
 year 
 
 lb/st 
 
 pound per short ton 
 
 $/yr 
 
 dollar per year 
 
 L/h 
 
 liter per hour 
 
 
 
 The Bureau of Mines expressly declares that there are no warranties express or implied that apply to the software 
 discussed in this report. By acceptance and use of said software, which is conveyed to the user without consideration by 
 the Bureau of Mines, the user expressly waives any and all claims for damage and/or suits for or by reason of personal 
 injury, or property damage, including special, consequential or other similar damages arising out of or in any way con- 
 nected with the use of the software discussed herein. 
 
IN SITU LEACH MINING 
 
 Proceedings: Bureau of Mines Technology Transfer Seminars, 
 Phoenix, AZ, April 4, and Salt Lake City, UT, April 6, 1989 
 
 By Staff, Bureau of Mines 
 
 ABSTRACT 
 
 As part of its research on advanced mining technology, the U.S. Bureau of Mines em- 
 phasizes studies on in situ leach mining. This research enables the Bureau to develop 
 technical information that can allow mining companies to adopt in situ leach mining 
 methods. Research results should help to remove the technical and economic barriers cur- 
 rently hindering the development of commercial in situ leach mining operations. 
 
 This report contains papers summarizing results of Bureau research on various aspects 
 of in situ leach mining. A featured topic is the cooperative field experiment of the Bureau 
 and Santa Cruz Joint Venture in a copper deposit near Casa Grande, AZ. Discussions in- 
 clude an overview of the entire project, cost modeling, environmental permitting, laboratory 
 core-leaching experiments, petrographic studies, and an evaluation of the geologic structure 
 in the region. In addition, papers describe hydrologic modeling and geophysical studies of 
 solution flow during leaching. Preliminary geochemical modeling studies are also sum- 
 marized. A final paper describes a nonionic chemical additive with potential for significantly 
 reducing the high costs of drilling in in situ leach mining. 
 
INTRODUCTION 
 
 During the 1980's, the U.S. minerals industry found it dif- 
 ficult to compete with foreign mining companies in world 
 markets. The U.S. copper companies were high-cost producers 
 in a shrinking world market. Copper mining companies began 
 restructuring in order to survive. The restructuring included 
 considerable closures to match production with lower demand. 
 Closing inefficient operations and adopting new processing 
 technology enabled companies to boost productivity. This pro- 
 ductivity, coupled with recent changes in the foreign exchange 
 rates of the U.S. dollar, returned many U.S. copper producers 
 to better fiscal health. 
 
 The present favorable market conditions will not, 
 however, last forever. Major foreign companies also cut costs 
 with considerable restructuring. U.S. companies remain high- 
 cost producers. For continued existence, the U.S. companies 
 must be able to sell mineral products at the same price as do 
 foreign companies and still make a profit. Low ore grades, 
 high labor rates, strict regulatory constraints, higher concern 
 over miner health and safety, and more costly environmental 
 safeguards all force U.S. mining costs upward. During the 
 next market contraction, U.S. companies will again be par- 
 ticularly vulnerable. Most of the restructuring during the early 
 1980's was a one-time cure; similar actions will not likely be 
 available to these companies in another down market. The 
 Bureau of Mines believes that the industry needs new 
 technology to overcome the high cost of conventional mining 
 operations. 
 
 The Bureau recently expanded its efforts to provide 
 technology that will help the minerals industry develop the 
 necessary new mining and processing methods. Bold new 
 methods of mining that improve worker productivity, 
 minimize worker exposure to hazardous underground condi- 
 
 tions, profoundly alter environmental impacts, and reduce 
 waste have moved to the forefront of Bureau research. 
 
 One of these new methods is called in situ leach mining. In 
 situ leach mining involves circulating dilute chemical solutions 
 through an ore deposit to dissolve the target metals. The solu- 
 tions may be applied to the top of the ore body and allowed to 
 seep down through it by gravity. Or the solutions may be 
 forced through the ore between series of vertical injection and 
 recovery wells (fig. 1). This method is sometimes referred to as 
 "true" in situ leach mining. In yet another variation, leach 
 solutions are trickled over rubbled ore in stopes or other 
 underground openings. Regardless of the application method, 
 the solutions containing the dissolved metals are then pumped 
 out of the ground and to a plant for metal recovery. After 
 removing the metals from solution, operators recycle the solu- 
 tion back to the ore body. In situ leach mining thus combines 
 mining and processing operations. 
 
 In situ leach mining offers several economic, safety, and 
 environmental advantages over conventional mining. The 
 Bureau believes that the method requires less capital invest- 
 ment. By eliminating ore extraction and crushing, in situ leach 
 mining also saves labor and energy costs. Other benefits in- 
 clude fewer health and safety risks as miners will not be ex- 
 posed to the hazards of underground mining. The technique 
 disrupts the environment less than does conventional mining; 
 in situ leach mining removes little from the ground except the 
 target metals. It also leaves a mine site that is easily reclaimed 
 to its original condition by capping wells and dismantling the 
 processing plant. 
 
 The big question with the method is the metal recovery. 
 Operators experienced low recovery from shallow rubbled 
 deposits that were leached in the 1970's. Bureau experiments, 
 
 Injection 
 well 
 
 Recovery 
 well 
 
 Processing plant 
 
 Ore mineralization 
 
 Leach solution 
 
 Figure 1.— Cross section of an in situ leach mining system in which the ore is leached via a system of vertical wells. 
 
however, indicate that certain deposits should leach well. Such 
 results underscore the necessity of conducting the tests and 
 analyses described in this Information Circular before deciding 
 whether to leach an ore body. The main environmental con- 
 cern in in situ leach mining is the control and collection of 
 leach solutions within the ore body. Developing control 
 technology has been a major focus of the Bureau's research. 
 The overall advantages provide two very important 
 benefits to mine operators, particularly to copper mining com- 
 panies: 
 
 1. Production costs will be reduced, thereby making 
 U.S. producers more competitive in the world market and 
 reducing foreign import dependence. 
 
 2. Copper can be recovered from small and/or low- 
 grade deposits that are currently not economical to mine by 
 conventional methods. 
 
 The Bureau began studying methods of leaching ore left 
 in the ground during the early 1970's. The early studies concen- 
 trated on fracturing shallow copper ore bodies to improve the 
 flow of leaching solutions. By the late 1970's, the Bureau 
 focused its in situ leaching research on uranium deposits in 
 shallow, permeable sandstones. These studies helped stimulate 
 commercial operations. Early in the 1980's, the Bureau began 
 evaluating techniques for in-place leaching of other metals, 
 such as manganese and gold. Column leaching experiments, 
 geologic characterization, and fluid flow modeling 
 predominated. 
 
 In 1986, the Bureau began a research program emphasiz- 
 ing in situ leach mining of shallow to moderately deep copper 
 oxide ores. Its goal is to provide industry with the technology 
 to design an in situ copper leach mining operation for any 
 specific deposit. A combined laboratory and field testing ef- 
 fort was started to obtain site-specific data from two oxide 
 copper deposits near Casa Grande, AZ. In support of the field 
 work, the Bureau is conducting whole-core laboratory 
 leaching experiments to determine the effects of microstruc- 
 ture, ore and gangue-mineral relationships, flow rates, and 
 other factors on copper recovery and acid consumption. 
 
 The last Bureau of Mines technology transfer seminar on 
 in situ leach mining technology was held in Denver, CO, in 
 1981. It emphasized uranium. This Information Circular con- 
 tains the results of the Bureau's copper in situ mining research 
 
 over the past 5 yr and also includes the results of relevant cop- 
 per block-cave leaching research. 
 
 The first five papers discuss several important aspects of 
 the Bureau's cooperative in situ copper mining research near 
 Casa Grande, AZ. The first presents an overview of the 
 research. The second describes the leach mine design manual 
 with cost model developed by Science Applications Interna- 
 tional Corp. under Bureau contract J0267001. The third ex- 
 plains the potential environmental impact of in situ copper 
 leach mining and the requirements of regulatory agencies in 
 Arizona. The fourth describes the laboratory tests necessary to 
 evaluate copper deposits for leaching. The fifth paper 
 discusses the potential effects of the geologic structures on 
 leaching results at the deposit chosen for the experiment near 
 Casa Grande, AZ. 
 
 The next paper discusses the potential applications of 
 geochemical modeling for estimating chemical reactions that 
 will occur during in situ leaching. The next two papers discuss 
 the importance of hydrologic modeling studies in helping to 
 understand the fluid flow during block-cave leaching. These 
 papers describe the recently developed MINEFLO computer 
 model and its usefulness in evaluating complex hydrologic 
 problems. 
 
 The Bureau is also investigating potential applications of 
 geophysical systems to monitor leachate during in situ mining. 
 Tomographic reconstruction of seismic cross-hole data to help 
 detect leach solutions is described in the next paper. 
 
 The final paper discusses the applicability of nonionic ad- 
 ditives to improve drilling performance. Commercial in situ 
 leach mining operators will spend much money on drilling 
 wells. Successful application of this drilling additive will save 
 considerable costs. 
 
 Additional technical information can be obtained by con- 
 tacting William C. Larson, supervisor of the Advanced Mini- 
 ing Division at the Twin Cities Research Center, 5629 Min- 
 nehaha Avenue S., Minneapolis, MN 55417, or by calling 
 612-725-4690. Information concerning the Bureau's research 
 programs on Advanced Mining Systems can be obtained by 
 contacting Peter G. Chamberlain, program manager, 2401 E 
 Street, NW., Washington, DC 20241, or by calling 
 202-634-9885. 
 
IN SITU COPPER MINING FIELD RESEARCH PROJECT 
 
 By Jon K. Ahlness 1 and Daniel J. Millenacker 2 
 
 ABSTRACT 
 
 To help meet the need of the domestic copper mining industry for new technology, the 
 U.S. Bureau of Mines is conducting a field test of in situ copper mining at the Santa Cruz 
 deposit near Casa Grande, AZ. A preliminary commercial mine design has been prepared for 
 the site, based on a costing manual developed under Bureau contract. This paper describes 
 the site and the well field design, as well as the steps involved in conducting the field test. 
 These steps include obtaining geologic and hydrologic data on the ore zone, applying for en- 
 vironmental permits, constructing the well field and a solvent extraction-electrowinning 
 plant, startup, leaching and processing operations, and well field decommissioning. 
 Geologic characterization of drill core from the deposit is currently under way, along with 
 whole-core laboratory leaching experiments. Test site geology and hydrology data will be 
 gathered for development of final project designs and environmental permit applications. 
 
 INTRODUCTION 
 
 In recent years, copper production in the United States 
 has experienced a significant decline; however, it is once again 
 increasing, owing to the recent increase in copper price. For 
 example, domestic production of copper in the years 1977-81 
 averaged 1.38 million mt/yr, while in the following 6 yr 
 (1982-87) production averaged only 1.14 million mt/yr (a 
 decrease of 18 pet) (7). 3 Production in 1987 was up to 1.27 
 million mt, which is still down 8 pet from the 1977-81 average. 
 Employment in the operating mines and mills fell from over 
 30,000 people in 1981 to less than 10,000 in 1987. 
 
 These decreases have occurred at a time when the United 
 States is importing 27 pet of its copper requirements from 
 Chile, Canada, Peru, Mexico, and other countries (2). Several 
 factors are responsible for the drop in domestic production, 
 including foreign competition, depletion of accessible, higher 
 grade domestic reserves, and environmental considerations. 
 Competition from developing countries is particularly intense 
 in those cases where foreign reserves are higher grade, labor 
 costs are low, and mines are nationalized or subsidized. 
 
 Maintaining a viable copper industry in the United States 
 is dependent on several factors. First, it is up to the copper 
 producers to maintain the lower operating costs and improved 
 efficiencies that have been developed in recent years. Second, 
 the development of new lower cost, environmentally sound 
 technologies (such as in situ mining) are required to maintain 
 and, it is hoped, to enhance domestic copper production from 
 the remaining small, low-grade, and/or deep deposits. The 
 
 'Supervisory physical scientist. 
 
 2 Hydrologist. 
 Twin Cities Research Center, U.S. Bureau of Mines, Minneapolis, MN. 
 
 J Italic numbers in parentheses refer to items in the list of references at the end 
 of this paper. 
 
 Bureau of Mines is conducting research to develop in situ min- 
 ing technology to meet this need. 
 
 In situ mining has been used on a commercial basis since 
 the mid-1970's to produce uranium from porous sandstone 
 deposits in Texas and Wyoming (5). In situ methods have also 
 been used to recover copper from low-grade ore and waste 
 rock in old open pit mines, block-caved zones, and backfilled 
 stopes as an afterthought to conventional mining (4). Ore can 
 also be placed on an impermeable surface in heaps and leached 
 by sprinkling a leach solution (lixiviant) over the surface of the 
 pile. Heap leaching is actively being used for extracting gold 
 and silver from low-grade ore (J). Recovery of the pregnant 
 leach solution in each of these operations occurs by fluid 
 movement to a recovery well or underground sump, or to a 
 surface collection pond. 
 
 The Bureau's research objective is to stimulate domestic 
 production of copper by the private sector, using in situ min- 
 ing methods. To achieve this goal, the Bureau intends to pro- 
 vide industry with the means to design the most economical in 
 situ mining operation for any specific copper oxide ore 
 deposit. A "Draft Generic In Situ Copper Mine Design 
 Manual" has been prepared by Science Applications Interna- 
 tional Corp. (SAIC), McLean, VA, under Bureau contract 
 J0267001 (6). Subcontractors that assisted SAIC included Ray 
 V. Huff and Associates, Golden, CO, Davy McKee Corp., San 
 Ramon, CA, and Sergent, Hauskins, and Beckwith, Phoenix, 
 AZ. This draft manual serves as a source document for 
 developing a commercial-scale design for an in situ operation. 
 The draft manual contains 
 
 1 . A listing and description of the design elements and 
 procedures for each component of an in situ copper mining 
 system, and a method of costing individual components. 
 
 2. A method of identifying the best of 42 possible in situ 
 mining scenarios, which maximizes the discounted-cash-flow 
 
rate of return (DCFROR) for commercial operation for a 
 specific site. 
 
 3. A listing of the site-specific parameters that must be 
 quantified to develop a commercial design and a description of 
 laboratory and field tests to measure these parameters. 
 
 4. A cash-flow model and computer program to con- 
 duct a DCFROR economic analysis incorporating all capital 
 and operating expenses associated with the well field, a solvent 
 extraction-electrowinning plant (SX-EW), and environmental 
 permitting. 
 
 5. A description of the environmental procedures, 
 specifications, designs, and costs for permitting, monitoring, 
 and restoration. 
 
 The generic approach provides flexibility in developing 
 designs and costs for a wide range of deposit characteristics 
 and operational parameters. To verify the draft manual and 
 demonstrate the technical feasibility of in situ copper extrac- 
 tion, a field test is being conducted in a copper oxide ore 
 deposit in Arizona. 
 
 IN SITU COPPER MINING FIELD RESEARCH PROJECT 
 
 The Bureau is conducting the field research project in 
 cooperation with the Santa Cruz Joint Venture (a mining part- 
 nership formed between ASARCO Santa Cruz Inc., a sub- 
 sidiary of ASARCO Incorporated, and Freeport Copper Co., 
 a subsidiary of Freeport-McMoRan Gold Co.). This research 
 is being conducted at the Santa Cruz site, 7 miles west of Casa 
 Grande, AZ. A commercial mine design has been prepared for 
 this site using the algorithms and model provided in the "Draft 
 Generic In Situ Copper Mine Design Manual" (6). Actual field 
 testing will utilize the unit-cell concept. A unit cell is a small 
 well field constructed to commercial-scale well size and spacing 
 specifications. Since the design and operational parameters are 
 comparable to a commercial-scale operation, the field test will 
 allow validation of the algorithms used in preparation of the 
 commercial design and will also demonstrate whether an in situ 
 
 facility can be operated as designed and within budget for a 
 sustained period of time. Validation of the model will consist 
 of comparison of actual versus calculated construction and 
 operating costs, as well as verification of technical parameters 
 (flow rate, fluid control, etc.). 
 
 The Santa Cruz site (fig. 1) is undeveloped and contains a 
 mineralized zone with estimated copper oxide reserves of 97 
 million mt at an average grade of 0.7 pet acid-soluble copper. 
 Chrysocolla and atacamite are the predominant copper oxides 
 present in the granite and porphyry host rocks. Available field 
 data indicate an in situ permeability range of 5 to 20 mD. The 
 bottom of the mineralized zone is located at an average depth 
 of 2,200 ft below the surface, with an average thickness of 345 
 ft. Surface terrain is essentially flat lying with relief of 10 
 ft/mi. Access to the site is provided by well-maintained State 
 
 Figure 1.— Drill rig on the Santa Cruz site. 
 
and township roadways. Geologic characterization of 
 available drill core is presently under way, along with 
 laboratory whole-core leaching experiments. 
 
 A well field of two connected five-spot patterns has been 
 selected as the design basis for the field test. Corner wells in 
 the pattern will serve production needs, and center wells will be 
 used for injection. Producer-to-producer well spacing is 
 designed to be 160 ft. The selected well spacing and calculated 
 flow rate allow a sufficiently large area to be leached to deter- 
 mine copper recovery, copper loading, and residence time ex- 
 pected for a commercial operation in a representative geologic 
 environment. Short-radius, propped hydraulic fractures may 
 be necessary to increase the effective wellbore radius. 
 
 The entire field test is scheduled to run for a 48-month 
 period. Actual fluid injection-recovery operations will, 
 however, run for only 18 months. Construction and decom- 
 missioning activities account for the balance. The 18-month 
 production period will allow sufficient time to assess opera- 
 tional problems and to determine if copper loadings can be 
 maintained over an extended period. 
 
 Several steps are involved in conducting the field test. The 
 first of these is to obtain hydrologic and geologic data from 
 the ore zone and vicinity. A test program will be initiated to 
 assess, first, flow characteristics within the selected ore block 
 and, subsequently, flow control within a well pattern. Initially, 
 two holes will be drilled to obtain core and to allow 
 geophysical logging and open hole testing. Testing will provide 
 data on porosity and permeability variations within the vertical 
 ore interval. Upon completion of testing, two additional wells 
 will be completed within 50 ft of the original wells. Tests will 
 be conducted to determine orientation of the natural fractures 
 (and effectiveness of wellbore stimulation) and to assess fluid 
 movement and control within the well pattern. These data will 
 be used for orienting the five-spot patterns. 
 
 Environmental permitting activities will be initiated dur- 
 ing the site investigation stage. Required ground water protec- 
 
 tion permit applications will be prepared and submitted to the 
 appropriate Federal and State regulatory agencies. Permitting 
 is the mechanism used by the regulatory agencies to ensure that 
 applicable water quality standards will be achieved in the areas 
 adjacent to the facility during both operation and decommis- 
 sioning. Requirements specific to protection of air resources 
 will additionally need to be addressed at the local government 
 level. Ground water protection requirements applicable to an 
 in situ mining operation in Arizona have previously been 
 discussed by Weeks (7). 
 
 Upon completion of site characterization work, the well 
 field will be constructed. The preliminary well field design 
 assumes a 50-pct copper recovery over an approximate leach 
 interval of 325 ft. The maximum flow rate expected for each of 
 the injection wells is 40 gal/min. The pregnant solution grade 
 is estimated to be 10 g/L. The final design (and parameter 
 values) may be modified as field data become available from 
 the site investigation. 
 
 A solvent extraction-electrowinning (SX-EW) plant will 
 be constructed upon completion of, or in conjunction with, 
 well field installation. The plant will extract copper from the 
 pregnant liquor, plate it as cathode copper, and recondition 
 the lixiviant. The facility design will achieve an anticipated 
 copper production rate of 1,095 st/yr cathode copper. 
 
 Startup operations will initially occur over a 2-month 
 period to establish flow rate equilibrium, determine steady- 
 state solution concentrations, and identify and troubleshoot 
 equipment problems. Upon completion of well field startup, 
 the lixiviant will be injected and actual leaching operations will 
 commence. Produced fluids will be pumped to the SX-EW 
 plant for processing. 
 
 Well field decommissioning will be initiated at the com- 
 pletion of leaching. During this time, the surface plant will be 
 dismantled and containment structures will be removed from 
 the site. Wells will be abandoned and the leach interval 
 restored as required by the regulatory agencies. 
 
 SUMMARY 
 
 The Bureau is conducting research to evaluate in situ min- 
 ing as a method to maximize the probability of the domestic 
 production of copper from small, deep, and/or low-grade cop- 
 per oxide ore deposits. The Bureau intends to provide the U.S. 
 copper mining industry with the technical and economic data 
 necessary to design an in situ mining operation for any specific 
 copper oxide ore deposit. A "Draft Generic In Situ Copper 
 Mine Design Manual," prepared for the Bureau by Science Ap- 
 
 plications International Corp., serves as a source document 
 for developing a commercial design for an in situ operation. 
 To verify the draft manual and demonstrate the technical 
 feasibility of in situ extraction of copper, a field test is being 
 conducted in a copper oxide ore deposit in Arizona. Field 
 research is expected to continue for the next several years. 
 Research results will be made available to the mining com- 
 munity as significant milestones are attained. 
 
 REFERENCES 
 
 1. U.S. Bureau of Mines. Mineral Commodity Summaries 1977-87. 
 Section on Copper. 
 
 2. U.S. Department of the Interior. The Mineral Position of the 
 United States— 1987. Annual Report of the Secretary of the Interior 
 Under the Mining and Minerals Policy Act of 1970. BuMines Spec. 
 Publ., 1987, 77 pp. 
 
 3. Larson, W. C. Uranium In Situ Leach Mining in the United 
 States. BuMines IC 8777, 1978, 68 pp. 
 
 4. Ahlness, J. K., and M. G. Pojar. In Situ Copper Leaching in the 
 United States: Case Histories of Operations. BuMines IC 8961, 1983, 
 37 pp. 
 
 5. Chamberlain, P. G., and M. G. Pojar. Gold and Silver Leaching 
 Practices in the United States. BuMines IC 8969, 1984, 47 pp. 
 
 6. Davidson, D. H., R. V. Huff, R. E. Weeks, and J. F. Edwards. 
 Generic In Situ Copper Mine Design Manual (contract J0267001, 
 Science Applications Int. Corp.). Volume II: Draft Generic In Situ 
 Copper Mine Design Manual. 1988, 454 pp.; for inf., contact J. K. 
 Ahlness, TPO, Twin Cities Res. Cent., BuMines, Minneapolis, MN. 
 
 7. Weeks, R. E., and D. J. Millenacker. Environmental Permitting 
 Considerations for True In Situ Copper Mining in the State of 
 Arizona. Soc. Min. Eng. AIME preprint 88-196, 1988, 7 pp. 
 
GENERIC DESIGN MANUAL: COST MODEL FOR IN SITU COPPER MINING 
 
 By Joseph M. Pugliese 1 
 
 ABSTRACT 
 
 The U.S. Bureau of Mines has provided a cost model for in situ copper mining that specifies 
 (1) site-specific parameters, which must be quantified, (2) a method for individual site design based 
 on these site parameters, and (3) a procedure for assessing the economic viability of the site-specific 
 design. This paper describes the model, a menu-driven computer program that performs calcula- 
 tions for developing commercial design specifications as well as capital and operating costs. The 
 program also provides discounted-cash-flow rate of return (DCFROR) and sensitivity analyses for 
 a true in situ copper mining operation at any specified site. This model can be used to compare the 
 relative costs of different in situ mining scenarios. It provides a systematic method for assessing the 
 commercial feasibility of applying true in situ copper mining at a selected site. 
 
 INTRODUCTION 
 
 The Bureau of Mines believes that the competitive position of 
 domestically produced copper can be significantly improved with 
 the application of in situ mining techniques. A long-term objective 
 of the Bureau is to increase the probability of the domestic produc- 
 tion of copper by the private sector, using in situ means. As part of 
 the effort to meet this objective, the Bureau conducted research on 
 cost modeling of in situ copper mining to provide the mining in- 
 dustry with the means to design the most economically successful in 
 situ copper operation for any specific deposit. The research was 
 conducted by Science Applications International Corp. (SAIC), 
 McLean, VA, under Bureau contract J0267001. The contractor had 
 three subcontractors to aid in conducting the work: Ray V. Huff 
 and Associates, Inc., Golden, CO; Davy McKee Corp., San 
 Ramon, CA; and Sergent, Hauskins, and Beckwith, Phoenix, AZ. 
 
 The contract effort started in October 1986, and the contractor's 
 final report was completed in April 1988. ; 
 
 The paper deals with the economic analysis found in Volume 
 II of SAIC's final report and the commercial-scale operation exam- 
 ple found in Volume IV of that report. Mention of these volumes 
 and of SAIC's final report throughout the paper will be made with 
 the understanding that the work is cited in footnote 2. 
 
 The algorithms developed from this research are combined in- 
 to the cost-modeling computer program for the design and 
 economic analysis of an in situ copper mining operation. The com- 
 puter program is on an IBM-compatible microcomputer; the 
 minimum hardware requirements for running the program are a 
 360-Kbyte memory and a math coprocessor. 
 
 COST MODEL ASPECTS 
 
 The cost model may be used to evaluate 42 in situ mining 
 scenarios. Three basic characteristics are used to define each of the 
 42 scenarios. 
 
 The first characteristic addresses the way in which the ore body 
 and adjacent rock will be penetrated, that is, the method of deposit 
 access. Seven types of deposit access treatments are— 
 
 1 . Drilling from the surface; 
 
 2. Drilling from existing underground entries; 
 
 3. Developing new underground entries, with drilling from 
 those entries; 
 
 4. Combinations of types 1 and 2; 
 
 5. Combinations of types 1 and 3; 
 
 6. Combinations of types 2 and 3; and 
 
 7. Combinations of types 1, 2, and 3. 
 
 The second characteristic involves the treatment of the 
 rock matrix from within the borehole, or matrix modification. 
 The three types of treatments considered by the cost model 
 are — 
 
 A. No modification; 
 
 B. Modification using hydraulic fracturing; and 
 
 C. Modification using explosives in the borehole. 
 
 'Mining engineer, Twin Cities Research Center, U.S. Bureau of Mines, Minneapolis, 
 
 MN. 
 
 -Davidson, D. H., R. V. Huff, R. E. Weeks, and J 
 per Mine Design Manual (contract J0267001, Science 
 I: Executive Summary, Apr. 1988, 93 pp.; Volume 
 Mine Design Manual, Apr. 1988, 454 pp.; Volume III 
 Design of Commercial Scale Operation, Apr. 1988, 
 Field Experiment and Design of Commercial Scale 
 Volume V: Field Testing at the Santa Cruz Site, Apr. 
 K. Ahlness, TPO, Twin Cities Res. Cent., BuMines, 
 
 F. Edwards. Generic In Situ Cop- 
 Applications Int. Corp.). Volume 
 II: Draft Generic In Situ Copper 
 Lakeshore Field Experiment and 
 371 pp.; Volume IV: Sanla Cruz 
 Operation, Apr. 1988, 385 pp.; 
 1988, 1 14 pp.; for inf., contact J. 
 Minneapolis, MN. 
 
The [bird characteristic involves the initial hydrologic set- 
 ting of the ore body. Two types of treatments are — 
 
 i. No hydrologic modification (already saturated); and 
 
 ii. Hydrologic modification (unsaturated conditions 
 originally). 
 
 Type ii treatment indicates that the ore body should be 
 saturated before leaching begins. 
 
 Forty-two mining scenario permutations can be deter- 
 mined from the types of treatments shown above. An example 
 of a mining scenario permutation would be l.A.i., that is, 
 drilling from the surface, no matrix modification, and an ore 
 body that is initially saturated. The approach used to estimate 
 which of the 42 possible mining scenarios may be best for a 
 specific site is found in Volume II, chapter 9, of SAIC's final 
 report. 
 
 The cost model does the following: 
 
 1. Develops detailed capital and operating costs for any 
 proposed copper in situ mining project. 
 
 2. Solves for pretax discounted-cash-flow rate of return 
 (DCFROR) or selling price. 
 
 3. Contains cost data (default values) from a broad ex- 
 perience base. 
 
 4. Allows for cost analyses where either minimum or 
 maximum data are available. 
 
 5. Accepts user-specified information. 
 
 6. Conducts cash-flow analyses. 
 
 7. Can be used for sensitivity analyses. 
 
 Each of these features will be covered in the following sec- 
 tions. 
 
 COST MODEL INPUT PARAMETERS 
 
 Input parameters are listed under the following 
 categories: (1) business-related parameters, (2) site-specific ore 
 body and well field characteristics, (3) copper leaching, (4) 
 program control, (5) well system specifications and costs, (6) 
 surface plant specifications and costs, and (7) environmental 
 costs. The first three categories contain those input parameters 
 that must be utilized to specify a certain mining scenario. The 
 fourth category contains the input parameters that must be 
 specified to control the printout of output data. The last three 
 categories contain the input parameters that correspond to 
 cost values and design specifications that translate a specific 
 mining scenario into capital and operating costs. 
 
 Category 1 consists of inflation indexes for plant and min- 
 ing construction costs, copper selling price or rate of return, 
 capital expenditure schedule, plant life, and copper produc- 
 tion. 
 
 Category 2 contains site-specific parameters that need to 
 be considered in design of the wells and well pattern, such as 
 depth to the bottom of the ore zone, thickness of the ore zone, 
 leach interval, ore grade, permeability and porosity of the ore 
 zone, type of mining access (surface or underground), status 
 of underground mine access (existing or nonexisting), well 
 
 field flow pattern, well spacing, type of well completion and 
 stimulation, and ore body water saturation. 
 
 Category 3 includes site-specific parameters that relate to 
 pregnant liquor copper loading, acid consumption by gangue 
 and copper minerals, and copper recovery efficiency. 
 
 Category 4 contains the parameters that control whether 
 to print out calculated variables, annual cash-flow summary, 
 and/or all input parameters. 
 
 Category 5 includes unit costs used to convert well system 
 input parameters to output specifications and costs. Examples 
 of these parameters are the cost of preparing the surface drill 
 site; wellhead, packer, tubing, and casing costs; and hydraulic 
 fracturing and explosive stimulation costs. 
 
 Category 6 includes unit costs used to compute surface 
 plant specifications and costs. These unit costs apply to such 
 input parameters as pipes (lines), extractant, and diluent. Flow 
 rate through the lines is covered under this category. 
 
 Category 7 addresses environmental aspects of the in situ 
 mining operation. Included here are monitoring well unit cost, 
 cost of initial environmental permitting, annual cost of en- 
 vironmental monitoring, and well field restoration cost. 
 
 COST MODEL OUTPUT PARAMETERS 
 
 Output parameters related to the well system, the surface 
 plant, and the environment are grouped under the following 
 categories: (1) design variables, (2) capital costs, and (3) 
 operating costs. 
 
 Category I includes leach area of unit cell; area of well 
 field; injection pressure at top of ore zone; injection flow rate; 
 total system flow rate; number of unit cells; well field life; in- 
 jection acid concentration; acid and lime consumption; metal 
 sulfates produced per year; number of injection and total wells 
 (excluding monitor wells); well field solution transfer piping 
 length and diameter; well-field-to-plant transfer piping 
 diameter; total weight of copper in unit cell; fraction of copper 
 recovered from the zone being leached; number of mine and 
 
 plant operators; tubing, casing, and wellhole diameters; and 
 injection and production pump power requirements. 
 
 Category 2 includes well site preparation cost; well casing, 
 cementing, drilling, completion, and stimulation costs; well 
 logging cost; well injection and production equipment costs; 
 cost per fan for fan wells; total surface facility capital cost; 
 total cost of the solution transfer system; total well field 
 capital cost; cost of initial environmental permitting; cost of 
 monitoring wells; and well field restoration cost. 
 
 Category 3 includes chemical and consumable costs, labor 
 cost, cost of utilities, cost of environmental monitoring, and 
 maintenance costs. 
 
CASH-FLOW ANALYSIS 
 
 This cost model can be used to conduct a cash-flow 
 analysis, using the DCFROR. A thorough discussion on 
 DCFROR is found in Stermole, ! and a brief summary from 
 that reference is presented in this section. The term "discount" 
 is generally considered to be synonymous with "present worth" 
 in economic evaluation work. The term "cash flow" is used to 
 refer to the net inflow or outflow of money that occurs during 
 a specified operating period such as a month or a year. The 
 term "discounted cash flow" evolved from the idea that in- 
 vestors often handle the time value of money using discounted 
 (present value) calculations. That is, in evaluating the mining 
 operation's economic potential, the investors used positive and 
 negative cash flows at their present worth anticipated from an 
 investment. 
 
 According to the Stermole text, the investment cash flow 
 in any one year represents the net difference between inflows 
 of money from all sources minus investment outflows of 
 money from all sources. This cash flow is calculated by 
 
 Cfn 
 
 R - CC - OC, 
 
 where cf n 
 
 R 
 
 CC 
 
 and OC 
 
 net cash flow in each project year, $/yr, 
 revenue, $/yr, 
 capital costs, $/yr, 
 operating costs, $/yr. 
 
 The cost model developed by SAIC for the Bureau does not 
 consider taxes, royalties, or predevelopment costs in the calcu- 
 lations because of the generic nature of the cost model and the 
 variation in these costs among States and companies. In addi- 
 tion, the model does not include sunk costs, such as for ex- 
 ploration and property acquisition; depletion allowances; 
 depreciation; and salvage value. The cost model should be 
 used to make relative production cost comparisons among dif- 
 ferent commercial in situ copper mining scenarios for the pur- 
 pose of selecting the one most economical. The cost model can 
 form the basis for a mining company to develop its own ab- 
 solute cost model. 
 
 In using DCFROR, the term "net present value" (NPV) 
 needs to be defined. ' The NPV is the cumulative present worth 
 of positive and negative investment cash flow using a specified 
 discount rate to handle the time value of money. As an equa- 
 tion: 
 
 NPV: 
 
 E Cf n , 
 
 n = (1 + i)" 
 
 where NPV = net present value in year of calculation, $/yr, 
 cf n = net cash flow in each project year, $/yr, 
 n = year number (first year number = 0), 
 N = total project life (plant life + construction 
 period), yr, 
 and i = discount rate, pet (expressed as a decimal in 
 equation). 
 
 The DCFROR is the value of "i" that makes NPV equal to 
 zero. In the cost model, the DCFROR is used to calculate cop- 
 
 Stermole, F. J., and J. M. Stermole. Economic Evaluation and Investment 
 Decision Methods. Investment Evaluations Corp., Golden, CO, 1987, 479 pp. 
 'Work cited in footnote 3. 
 
 per selling price if DCFROR is specified. Conversely, copper 
 selling price is used to calculate DCFROR if copper selling 
 price is specified. 
 
 The information required to define a commercial in situ 
 copper mining operation for the Santa Cruz site near Casa 
 Grande, AZ, for which design specifications and costs can be 
 developed is found in Volume IV of the contractor's final 
 report. In September 1986, the Bureau and the Santa Cruz 
 Joint Venture (ASARCO Santa Cruz Inc. and Freeport Cop- 
 per Co.) agreed to cooperate for the purpose of conducting 
 data acquisition field tests and developing in situ mine designs 
 for the Santa Cruz site. In May 1987, the Bureau contractor, 
 SAIC, met with Asarco staff and obtained business decision 
 and ore deposit information for use in defining a commercial- 
 scale operation. In defining this operation, a scoping analysis 
 was employed that did not use site-specific construction data 
 pertaining to local Casa Grande conditions and did not con- 
 sider local taxes and utility costs. 
 
 Based on SAIC's analysis, the initial mining scenario (best 
 from information available at the time) involved vertical wells 
 drilled from the surface. Vertical wells seemed to be a reason- 
 able place to start since no underground workings existed. For 
 this Santa Cruz commercial-scale operation example, a five- 
 spot pattern with a radial flow pattern to be used for leaching 
 was developed (fig. 1). In this figure, the central injection well 
 region is an expanded view showing the wellbore radius prior 
 to small-radius hydraulic fracturing and the radius after this 
 type of stimulation. A value of 8 ft is assumed as the effective 
 wellbore radius, or radius after stimulation. The four producer 
 wells are at the corners of the pattern and have a spacing (S) of 
 160 ft. The depth to the bottom of the ore zone is 2,136 ft, and 
 the ore zone thickness is 345 ft. The ore body is in a saturated 
 state. The copper production rate (plant capacity) is estimated 
 at 23,600 st/yr, and the plant life is 15 yr. The grade of ore is 
 0.72 pet, and the permeability of the ore zone is estimated at 2 
 mD. Underground copper recovery is assumed at 50 pet, and 
 the copper loading is estimated to be 10 g/L. The other input 
 and calculated parameter values for this Santa Cruz 
 commercial-scale example can be found in Volume IV of 
 SAIC's final report. 
 
 Table 1 shows the annual cash flows for the commercial- 
 scale copper in situ mining operation example at the Santa 
 Cruz site and is taken from Volume IV of SAIC's final report. 
 The DCFROR is 20 pet, which resulted in a copper selling 
 price of 51 (50.94) c/lb. The project life is 18 yr, including the 
 first 3 yr for construction, which results in a plant life of 15 yr. 
 As well fields are depleted of copper, they are replaced with 
 new well fields. In this example, well field construction is esti- 
 mated to take 1 !/2 yr, and the well field life is 3 yr. As each well 
 field is put into operation, a well priming time is required. For 
 the years 4, 7, 10, 13, and 16, the lower production values 
 shown reflect this priming effect. That is, each time a new well 
 field is started up, no salable cathode copper is assumed to be 
 available until 1 pore volume displacement has taken place. 
 This lost production of cathode copper is subtracted from the 
 normal annual production. The total operating and mainte- 
 nance costs are incurred from the first year of production (year 
 4) through the last year of project life (year 18). 
 
10 
 
 KEY 
 
 Wei I bore radius prior to stimulation 
 T^ ] Effeotlve wellbore radius 
 = S > Radial flow pattsrn 
 
 ^P Produoar well 
 3 Produoer-well spacing 
 
 Injeotlon-well 
 region boundary, 
 
 3 
 
 Not to aoale 
 
 Figure 1.— Plan view of five-spot pattern, commercial-scale operation example. 
 
11 
 
 Table 1.— Annual cash flow, Santa Cruz commercial-scale operation example 
 
 Approximate Production Costs, 10 3 $/yr Net cash Discounted 
 
 Year production 1 revenue 2 Plant capital Well field Total operating flow cash flow 3 
 (Y), (R), cost capital and maintenance (cf n ), (DCF), 
 10 3 st/yr 10 3 $/yr (PCC) cost (WCC) cost (TOC) 10 3 $/yr 10 3 $/yr 
 
 1 4,945 - 4,945 - 4,945 
 
 2 5,374 2,187 -7,561 -6,301 
 
 3 11,179 4,245 -15,424 -10,711 
 
 4 18 18,286 11,908 6,378 3,691 
 
 5 22 21,928 2,217 11,908 7,803 3,763 
 
 6 22 21,928 4,304 11,908 5,716 2,297 
 
 7 18 18,286 11,908 6,378 2,136 
 
 8 22 21,928 2,217 11,908 7,803 2,178 
 
 9 22 21,928 4,304 11,908 5,716 1,329 
 
 10 18 18,286 11,908 6,378 1,236 
 
 11 22 21,928 2,217 11,908 7,803 1,260 
 
 12 22 21,928 4,304 11,908 5,716 769.3 
 
 13 18 18,286 11,908 6,378 715.3 
 
 14 22 21,928 2,217 11,908 7,803 729.3 
 
 15 22 21,928 4,304 11,908 5,716 445.2 
 
 16 18 18,286 11,908 6,378 414.0 
 
 17 22 21,928 11,908 10,020 542.0 
 
 18 22 21,928 11,908 10,020 451.6 
 
 ' Actual production: priming years, 17,949 st/yr; nonpriming years, 21,523 st/yr. On-stream plant operating time per year, 333 days (91.2 pet). 
 
 2 Production revenue = actual production x 2,000 Ib/st x selling price. Selling price is 51 (50.94) c/lb, and DCFROR is 20 pet. 
 
 3 Sum of these values = net present value (NPV) = 0. 
 
 NOTE.— Underground capital cost (UCC) is $/yr for life of project. 
 
 Source: Davidson, D. H., R. V. Huff, R. E. Weeks, and J. F. Edwards. Generic In Situ Copper Mine Design Manual (contract J0267001, Science Ap- 
 plications Int. Corp.). Volume IV: Santa Cruz Field Experiment and Design of Commercial Scale Operation, Apr. 1988, 385 pp. 
 
 SENSITIVITY ANALYSIS 
 
 This cost model may be used to conduct sensitivity 
 analyses. Sensitivity analyses can be used to evaluate the ef- 
 fects of uncertainty on investment by determining how the in- 
 vestment profitability varies as parameters are varied one at a 
 time. 5 Further, sensitivity analyses may be used to identify 
 those critical variables that, if changed, could affect substan- 
 tially the profitability measure, such as DCFROR. When car- 
 rying out a sensitivity analysis, the value for an individual 
 variable is changed, and the effect of the change on the ex- 
 pected rate of return (or other profitability measure) is 
 calculated. Once the strategic variables that have major effects 
 on the profitability measure are identified, they can be given 
 special attention by the mine management. Several examples 
 of parameters that may be allowed to vary in an in situ mining 
 sensitivity analysis are copper loading, ore grade, well comple- 
 tion and type of stimulation, and well spacing. These may be 
 plotted against selling price while holding DCFROR constant 
 or against DCFROR while holding the selling price constant. 
 
 According to the Stermole book, the range approach to 
 sensitivity analysis involves estimating the most optimistic and 
 most pessimistic values (or the best and worst cases) for each 
 parameter in addition to estimating most expected values. The 
 best, worst, and most expected case sensitivity analysis results 
 provide important information that bracket the range of proj- 
 ect rate of return results that can reasonably be expected. 
 
 'Work cited in footnote 3. 
 
 Figures 2 and 3 represent examples of sensitivity plots for 
 the Santa Cruz test site with data taken from Volume IV of 
 SAIC's final report. In using sensitivity analysis, different 
 values are assigned to one input parameter, plotted against the 
 x-axis, while other input parameters are held constant. As dif- 
 ferent values are given to the one input parameter, the corre- 
 sponding changes in selling price are plotted against the y-axis. 
 The selling price values are determined through the cash-flow 
 analysis. For these sensitivity analysis examples, the DCFROR 
 is 20 pet. The vertical line within each plot indicates the initial 
 case value estimate for the parameter to be changed. 
 
 In figures 2 and 3, the initial case parametric value pro- 
 vides a selling price of 51 c/lb for the DCFROR of 20 pet. In 
 general, in all of the examples an increase in the parametric 
 value decreases the selling price. No estimates of ore grade 
 above the initial case value of 0.72 pet were made (fig. 2C) 
 because this value was thought to be at the high end for the 
 Santa Cruz example. 
 
 Figure 3, the plot of selling price versus permeability, is an 
 illustration of combined effects in the cash-flow analysis. The 
 permeability increases reduce the number of wells and thus the 
 total cost of one well field. By maintaining a fixed well spac- 
 ing, the well field will be depleted of copper in less time 
 because of the higher injection flow rate. (More well fields will 
 be needed over the project life, which will cause additional 
 capital costs.) The higher flow rate (higher permeability) can 
 result in higher negative discounted cash flow if the initial well 
 field life is less than 3 yr, thus requiring a higher selling price to 
 maintain the 20-pct DCFROR. 
 
12 
 
 
 
 
 
 
 DO 
 
 60 
 
 i 
 
 i i 
 
 A 
 
 DO 
 
 60 
 
 
 1 - 1" 
 
 i 
 
 B 
 
 
 
 
 55 
 
 
 
 55 
 
 
 
 
 
 
 .a 
 o 
 
 50 
 
 
 
 50 
 
 «ft 
 
 
 
 
 
 
 45 
 
 i 
 
 i i 
 
 
 1 I 
 
 i 
 
 
 
 20 30 40 5 
 
 )0 
 
 125 150 175 
 
 200 
 
 LxJ 
 O 
 
 a: 
 
 CL 
 
 COPPER PRODUCTION RATE, 1,000 st/yr 
 
 
 PRODUCTION WELL SPACING, ft 
 
 
 
 
 
 
 
 
 
 
 
 o 
 
 LU 
 CO 
 
 0.5 0.6 0.7 
 
 ORE GRADE. * 
 
 0.8 
 
 5 10 15 
 
 EFFLUENT COPPER LOADING, g/L 
 
 20 
 
 Figure 2.— Sensitivity analyses, Santa Cruz commercial-scale operation example. (Vertical line represents initial case.) 
 
 70 
 JQ 
 
 s 
 
 « 65 
 
 •> 
 
 Ld 
 O 
 
 a. 
 
 o 
 
 z 
 
 Zj 55h 
 
 60 
 
 Ld 
 
 to 
 
 50 
 
 \02) 
 
 p r- i 
 
 KEY 
 
 
 (No.) Calculated 
 well field life, yr 
 
 W) 
 
 i 
 
 (2) (2) (2)- 
 
 
 1 2 3 
 
 PERMEABILITY, mD 
 
 Figure 3.— Combined effects in cash-flow analysis, Santa Cruz commercial-scale operation 
 example. (Vertical line represents initial case.) 
 
13 
 
 SUMMARY 
 
 A menu-driven computerized cost model was designed 
 that may be used in developing a commercial in situ copper 
 mining operation. Specifically, this cost model may be used to 
 determine the economics of mining copper resources by the in 
 situ mining method. This model permits the user to make cost 
 inputs and inputs that are converted to mine design elements. 
 Outputs that are generated include specific design parameters 
 and capital and operating costs. Other outputs include a pretax 
 
 selling price or DCFROR and a discounted-cash-flow table. 
 The computer model may be used to make sensitivity analyses, 
 whereby one input parameter's value is changed and the other 
 input parameters' values are held constant. The effect of this 
 change on selling price or DCFROR is then determined. The 
 computer program is on an IBM-compatible microcomputer; 
 the minimum hardware requirements for running the program 
 are a 360-Kbyte memory and a math coprocessor. 
 
14 
 
 INTRODUCTION TO THE ENVIRONMENTAL PERMITTING 
 PROCESS FOR IN SITU COPPER MINING 
 
 By Daniel J. Millenacker 1 
 
 ABSTRACT 
 
 This U.S. Bureau of Mines paper introduces the prospective mine operator to the en- 
 vironmental permitting process for an in situ copper mining activity in the State of Arizona. 
 It identifies the environmental aspects of in situ copper mining through a discussion of the 
 regulatory authorization process. The Bureau intends to use this collected information in 
 subsequent phases of its research to obtain permits to conduct a field test at the Santa Cruz, 
 AZ, field site. 
 
 INTRODUCTION 
 
 In situ copper mining is a relatively new concept, which 
 involves pressure injection of a dilute sulfuric acid leach solu- 
 tion (lixiviant) into an ore zone through a geometrically ar- 
 ranged well pattern. Dissolution of the mineralization located 
 along fractures or disseminated within the ore occurs as the 
 minerals are contacted by the injected lixiviant. Resulting 
 copper-bearing solution is drawn to recovery wells where it is 
 pumped to a surface processing plant. The solvent extraction- 
 electrowinning (SX-EW) method is used to remove copper 
 from solution. 
 
 As part of its program to ensure a dependable domestic 
 supply of minerals, the Bureau of Mines is evaluating the use 
 of in situ methods for extracting copper from previously un- 
 mined oxide ore deposits. Successful development of this min- 
 ing technique and its acceptance by the mining industry will 
 require an engineering design both technically feasible and en- 
 vironmentally compatible. One task of the Bureau's research is 
 to identify the environmental issues of in situ copper mining 
 and the corresponding regulatory elements of environmental 
 permitting and protection. 
 
 Environmental issues specific to in situ copper mining are 
 addressed through a permitting process administered by 
 Federal, State, and local governmental regulatory agencies. 
 Each agency maintains jurisdiction over a specific en- 
 vironmental element of the mining operation. Permitting in- 
 volves preparation of comprehensive application documents 
 by the mining company and subsequent document submittal to 
 the appropriate agency. Authorization to proceed with the 
 mining activity may be either granted or denied by the 
 regulatory agency depending upon anticipated facility com- 
 pliance with environmental protection requirements. Once 
 permits are approved and the facility is constructed, facility 
 operation requires strict adherence to environmental protec- 
 tion performance standards. It is incumbent upon the mine 
 operator to consider all environmental requirements early in 
 
 the mine planning process since permitting forms the first ma- 
 jor step in establishing the feasibility, costs, and schedule for 
 developing an environmentally sound mine design. 
 
 Environmental resources afforded special protection 
 through permitting include ground water, air, biological, and 
 cultural resources. By the very nature of the in situ mining 
 process, the greatest potential for environmental effect rests 
 with the fluid injection-recovery system. Accordingly, the 
 principal environmental permitting issue is protection of 
 ground water resources. A permit application specific to 
 ground water protection must include, at a minimum, infor- 
 mation on site characteristics and hydrogeologic conditions, 
 the extent of the area to be affected by the injection activity, 
 anticipated effects to environmental resources, well and sur- 
 face facility construction specifications and plans for opera- 
 tion, site monitoring methods and procedures, contingency 
 plans, and plans for facility closure. Federal and State re- 
 quirements restrict injection activities that have the potential 
 to degrade ground water resources. Requirements that apply 
 to ground water protection in Arizona have previously been 
 discussed by Weeks (/). 2 
 
 Protection of air, cultural, and biological resources are 
 subordinate to the issue of ground water protection. Air emis- 
 sions expected to result from an in situ operation include 
 fugitive dust from access road traffic, and volatile organics 
 and sulfuric acid mist from operation of the SX-EW facility. 
 Air resources protection requirements are intended to mini- 
 mize the introduction of contaminants into the open air from a 
 source or operation. Biological and cultural resources may be 
 afforded special protection if ownership of "the property upon 
 which the facility is sited is public. Given the potential for ef- 
 fects on ground water and air quality from this type of mining 
 activity, the focus of this paper is on permitting requirements 
 as they apply to protection of these two environmental re- 
 sources. 
 
 'Hydrologist, Twin Cities Research Center, U.S. Bureau of Mines, Min- 
 neapolis, MN. 
 
 ^Italic numbers in parentheses refer to items in the list of references at the end 
 of this paper. 
 
15 
 
 REGULATORY REQUIREMENTS IN ARIZONA 
 
 Environmental requirements addressing protection of 
 ground water and air resources in Arizona are currently ad- 
 ministered by several governmental regulatory agencies. With 
 respect to ground water protection, the U.S. Environmental 
 Protection Agency (EPA), the Arizona Department of En- 
 vironmental Quality (DEQ), and the Arizona Department of 
 Water Resources (DWR) maintain jurisdiction. For injection 
 activities occurring on Federal or Indian lands, requirements 
 of the specific Federal land management agency (Forest Serv- 
 ice, Bureau of Land Management, or Bureau of Indian Af- 
 fairs) may also apply. Requirements of these agencies may be 
 addressed in a facility plan of operation or other suitable docu- 
 ment. Requirements addressing protection of air resources are 
 administered through either DEQ or county air quality control 
 boards. 
 
 GROUND WATER PROTECTION 
 Federal Water Quality Requirements 
 
 EPA regulations relevant to fluid injection activities are 
 authorized under the Underground Injection Control (UIC) 
 program promulgated under part C of the Safe Drinking 
 Water Act, Public Law 93-523. The technical requirements of 
 this program are identified under 40 CFR 144 and 146 (2). Re- 
 quirements specified under part 144 address permitting and 
 program specifications. Part 146 addresses technical criteria 
 and standards for operation. 
 
 According to EPA, no fluid injection shall be authorized 
 if it results in the movement of fluid containing a contaminant 
 into an underground source of drinking water (USDW), if the 
 presence of that contaminant may cause a violation of primary 
 drinking water standards or may adversely affect public 
 health. If an aquifer within a prospective leach interval meets 
 the definition of a USDW, it must be determined to be an "ex- 
 empted aquifer" before lixiviant injection may commence. An 
 exempted aquifer is one that would otherwise qualify as a 
 USDW but has been determined to have no real potential as a 
 drinking water source. 
 
 Five well classes have been established under the UIC pro- 
 gram to regulate fluid injection activities. Two of the five well 
 classes are applicable to in situ mining. A Class III well desig- 
 nation applies to injection wells used for commercial in situ 
 production from ore bodies that have not been conventionally 
 mined. A Class V well is one used for solution mining of con- 
 ventional mines, such as stope leaching, and for wells used in 
 development of new technologies. The latter well class is 
 generally assigned to those wells that inject nonhazardous 
 fluids and are considered less of an environmental risk. An ex- 
 ample of a Class V well is one used during a field-scale test of a 
 mineral deposit to assess viability for commercial-scale in situ 
 production. 
 
 For in situ mining, the UIC program authorizes well con- 
 struction by either permit or rule. Authorization by permit ap- 
 plies to a Class III commercial well facility, while a Class V 
 facility is authorized by rule. Application for a Class III well 
 permit requires explicit and detailed information on the hydro- 
 geologic resources present within the facility area, the well 
 field design and operational plan, reporting and monitoring 
 schedules, and plans for site closure. The principal emphasis 
 of the Class III permit is design and operation of the facility to 
 avoid migration of process fluids into, and degradation of, an 
 
 overlying or adjacent USDW. The Class V authorization-by- 
 rule process, on the other hand, requires an operator to pro- 
 vide EPA with generalized inventory information on the facili- 
 ty, along with any additional data that EPA may require. The 
 EPA maintains the prerogative to require a UIC permit for 
 any Class V well authorized by rule at any point in time if 
 operation of that well does not meet minimum operating 
 standards. 
 
 In the application for a Class III permit, the physical 
 limits of the study area must be established by defining the 
 "area of review." This establishes the potential area of 
 hydraulic influence of the well field. For a well array, the area 
 of review is defined under 40 CFR 146 as the project area plus 
 a circumscribing area the width of which is the lateral distance 
 from the perimeter of the facility, in which the pressures in the 
 injection zone may cause the migration of the injection and/or 
 formation fluid into a USDW. EPA provides the option of ap- 
 plying either this computational method for determining the 
 area of influence or selecting a fixed radius of not less than 
 one-fourth of a mile around the site. 
 
 Closure of a Class III facility requires proper well field 
 abandonment. The need to restore a leached interval is de- 
 pendent upon the hydrologic conditions expected to return to 
 the site upon completion of leaching activities. If hydraulic 
 conditions are favorable for migration of residual process 
 fluids from the site and the integrity of a receiving USDW can- 
 not be assured, restoration will be necessary. Restoration is the 
 rehabilitation of water quality within the leached interval to a 
 baseline or predetermined water quality standard. A permittee 
 is also required to maintain the financial resources to close, 
 plug, and abandon the well field as required. Financial assur- 
 ances (performance bonds, financial statements, etc.) must be 
 provided to EPA as a component of the permit application. 
 
 Arizona Water Quality Requirements 
 
 The State of Arizona presently administers its own water 
 quality protection program under authority of the Arizona En- 
 vironmental Quality Act (EQA). The EQA was enacted by the 
 State legislature in August 1986 to protect water resources 
 from all activities that have the potential to degrade water 
 quality. The DEQ was established as the principal agency re- 
 sponsible for enforcing provisions of the EQA. Statutory 
 language addressing protection of water resources from injec- 
 tion activities is found under Arizona Revised Statutes (ARS) 
 49-201 through 208, ARS 49-221 through 225, and ARS 49-241 
 through 244 (5). The DEQ is required to adopt, by rule, an 
 aquifer protection permit program to control discharges of 
 any pollutant or combination of pollutants that are reaching 
 or may with reasonable probability reach an aquifer. Efforts 
 to establish this permit program are presently under way. With 
 certain exceptions, an aquifer protection permit will be re- 
 quired of any person who owns or operates a facility that dis- 
 charges fluids. Given the recent date of statutory authoriza- 
 tion, the rulemaking process will require 1 to 2 yr before final 
 permit requirements are implemented. The DEQ is also re- 
 quired to adopt, by rule, the UIC permit program described in 
 the Safe Drinking Water Act. Any effort to obtain primacy 
 from EPA for regulating underground injection activities will 
 require the DEQ to adopt rules that closely parallel the EPA- 
 UIC regulations. In developing its own UIC program, a State 
 is not precluded from developing regulations more stringent 
 than those established by EPA. Language specific to State 
 UIC program requirements is identified under 40 CFR 145 (2). 
 
16 
 
 
 Prior to enactment of the EQA, the Arizona Department 
 of Health Services (ADHS) was responsible tor enforcing State 
 \\ ater quality protection standards. (The ADHS was the prede- 
 cessor to DEQ.) The ADHS regulated injection activities 
 through a Notice of Disposal or ground water protection per- 
 mit. Information requirements for a permit were more com- 
 prehensive than those for a Notice of Disposal. Under ADHS, 
 a permit application required information on site hydro- 
 geology and anticipated hydrologic impacts, facility design 
 and operation, monitoring, reporting, contingency planning, 
 and decommissioning. The aquifer protection permit estab- 
 lished under the new DEQ regulatory program will require 
 consideration of many of the same elements originally in- 
 cluded under the ground water protection permit. A new 
 criterion, however, specifies that an operator apply the best 
 available demonstrated control technology (BADCT) to en- 
 sure aquifer protection. BADCT promotes the use of tech- 
 nologies that result in discharge reduction or, if practical, 
 complete elimination of the discharge. 
 
 A key element of the new EQA specifies that all water- 
 bearing units in Arizona meeting the definition of an aquifer 
 are to be classified for drinking water protected use. The DEQ 
 has defined an aquifer as a geologic unit that contains suffi- 
 cient saturated permeable material to yield 5 gal/d of water to 
 a well or spring (4). Water quality standards in effect for an 
 aquifer are the primary maximum contaminant levels estab- 
 lished by EPA. The DEQ maintains flexibility to establish 
 aquifer water quality standards for additional pollutants for 
 which no Federal or State standard exists. The classification of 
 an aquifer, or portion thereof, can be changed by DEQ to a 
 use other than drinking water if specific criteria are met. If a 
 leaching operation is to occur in a geologic unit that meets the 
 definition of an aquifer and water quality standards cannot be 
 met at the designated facility point of compliance, the 
 operator must petition DEQ for reclassification of the aquifer 
 (or portion thereof) to a nondrinking water protected use. 
 
 Arizona Water Conservation Requirements 
 
 A second Arizona agency involved in the permitting pro- 
 cess for an in situ copper mining operation is the DWR. The 
 
 DWR is principally responsible for ground water conservation 
 but also enforces stringent criteria for drilling, rehabilitation, 
 completion, replacement, and abandonment of wells. Active 
 management areas (AMA) have been established under DWR 
 authority for regulating ground water use. Removal of ground 
 water from a location within an AMA may occur only if a right 
 to that water has been established. The DWR specifies pro- 
 cedures to be followed in establishing a ground water 
 withdrawal right. The DWR criteria specific to water use and 
 well construction are defined under ARS 45-591 through 604 
 (5) and R12-15-801 through 821 (6). 
 
 AIR QUALITY PROTECTION 
 
 Construction of an SX-EW plant will require an operator 
 to obtain an air quality installation permit from either DEQ or 
 an appropriate county air quality control board (7). (Arizona 
 counties with approved State air quality control programs are 
 given jurisdiction over facilities that produce a total emission 
 quantity up to 75 st/d. Facilities producing emissions in excess 
 of this rate are regulated by DEQ.) An installation permit is re- 
 quired for each "permit unit" of basic equipment and its atten- 
 dant air pollution control equipment. A permit unit contains 
 all items of equipment that operate as a single functional unit. 
 Additional permit units may be considered for storage of 
 solids and liquids, incinerators, and fuel burning equipment. 
 The installation permit application must provide a detailed 
 description of all processes, process equipment, storage units, 
 and fuels to be burned, and other information necessary to 
 describe the proposed project and its air pollutant emission 
 sources, together with estimated emission quantities. The per- 
 mit applicant must describe the system or technique to be used 
 to control emission of all pollutants and additionally provide a 
 description of ambient air quality at the proposed site. An in- 
 stallation permit remains in effect throughout facility con- 
 struction and facility startup. Once the facility becomes com- 
 mercially operational, an operating permit is granted by the 
 regulatory agency if specified operational criteria are met. The 
 operating permit establishes the emission standards to be met 
 during facility operation. 
 
 IN SITU COPPER MINING FIELD TEST 
 
 The Bureau, in cooperation with the Santa Cruz Joint 
 Venture (a mining partnership formed between subsidiaries of 
 ASARCO Incorporated and Freeport-McMoRan Gold Co.), 
 intends to conduct an in situ copper mining field test at the 
 Santa Cruz site in the south-central part of Arizona, approx- 
 imately 7 miles west of the town of Casa Grande. The project 
 involves construction and operation of two interconnected 
 five-spot well patterns and a surface plant capable of process- 
 ing copper-bearing leach solution. The facility design was 
 prepared using specifications calculated from the Generic In 
 Situ Copper Mine Design Manual (8). The field test will utilize 
 the unit-cell concept, which is essentially a small well field con- 
 structed to commercial-scale well size and spacing specifica- 
 tions. Corner wells in the pattern will serve production needs, 
 and center wells will be used for lixiviant injection. The field 
 test is scheduled to run for a 48-month period. Fluid injection- 
 recovery operations will account for 18 months of this total, 
 while construction, system testing, and decommissioning ac- 
 tivities will account for the balance. Although the field test is a 
 scaled-down version of a commercial-size operation, it is sub- 
 
 ject to the same environmental analysis as would be required 
 of a full-scale in situ copper mine. 
 
 The Bureau, together with the Santa Cruz Joint Venture, 
 has outlined three environmental tasks to be conducted during 
 the 4-yr project life. The first of these is to establish a plan for 
 permitting the field test. This includes developing permitting 
 schedules and identifying the regulatory requirements and 
 design criteria applicable to well field and surface plant con- 
 struction and operation. The second task is to conduct a site 
 hydrogeologic investigation to define in situ fluid flow condi- 
 tions and to determine the area of hydraulic influence around 
 the facility. Data collected through this investigation will be 
 used both for permitting and for developing final engineering 
 designs. The third and final task will be to construct the well 
 field and surface plant and operate it in full compliance with 
 permit terms and conditions. This task additionally involves 
 closure of the facility in a manner consistent with permit re- 
 quirements. 
 
 At the present time, the first of the environmental tasks 
 has been completed. Agency jurisdictions have been identified, 
 
17 
 
 and permitting schedules have been estimated. Environmental 
 permits that must be obtained include a Class V well 
 authorization from EPA region 9, an aquifer protection per- 
 mit from DEQ, a permit to construct wells from DWR, and an 
 air quality installation permit from Pinal County, AZ. The 
 Santa Cruz Joint Venture, as landowner of the property upon 
 which the field test is to be sited, maintains a grandfathered 
 ground water right and is free to withdraw water from the site 
 without obtaining a separate ground water withdrawal permit 
 from DWR. Design and performance criteria of each of these 
 agencies have been incorporated into the preliminary facility 
 design. 
 
 A site hydrogeologic investigation, the second task, has 
 been initiated but not yet completed. Collected data will be 
 analyzed and used for development of the DEQ aquifer pro- 
 tection permit, as well as for final facility design. The DEQ 
 permit is by far the most difficult environmental authorization 
 to be obtained for the field test. The estimated time required 
 for permitting is 12 to 16 months. This period is all inclusive 
 from initiation of the site investigation through issuance of the 
 permit. Other agency permits will be sought in a manner con- 
 sistent with the DEQ permit application schedule. The third 
 task will be initiated during the second year of the field test 
 and will carry through to field test completion. 
 
 SUMMARY 
 
 The Bureau is evaluating the feasibility of extracting cop- 
 per from previously unmined oxide ore deposits using in situ 
 methods. This research program gives equal weight to develop- 
 ing a mining method both technically feasible and en- 
 vironmentally compatible. 
 
 The Bureau has identified the environmental re- 
 quirements of in situ copper mining and the regulatory agen- 
 cies that maintain jurisdiction over fluid injection activities in 
 Arizona. Collected information is being used to develop a pro- 
 gram for obtaining all necessary permits to conduct an in situ 
 copper mining field test. The intent of the field test is not only 
 to demonstrate the engineering feasibility of in situ mining but 
 
 also to establish the environmental acceptability of this 
 method through permitting, and to operate the facility in full 
 compliance with applicable environmental protection per- 
 formance standards. 
 
 Although the environmental approach to this research is 
 specific to in situ copper mining in Arizona, it can also apply 
 to in situ extraction of other metals or to operations that might 
 occur in other States. The experience gained from this en- 
 vironmental program will provide the mining community with 
 the information necessary to prepare a comparable and effec- 
 tive program for permitting a future in situ mining operation. 
 
 REFERENCES 
 
 1. Weeks, R. E., and D. J. Millenacker. Environmental Permit- 
 ting Considerations for True In Situ Copper Mining in the State of 
 Arizona. Soc. Min. Eng. AIME preprint, 88-196, 1988, 7 pp. 
 
 2. U.S. Code of Federal Regulations. Title 40 — Protection of En- 
 vironment; Chapter I — Environmental Protection Agency; Sub- 
 chapter D — Water Programs; Parts 144, 146 — Underground Injection 
 Control Program; July 1, 1986. 
 
 3. Arizona Legislature. Arizona Environmental Quality Act. 
 Laws of 1986. Chapter 368, Arizona Revised Statutes, Title 49, 
 Chapter 2, Articles 1-7; 1986. 
 
 4. Arizona Department of Environmental Quality. Aquifer Boun- 
 dary and Protected Use Classification. Arizona Administrative Rules 
 and Regulations, Title 18, Chapter 11, Article 5; 1987. 
 
 5. Arizona Legislature. Arizona Groundwater Code. Laws of 
 1980. Fourth Special Session. Chapter 1, Arizona Revised Statutes, 
 Title 45, Chapter 2, Articles 1-12; 1980. 
 
 6. Arizona Department of Water Resources. Well Construction 
 and Licensing of Well Drillers. Arizona Administrative Rules and 
 Regulations, Title 12, Chapter 15, Article 8; 1984. 
 
 7. Arizona Legislature. Arizona Environmental Quality Act. 
 Laws of 1986. Chapter 368, Arizona Revised Statutes, Title 49, 
 Chapter 3, Articles 2-3; 1986. 
 
 8. Davidson, D. H., R. V. Huff, R. E. Weeks, and J. F. Edwards. 
 Generic In Situ Copper Mine Design Manual (contract J0267001, 
 Science Applications Int. Corp.). Volume I: Executive Summary, 
 Apr. 1988, 93 pp.; Volume II: Draft Generic In Situ Copper Mine 
 Design Manual, Apr. 1988, 454 pp.; Volume III: Lakeshore Field Ex- 
 periment and Design of Commercial Scale Operation, Apr. 1988, 371 
 pp.; Volume IV: Santa Cruz Field Experiment and Design of Com- 
 mercial Scale Operation, Apr. 1988, 385 pp.; Volume V: Field Testing 
 at the Santa Cruz Site, Apr. 1988, 114 pp.; for inf., contact J. K. 
 Ahlness, TPO, Twin Cities Res. Cent., BuMines, Minneapolis, MN. 
 
18 
 
 LABORATORY CORE-LEACHING AND PETROLOGIC STUDIES 
 TO EVALUATE OXIDE COPPER ORES FOR IN SITU MINING 
 
 By Steven E. Paulson 1 and Harland L. Kuhlman 2 
 
 ABSTRACT 
 
 This paper describes U.S. Bureau of Mines laboratory core-leaching and geologic characteriza- 
 tion studies of oxide copper ores. These studies can provide detailed information about the nature 
 of the various chemical and physical processes operating during leaching of oxide copper ores and 
 determine the impact these processes will have on in situ mining. Three types of core-leaching 
 systems are being utilized to conduct these experiments. The leaching experiments provide a very 
 useful means of evaluating the effects of several variables on leaching, such as ore and gangue 
 mineralogy and chemistry, textural relationships between ore and gangue minerals and fluid flow 
 paths, leach solution concentration, flow rate, and pressure and temperature conditions during 
 leaching. Results from two sets of experiments conducted on two different ore types from Pinal 
 County, AZ, are discussed. These ores exhibited very different leaching characteristics because 
 of their distinctly different mineralogy and texture. 
 
 INTRODUCTION 
 
 The Bureau of Mines has developed a laboratory research 
 program with the main objectives of (1) gaining an understanding 
 of both the physical and chemical factors that have an impact on 
 the in situ mining of oxide copper ores and (2) developing techniques 
 that can be used to evaluate ore deposits for their amenability to 
 in situ mining. Three main research endeavors are utilized in this 
 program: conducting laboratory core-leaching experiments, deter- 
 mining the resulting fluid chemistry from these experiments, and 
 conducting preleach and postleach petrologic studies. The infor- 
 mation obtained from such studies on ore from a given deposit is 
 combined with field data from that deposit to provide a better 
 understanding of what can be expected if in situ mining is applied 
 to the deposit. 
 
 Up to this point, Bureau core-leaching studies have been con- 
 ducted only on oxide copper ores from Arizona. The purposes of 
 this paper are to explain in detail how the laboratory core-leaching 
 experiments are conducted so that others may perform similar ex- 
 periments on other oxide copper ores, to give examples of the types 
 of data that are generated by such studies, and to illustrate the 
 usefulness of these data for evaluating ore deposits for their suit- 
 ability for in situ mining. 
 
 There are two main reasons why the Bureau is conducting 
 leaching experiments on core samples rather than crushed material. 
 First, it is likely that the core-leaching experiments more closely 
 simulate the nature of the solution-rock interface in situ. In many 
 oxide copper deposits, the ore mineralization is distributed in the 
 fractures and pore spaces of the rock, which are where the leach 
 solutions will travel, both in situ and in core-leaching experiments. 
 Thus, the fluid chemistry resulting from such experiments will more 
 
 likely resemble that generated during in situ mining, giving a more 
 realistic indication of what the fluid composition will be during in 
 situ mining. In many cases, experiments performed with aggregate 
 material would expose a much greater proportion of gangue 
 mineralization to the leach solutions, perhaps resulting in a leachant 
 chemistry significantly different from that generated during in situ 
 mining. Therefore, it is likely that core-leaching experiments provide 
 a better assessment of parameters such as copper concentration in 
 solution, copper recovery from the rock, and acid consumption by 
 gangue mineralization (7). 3 
 
 Second, core-leaching experiments allow for an evaluation of 
 permeability changes taking place during leaching. Influences on 
 permeability include mineral dissolution and precipitation, clay 
 mineral hydration, and physical sedimentation. Mineral dissolution 
 may act to increase the size of the flow channels and thus increase 
 permeability. However, the elevated sulfate (S0 4 -2 ) content in the 
 leach solution may result in the combining of S0 4 -2 ions with 
 cations mobilized by mineral dissolution or ion exchange with clay 
 minerals, causing precipitation of sulfate minerals. This precipita- 
 tion may decrease permeability. Also, permeability may be 
 adversely affected by any carbon dioxide that may be generated 
 by acid attack on any carbonate minerals in the rock. Although the 
 permeability in a deposit will likely be much different from that 
 measured during laboratory core-leaching experiments, correlation 
 of permeability changes during the experiments with data on the 
 fluid chemistry and sample mineralogy will yield information that 
 may be useful for determining the potential impact of these 
 mechanisms during actual in situ mining in the field. 
 
 The three basic requirements for in situ mining are controlled 
 by the inherent petrology and geochemistry of the ore body. First, 
 
 ■Geologist. 
 
 Engineering technician. 
 Twin Cities Research Center, U.S. Bureau of Mines, Minneapolis, MN. 
 
 'Italic numbers in parentheses refer to items in the list of references at the end of 
 this paper. 
 
19 
 
 the leach solutions must be able to contact the ore minerals. This 
 is controlled by the rock texture, the nature of the ore mineral 
 distribution, and structural features of the ore body that influence 
 the flow of leach solution through the deposit. Second, the target 
 metals must be selectively solubilized by the leaching reagent. This 
 is accomplished by the type of leach solution utilized and its con- 
 centration. Finally, the metallic species must be mobilized to the 
 recovery wells. This is influenced by the solution pH and the con- 
 centrations of the various ions in solution, which are controlled by 
 the reactivity of the minerals and the flow rate of the leach solution 
 through the ore body. Precipitation may occur if ion concentration 
 and/or solution pH reaches sufficiendy elevated levels. Metal ion 
 mobilization can also be influenced by the interaction of clay 
 minerals with the metal ions in solution, as metal ion sorption onto 
 clay minerals can be significant in reducing metal ion concentra- 
 
 tion in solution. Laboratory core-leaching and petrologic studies 
 can help to determine what parameters have the greatest influence 
 on these mechanisms, as well as the magnitudes of these influences. 
 Among the variables being studied are the compositions of the ore 
 and gangue minerals present in the system, the textural and struc- 
 tual relationships among the minerals, the leach solution concen- 
 tration, the flow rate of the leach solution through the rock (residence 
 time), and the pressure and temperature conditions during the 
 experiment. By incorporating the laboratory data with field data 
 from a deposit, one can make predictions of what will actually occur 
 during full-scale in situ mining of that deposit. Experiments will 
 yield information such as copper loading in solution, fractional 
 copper removal versus time, and lixiviant consumption per unit of 
 copper produced. Such information is essential when assessing the 
 feasibility of applying in situ mining to a deposit. 
 
 EXPERIMENTAL APPARATUS 
 
 The Bureau has a complete laboratory program to study the 
 petrologic and chemical influences on in situ mining. This program 
 includes the development of sample preparation techniques for both 
 the petrologic studies and core-leaching experiments, as well as 
 several different versions of core-leaching apparatus. Among the 
 desirable features of a core-leaching system are flexibility to 
 accommodate a variety of sample sizes and shapes; nonreactivity 
 of the system with the leach solution; relatively low cost, thereby 
 permitting a large number of samples to be tested; and ability for 
 easy recovery of the postleach sample for petrologic studies (2). 
 
 Three systems have been designed to conduct experiments under 
 a variety of conditions. The main difference among the systems 
 is the type of reaction cell used to house the core samples. 
 
 Low-Pressure Core-Leaching System 
 With Acrylic Reaction Cell 
 
 The first type of reaction cell is constructed from acrylic pipe 
 and can be used at pressures up to 50 psi. It requires a small amount 
 of machining (figs. 1-2) and can be used again for subsequent 
 experiments (2). Portions of the core samples selected for the 
 experiments are submitted for chemical analyses, and thin sections 
 are prepared for examination of the preleach geochemistry and 
 petrology. The remainder of each core specimen is encased in epoxy 
 prior to placement into the reaction cell. This is done because in 
 many instances the samples become very friable because of reac- 
 tion with the acid, leading to difficulty in preparing postleach thin 
 sections. 
 
 The steps in sample preparation are as follows. The sample 
 is first painted with a thin coat of epoxy that cures at room 
 temperature, has low shrinkage, and is not reactive with the sulfuric 
 acid leach solutions used in these experiments (such as 3M 1838-L 
 Epoxy Adhesive). 4 This thin coating fills all of the surface 
 irregularities of the rock and ensures that the surface is completely 
 sealed. The sample is then placed on end in a segment of polyvinyl 
 chloride (PVC) pipe and the void space filled with epoxy. The inner 
 surface of the pipe is coated with a silicone lubricant prior to the 
 sample insertion to facilitate the removal of the sample after the 
 epoxy has cured. The ends of the cylinder are trimmed to expose 
 the surface of the rock and ground to provide a smooth surface 
 perpendicular to the core axis. (This procedure can also be used 
 for split core samples or other irregularly shaped pieces of ore 
 material if whole core samples are not available.) 
 
 After the sample is placed in the reaction cell, O-ring seal caps 
 are placed against each end of the sample. The acid-resistant O-ring 
 
 forms a seal between the wall of the reaction cell and the epoxy 
 jacket encasing the sample, which forces the leach solution to travel 
 through the core rather than along the outside edge (i.e. , short cir- 
 cuiting). The seal caps contain a hole in the center through which 
 the leach solution is pumped, and the seal cap surface facing the 
 rock is slightly concave to allow access of the leach solution to the 
 
 End cop 
 
 O-ring 
 seal cap 
 
 PVC spacer 
 
 Cell body 
 
 Epoxy jacket 
 
 Core sample 
 
 Scale, in 
 
 PVC spacer 
 
 O-ring 
 
 End cap 
 
 "Reference to specific products does not imply endorsement by the Bureau of Mines. 
 
 Figure 1.— Cross-sectional diagram of acrylic reaction cell for low- 
 pressure (50 psi) core-leaching experiments. 
 
20 
 
 entire rock faee. The necessary force to seal the O-ring at the sam- 
 ple surface is provided by the end caps of the reaction cell. Screws 
 are used to fasten the end caps to the main body of the reaction 
 cell. The end caps are also fitted with O-rings to prevent leakage 
 of leach solution from the reaction cell (fig. 1). If the core sample 
 is shorter than the reaction cell, the excess space is taken up with 
 a solid PVC cylinder of the same diameter, through which a hole 
 has been drilled to permit transport of the leach solution. The solid 
 PVC cylinder transfers the force from the end caps to seal the 
 O-rings at the sample surface. Tubing and fittings made of Teflon 
 
 Figure 2.— Two reaction cells for low-pressure core-leaching 
 experiments. Pressure gauges assess permeability changes during the 
 experiment. 
 
 fluorocarbon polymer are used for the plumbing on the system, and 
 the fluid samples are collected in closed polyethylene bottles (fig. 2). 
 
 The leach solution is injected at the bottom side of the sample, 
 and the injection pressure is monitored with a pressure gauge in 
 order to assess permeability changes that may occur during the 
 experiment (fig. 2). The gauges are fitted with plastic isolators to 
 prevent reaction of the sulfuric acid solutions used in the experiments 
 with the gauges. Reactions between the leach solutions and any com- 
 ponents of the leaching system would lead to contamination of the 
 fluid samples and, hence, severely complicate the study of the 
 leachant chemistry and reaction mechanisms operating during 
 leaching. A multichannel peristaltic pump is used to deliver the leach 
 solution to the sample with this system (fig. 3). The pump delivers 
 a constant flow of solution at low pressures (<30 psi), and the flow 
 rate of each channel can be varied by changing the diameter of the 
 flexible tubing that the metal rollers of the pump compress as they 
 rotate. 
 
 After completion of the experiment, the core sample is removed 
 from the reaction cell, and the reaction cell can be reused. A portion 
 of the sample is ground and analyzed for its postleach chemical 
 composition. In most cases the sample is quite friable because of 
 degradation during leaching. Therefore, the sample is dried and 
 vacuum impregnated with epoxy so that the cutting and grinding 
 needed for the preparation of the thin section can be easily 
 accomplished. Usually, several thin sections are prepared from each 
 sample to determine the extent of leaching within that sample. 
 
 Medium-Pressure Core-Leaching System 
 With PVC Reaction Cell 
 
 The second type of reaction cell is constructed from PVC pipe 
 and flanges (fig. 4). The main portion of this reaction cell can be 
 used only once, but it is relatively inexpensive and requires very 
 little machining. When this type of system is used, the sample is 
 
 Figure 3.— Low-pressure experimental core-leaching system. This photograph displays six experiments being conducted simultaneously, using 
 a multichannel peristaltic pump to pump the leach solution. 
 
21 
 
 5E 
 
 5 
 
 ZE 
 
 ir 
 
 3 
 
 E 
 
 rr 
 
 Blind flange 
 Gasket 
 
 Socket flange 
 
 Flange socket 
 
 PVC pipe 
 Epoxy potting 
 Test core 
 
 PVC spacer 
 
 3d 
 
 5 
 
 :e 
 
 Figure 4. — Cross-sectional diagram of PVC reaction cell for core- 
 leaching experiments. 
 
 cast directly into the PVC pipe with epoxy. After the epoxy has 
 cured, the ends of the whole assembly, which consists of core, 
 epoxy, and pipe, are again trimmed and ground to provide two 
 smooth surfaces. Each end is then fitted with a flange assembly, 
 which consists of a flange, a flange socket, a gasket, and a blind 
 flange (fig. 4). The flange itself is attached to the PVC pipe by 
 means of a PVC epoxy adhesive, and the blind flange is attached 
 with steel bolts. In order to reduce the volume of fluid stored in 
 the reaction cell, the void space between the core sample and the 
 blind flange is reduced by using a concave PVC spacer plate, 
 through which a hole has been drilled (fig. 4). Again, as with the 
 other type of reaction cell, a peristaltic pump and Teflon flurocarbon 
 polymer tubing are used to deliver the leach solution to the core 
 sample when working at pressures below 30 psi. 
 
 The PVC reaction cell can be used at pressures up to 150 psi, 
 but it is necessary to utilize a different pump and tubing when work- 
 ing in this pressure range. Since the flow rate of the peristaltic pump 
 is only constant at pressures up to 30 psi, a constant-flow, piston- 
 type pump is used for higher pressure experiments. The pump 
 selected for the current experiments is a constant-flow pump that 
 can deliver fluid at a very low flow rate, if necessary (0.001 
 mL/min) and can also be used at pressures up to 6,000 psi 
 (fig. 5). This allows for both maximum sensitivity and flexibility. 
 This pump was designed primarily for high-performance liquid 
 chromatography applications. 
 
 A pair of piston-type accumulators are used to isolate the pump 
 from the leach solution. Each accumulator has a capacity of 500 
 mL and is constructed of AISI Type 316 stainless steel (fig. 6). 
 Water is pumped to drive the piston and force leach solution into 
 the core sample. The two accumulators are connected in parallel 
 so that as one delivers leach solution to the sample, the other can 
 be refilled. A series of valves is used to isolate and switch the 
 accumulators. Type 316 stainless steel tubing, fittings, and valves 
 are used in place of Teflon fluorocarbon polymer at pressures greater 
 than 100 psi. All stainless steel wetted parts are passivated with 
 3M nitric acid for approximately 1 h prior to use in the system in 
 order to further reduce the possibility of reaction with the experi- 
 mental fluids. 
 
 Upon completion of the experiments using the PVC reaction 
 cell, one of two methods is used to remove the sample from the 
 cell: The specimen is overcored using a core bit slightly larger than 
 
 Figure 5.— Three experimental core-leaching systems utilizing PVC 
 reaction cells and high-pressure constant-flow pumps. 
 
 the sample diameter, or a lathe is used to trim away the PVC pipe 
 from the sample. After the sample is recovered from the reaction 
 cell, it is prepared in the same manner as described for the low- 
 pressure system, with a sample submitted to the analytical chemistry 
 laboratory before epoxy impregnation and thin section preparation. 
 
 Elevated-Pressure Core-Leaching System 
 With Stainless Steel Components 
 
 A system has also been developed to conduct experiments at 
 pressures above 150 psi. The maximum pressure at which this 
 system can be used is 1,000 psi. There are two reasons for con- 
 ducting experiments at these elevated pressures. First, the effects 
 of pressure on the leaching characteristics of various ores need to 
 be assessed, as it is likely that elevated pressures will be encountered 
 during in situ mining of oxide copper ores because of the signifi- 
 cant depths of some deposits. Therefore, elevated-pressure 
 experiments would be conducted to determine to what extent, if 
 any, pressure affects such parameters as primary mineral solubility, 
 secondary mineral precipitation, and reaction kinetics. Second, many 
 of the core samples tested in the laboratory exhibit very low 
 permeabilities, thereby making it very difficult or impossible to 
 establish an adequate flow of leach solution through these samples. 
 Thus, a higher pressure is needed to force the leach solutions through 
 the samples than can be attained with either of the two leaching 
 
22 
 
 -*HI 
 
 ■End caps 
 
 End plug 
 
 ■* Through bolts 
 
 Barrel (316SS) 
 
 Guide bushings 
 Sealing ring 
 Piston (316SS) 
 
 Sealing ring 
 
 Figure 6. — Cross-sectional diagram of stainless steel accumulator used 
 with high-pressure pump to deliver leach solution to the core sample. 
 
 systems previously described. This flexibility to conduct experiments 
 over a relatively wide pressure range allows for a larger variety 
 of samples to be tested and evaluated. 
 
 Sample preparation for experiments using the high-pressure 
 reaction cell is similar to that used with the PVC reaction cell. The 
 core sample is placed upright in a piece of ANSI schedule 40 steel 
 pipe, which is then filled with epoxy. After the epoxy has cured, 
 the ends are trimmed and ground to produce smooth surfaces perpen- 
 dicular to the axis of the sample. End caps for the reaction cell are 
 composed of Type 316 stainless steel and are held in place with 
 threaded steel rod passing through each cap (fig. 7). The end caps 
 are machined to fit over the end of the steel pipe, and they are fitted 
 with an acid-resistant O-ring, which forms the seal between the end 
 cap and the epoxy jacket surrounding the core sample. The end 
 caps are fitted with sintered titanium disks, which act as in-line filters 
 (3) while also providing support to the sample ends, which 
 frequently exhibit some spalling of material from the sample face. 
 Passivated Type 316 stainless steel tubing, valves, and fittings are 
 used throughout this leaching system. The 0. 16-cm tubing used has 
 an inside diameter of only 0.06 cm in order to reduce the amount 
 of fluid stored in the system. This small size also results in reduced 
 surface area of the tubing exposed to the leach solution and reduced 
 potential of back-diffusion of metals into the tubing and fluid 
 reservoir. Back-diffusion may occur in experiments with very 
 impermeable samples and, therefore, low flow rates. 
 
 As with the PVC reaction cell, a constant-flow, piston-type 
 pump and stainless steel isolators are used to deliver the leach solu- 
 
 
 Epoxy with 
 thickener 
 
 Epoxy jacket 
 
 Steel jacket 
 
 Through bolts 
 
 -rings 
 
 Porous discs 
 
 Figure 7.— Cross-sectional diagram of Type 316 stainless steel reaction 
 cell for elevated-pressure core-leaching experiments. 
 
 tion to the samples (fig. 8). The pump contains a high-pressure cutoff 
 switch, which is triggered if the pressure reaches the preset upper 
 pressure limit. At this point, the pump then becomes a constant- 
 pressure pump rather than a constant-flow pump, until the pressure 
 again drops below the preset upper pressure limit. The differential 
 pressure between the injection and recovery sides of the sample 
 is measured with a differential pressure transducer that is coupled 
 to an electronic data recorder to provide a continuous readout of 
 pressure throughout the experiment. The injection pressure of the 
 pump is also monitored with the data recorder. 
 
 If desired, a back pressure can be applied to the sample by 
 means of compressed nitrogen gas. A pressure gauge is used to 
 measure the actual back pressure applied to the sample. Fluid 
 samples are collected in a Type 316 stainless steel collection vessel, 
 which has been passivated with nitric acid (fig. 8). Experimental 
 fluid samples are removed from the collection vessel on a regular 
 basis and submitted to the analytical chemistry laboratory for 
 analysis. As with the PVC reaction cell, the postleach rock sample 
 can be recovered by overcoring or by turning down the pipe on 
 a lathe. 
 
23 
 
 Figure 8.— Elevated-pressure experimental core-leaching system (isolators not shown). 
 
 LEACHANT RECIRCULATION EXPERIMENTS 
 
 In addition to determining the fluid chemistry from experiments 
 in which fresh leach solution is continuously injected throughout 
 the duration of the experiment, it is also important to determine 
 the effects on leaching chemistry of recirculating the leachants back 
 through the rock after the copper has been removed, since this is 
 what will actually occur during in situ mining. Previous experiments 
 have shown that in addition to copper, appreciable amounts of 
 aluminum, calcium, iron, potassium, and sodium can also be 
 solubilized during leaching of oxide copper ores {1,4). Also, the 
 amount of S0 4 " 2 in the system increases with time as additional 
 
 sulfuric acid (H 2 S0 4 ) is added to the system to replace the acid con- 
 sumed by gangue mineralization and to maintain the desired pH. 
 Appreciable C 1 " can also accumulate in the leachant if a signifi- 
 cant amount of a copper chloride mineral, such as atacamite, is 
 present in the ore. Therefore, experiments were designed to study 
 the effects of fluid recycling on parameters such as mineral solubil- 
 ity, acid consumption, and net permeability. 
 
 Leachant recirculation experiments are designed to examine 
 the fluid chemistry after each pore volume of leach solution is 
 pumped through the core. As a convention, pore volume is defined 
 
24 
 
 here as the initial pore volume of the sample, since the actual pore 
 volume may change during the experiment because of mineral 
 dissolution and/or precipitation. Approximately 3 to 5 mL of sample 
 is needed for each set of chemical analyses performed on the 
 leachants. Therefore, it is desirable to recycle a total volume of 
 solution that will not result in significant depletion of fluid because 
 of sampling for the analytical determinations. Thus, in order to 
 ensure that an adequate amount of fluid is available for the dura- 
 tion of the experiment, a core sample with a pore volume >100 
 mL is suggested. Since the porosity of many of the oxide copper 
 ores tested in the laboratory ranged from 5 to 10 pet, a total rock 
 volume between 1 ,000 and 2,000 cm 3 is recommended. Therefore, 
 two types of experiments have been designed: one that utilizes a 
 large-diameter core sample (e.g., 10 or 15 cm) and one that uses 
 a longer segment of standard 5-cm-diameter core. 
 
 A large-diameter core sample is more desirable than a smaller 
 diameter sample because it has a greater surface area over which 
 to inject the leach solution, thus enhancing the possibility of inter- 
 secting mineralized fractures and fluid flow channels. Use of a large- 
 diameter sample also decreases the length of core necessary to meet 
 the minimum sample volume requirement, thereby allowing for a 
 faster flow rate than in a smaller diameter core sample with the 
 same permeability. This faster flow rate results in a reduction in 
 the time necessary to complete an experiment. Since 10- or 15-cm 
 core is not commonly available from exploratory drilling, larger 
 blocks of ore obtained from such sources as open pit or underground 
 mining operations need to be drilled to obtain these large-diameter 
 core samples. However, it is possible to conduct such recycle ex- 
 periments using smaller diameter core if larger samples are not 
 available. 
 
 Either the PVC or elevated-pressure reaction cell can be used 
 for the recycle experiments. The same basic design is utilized, 
 although the sizes of the components in the system need to be 
 matched with the diameter of the core samples to be used in the 
 experiments. After the diameter and approximate porosity of the 
 core samples have been determined, the length of core needed to 
 meet the minimum volume requirement can be calculated. In many 
 cases it will be necessary to use several pieces of core in one reac- 
 tion cell in order to meet the minimum sample size required to con- 
 duct the recycle experiment. The ends of core samples selected for 
 the experiment are trimmed on a rock saw and ground to provide 
 flat and reasonably smooth surfaces. The samples are then stacked 
 upright in the desired order, and a continuous composite rock core 
 
 is created by using epoxy adhesive, thickened with amorphous 
 fumed silica (3 wt pet), to spread around the joints between adja- 
 cent core segments. The fumed silica, which is unreactive with the 
 acidic leach solutions, gives the epoxy a pastelike consistency, which 
 allows it to be spread across the joints between sample segments 
 (fig. 7). Care must be taken to ensure that the thickened epoxy is 
 not forced into the interface between the sample segments, as this 
 will result in reduced solution flow through the composite sample. 
 After this thickened epoxy has cured, the composite core sample 
 is cast in epoxy and prepared as previously described. 
 
 After an adequate amount of leachant has been collected and 
 a 3- to 5-mL sample has been removed for the analytical chemistry 
 determinations, one of two methods is used to remove the copper 
 from the leachant prior to reinjection. The first method utilizes a 
 small electroplating cell to plate the copper directly out of the 
 leachant. The leachant is placed in a beaker and stirred as copper 
 is plated onto a copper cathode. A platinum basket is used as the 
 anode in this system. The second method of copper removal utilizes 
 an organic solvent extractant. Examples of solvent extractants that 
 worked well are Henkel LIX-622 and Acorga M-5397. A mixture 
 consisting of two parts of kerosene and one part of extractant is 
 added to an equal volume of leachant in a separatory funnel. The 
 mixture is agitated for several minutes and then allowed to settle 
 before the leachant is drawn off. This procedure is repeated until 
 the leachant no longer exhibits any blue color (usually 2 to 5 times, 
 depending on copper concentration). The leachant is then filtered 
 through a 0.45-/im membrane filter to remove additional organic 
 material that may be suspended. 
 
 The process of removing copper from the leachant results in 
 an exchange of H + for Cu +2 . However, the H + consumed by 
 gangue mineralization is not returned to solution during the cop- 
 per removal step. Therefore, it is necessary to restore the acid con- 
 centration to its initial value. This can be achieved by adding either 
 a relatively small amount of concentrated sulfuric acid or a larger 
 amount of dilute sulfuric acid. Which option to use will be dictated 
 somewhat by the total volume of solution being recycled. Adding 
 concentrated sulfuric acid can cause a rapid increase in the S0 4 
 concentration, which may be undesirable. However, the addition 
 of dilute sulfuric acid may lead to undesirable dilution of the 
 leachant. The advantages and disadvantages of both methods should 
 be considered. Perhaps the most important point is to maintain con- 
 sistency among experiments, as this will simplify data interpreta- 
 tion and comparison. 
 
 CHEMICAL AND PHYSICAL MEASUREMENTS 
 
 Prior to leach solution injection, distilled, deionized water is 
 pumped through the sample to completely saturate it and establish 
 the flow rate conditions for the experiment. Fluid samples from 
 the leaching experiments are collected, weighed, and analyzed on 
 a regular basis, generally every 48 to 72 h, to thoroughly docu- 
 ment the exact flow rate and chemical composition throughout the 
 duration of the experiments. This sampling frequency was chosen 
 to ensure that a large number of data points are obtained for each 
 experiment. The acidity of each sample is measured using two dif- 
 ferent methods. The first is a standard pH measurement and the 
 second is an end-point titration technique to measure the free-acid 
 content of the samples. Free acid is defined here as the H + present 
 in the leachant, excluding any contribution due to cation hydrolysis 
 after leaching (4). Because of the elevated levels of the various 
 cations in solution, the potential exists for appreciable H + 
 generation by hydrolysis reactions with these cations after the 
 leachant is collected from the experiment. Such reactions may lead 
 to erroneous assessments of acid consumption during leaching. 
 
 Therefore, a potassium oxalate buffer solution is added to a small 
 subsample of the leachant to complex the cations present. This solu- 
 tion is then titrated with a base to determine the actual amount of 
 H + remaining in the leachant collected, and therefore, how much 
 of the initial acid was consumed by reaction with the rock. An 
 automated titration system is utilized, which enables the rapid titra- 
 tion of many samples. 
 
 Approximately 2 to 3 g of leachant is diluted with distilled, 
 deionized water, acidified, if necessary, to prevent any precipita- 
 tion after dilution, and analyzed for the following components: 
 dissolved silica, and aluminum, calcium, copper, iron, magnesium, 
 potassium, and sodium. Atomic absorption (AA) spectroscopy and 
 inductively coupled argon plasma (ICAP) spectroscopy are used 
 to perform the analytical determinations for these components. An 
 additional 0.25 to 0.50 g of solution is diluted and anlayzed for 
 S0 4 ~ 2 and Cl~ using ion chromatography. Past experience has 
 shown that the leachant chemistry from the experiments using oxide 
 copper ores is composed chiefly of these 10 components. However, 
 
25 
 
 the specific analytes to be measured should be determined after a 
 close examination of the mineralogy of the rock samples to be used 
 for the experiments. Also, when conducting recycle experiments, 
 it may be necessary to expand the number of analytes, as elements 
 present only in trace amounts in the rock may accumulate to 
 appreciable concentrations after several recycle steps. 
 
 Prior to the initiation of each experiment, a portion of the core 
 sample is crushed, ground to pass a 100-mesh sieve, decomposed 
 by a fusion or acid dissolution technique, and then analyzed for 
 its elemental composition. ICAP and AA spectroscopy are used to 
 measure aluminum, calcium, copper, iron, magnesium, manganese, 
 potassium, silicon, sodium, and titanium. Combustion techniques 
 are used to measure carbon and sulfur, while ion chromatography 
 is used to measure chlorine in the rock. Trace elements such as 
 antimony, arsenic, barium, cadmium, chromium, lead, mercury, 
 selenium, silver, and zinc are measured using graphite furnace AA 
 spectroscopy. 
 
 Preleach and postleach petrologic and microprobe examina- 
 tions of the ore samples provide a great deal of useful information 
 about the texture and chemistry of the samples, which can be com- 
 bined with the results of the leaching experiments to help explain 
 the reactions occurring during leaching. Polished thin sections of 
 the preleach and posdeach samples are prepared and examined using 
 a polarizing microscope. Qualitative and quantitative chemical in- 
 formation is obtained using an electron probe microanalyzer 
 equipped with an X-ray analyzer. This system has the capability 
 to analyze an area only a few micrometers in diameter, as well as 
 X-ray mapping capabilities, which are used to display elemental 
 distributions within a sample. 
 
 In addition to the chemical analyses, porosity and permeability 
 measurements are performed on the core samples used in the 
 laboratory experiments. To determine the effective porosity of a 
 sample, it is first dried under vacuum and weighed, followed by 
 a vacuum saturation with distilled, deionized water and a second 
 weighing. After the total volume of the sample is measured, the 
 effective porosity is calculated using the volume of pore water after 
 saturation. The permeability of each sample is determined after the 
 injection of distilled, deionized water is initiated at the beginning 
 of the experiment. The following equation, derived from Darcy's 
 law, is used to calculate the permeability (5): 
 
 K = nVL , 
 APt 
 
 where K = permeability, D, 
 
 n = viscosity of fluid at measured temperature, cP, 
 
 V = volume of fluid, cm 3 , 
 
 L = length of sample, cm, 
 
 A = cross-sectional area of sample, cm 2 , 
 
 P = pressure, atm, 
 and t = time, s. 
 
 The field permeability of a particular ore deposit will most 
 surely be quite different from that measured in a core sample in 
 the laboratory, owing to the appreciable impact of large fractures 
 and faults occurring in the deposit. However, the main purpose of 
 measuring and monitoring permeability during the experiments is 
 to assess the relative changes in permeability that occur as the 
 experiment progresses, rather than to focus on the absolute 
 permeability of the core samples. 
 
 EXPERIMENTAL RESULTS AND DISCUSSION 
 
 Fluid chemistry data generated in the core-leaching experiments 
 are evaluated to provide input for making decisions in the field on 
 such parameters as the number of wells operating at any given time, 
 well spacing, and residence time of the fluid in the ore body. These 
 three parameters have an effect on the copper loading in solution 
 and the recovery rate of copper from the deposit. The minimum 
 copper concentration in solution necessary to ensure the efficient 
 operation of the surface recovery facility is maintained by controlling 
 these three parameters. Among the mechanisms that influence the 
 copper recovery rate are (1) chemical kinetics of mineral-acid reac- 
 tivity, (2) acid and copper diffusion through reaction product layers, 
 (3) acid and copper diffusion between the main flow channels and 
 the other pores and microfractures in the rock, (4) chemical 
 equilibria affecting the mobilization of copper to the recovery wells 
 after dissolution, particularly pH controls, and (5) adsorption of 
 mobilized copper onto mineral surfaces in the rock [Davidson (6)]. 
 Core-leaching experiments provide a measure of the net effect of 
 these mechanisms, which can then be used as input for the design 
 of a true in situ mining well field. 
 
 Some examples of data from both the core-leaching experiments 
 and the petrologic studies are presented here to illustrate the type 
 of data such studies yield and how this information is evaluated 
 with respect to in situ mining. Each ore deposit displays its own 
 unique characteristics; however, if the fundamental reaction 
 mechanisms operating during leaching can be identified and quan- 
 tified in the laboratory, then the evaluation of a particular deposit 
 for in situ mining will be simplified. Thorough petrologic and elec- 
 tron microprobe studies of the rock samples both before and after 
 the experiments provide information that is very useful in explain- 
 ing the resulting leachant chemistry and identifying the various reac- 
 tions operating during leaching. 
 
 To illustrate the significance of mineralogy and texture on the 
 leaching characteristics of an ore, results of experiments using two 
 oxide copper ores with distinct differences in their texture and 
 mineralogy are presented, which show the very different reactivities 
 that these ores exhibit. Both ores were from deposits in Pinal 
 County, AZ. 
 
 Chrysocolla Hosted in Granodiorite Porphyry 
 
 The first type of ore studied was a granodiorite porphyry that 
 contained chrysocolla as the principal copper mineral. The porphyry 
 is significantly altered, resulting in a permeable, porous rock. 
 Thorough petrologic studies of many thin sections of preleach 
 samples determined that the average mineralogy of this ore was 
 33 pet quartz, 30 pet plagioclase feldspar (one-half altered to copper- 
 bearing clay), 16 pet potassium-feldspar, 12 pet biotite (one-half 
 altered to chlorite and clay), 6 pet chrysocolla in fractures, and 3 
 pet limonite. Chemical analyses of this ore resulted in an average 
 copper content of 2.3 wt pet. To assess the potential copper removal, 
 it is important to determine the extent and distribution of the cop- 
 per in various minerals. This particular ore displayed the follow- 
 ing copper distribution: 55 pet in chrysocolla, distributed mainly 
 in fractures but also in altered plagioclase feldspar phenocrysts; 40 
 pet in clay minerals, distributed in altered plagioclase feldspar and 
 biotite phenocrysts; and 5 pet in limonite mineralization (1) (fig. 
 9). Copper mobilization from these various minerals is dependent 
 on both the accessibility of the ore minerals to the leach solution 
 and the mineral solubility. The textural relationships between the 
 copper-bearing minerals and the fractures in the rock will deter- 
 mine the leach solution accessibility. Copper mineralization located 
 along fractures (figs. 9-10) will have greater accessibility to the 
 
26 
 
 2 6 1 y 
 
 m 
 
 Figure 9.— Left, backscattered electron image (BED of chrysocolla veinlet (center) crosscutting an altered biotite phenocryst in preleach sample 
 of granodiorite porphyry ore. Right, BEI image with computer-enhanced X-ray overlay maps that show the distribution of single elements or 
 combinations of elements: copper (red), iron (green), potassium (dark blue), and iron and potassium (light blue). Copper is distributed in this 
 sample not only in chrysocolla but also in the altered biotite. The dark-blue area at the top of this map is potassium-feldspar. (These images 
 were obtained from polished thin sections on the electron probe microanalyzer. The colorized X-ray dot maps were generated by rastering an 
 electron beam across the sample annd counting the X-ray emissions of the selected elements with an energy-dispersive spectrometer.) 
 
 Figure 10.— -Quartz monzonite crosscut by numerous factures filled by atacamite and clay mineralization; photomicrograph. 
 
 ^^H 
 
 ■H 
 
27 
 
 leach solution than will disseminated copper mineralization (fig. 
 11), especially if in the latter case the matrix permeability and frac- 
 ture density are relatively low. Postleach petrologic examinations 
 of areas that were contacted by the leach solution will also yield 
 information on the relative solubilities of both the ore and gangue 
 minerals. 
 
 Three experiments were conducted on the chrysocolla ore using 
 three different leach solution concentrations: 5, 15, and 25 g/L 
 sulfuric acid (/). These experiments were conducted in the acrylic 
 reaction cells with the leach solution delivered to the samples with 
 a peristaltic pump (low-pressure system described earlier). Leachant 
 was not recirculated during these experiments; thus, fresh leach 
 solutions were continuously injected into the samples. The core 
 samples were approximately 13 cm in length and had diameters 
 near 5 cm. The experiments were conducted at ambient temperature 
 and pressure conditions. 
 
 Figures 12, 13, and 14 display the resulting fluid chemistry 
 from experiments conducted on this ore at the three different con- 
 centrations of sulfuric acid. When this information is combined with 
 the results of the preleach and postleach petrologic and microprobe 
 studies, conclusions can be drawn about the various reactions 
 operating during leaching. For example, the graph of copper con- 
 centration versus time (fig. 12) reveals that no copper was mobilized 
 during the first 200 to 500 h of the experiment. However, ap- 
 preciable amounts of calcium and sodium were mobilized during 
 these initial stages of the experiments (figs. 13-14). X-ray diffrac- 
 tion and microprobe studies of the clay minerals in this ore iden- 
 tified significant calcium and sodium in the exchangeable sites of 
 the clays. This, along with the elevated pH of the fluids during this 
 
 time, led to the conclusion that no copper was mobilized in the in- 
 itial period of the experiments because the acid was totally con- 
 sumed by ion-exchange reactions taking place, which left no H + 
 available for copper dissolution. After H + replaced all the ex- 
 changeable cations in the clays, as evidenced by the rapid decrease 
 in calcium and sodium in solution, copper dissolution proceeded. 
 Thus, if a significant amount of clay mineralization is present in 
 a deposit, there exists a potential for considerable acid consump- 
 tion. The fact that there is an initial delay in copper production, 
 yet high reagent consumption, may have a significant impact on 
 the economic analysis of a deposit. Therefore, initial microprobe 
 and leaching studies will help to quantify the significance of ion- 
 exchange reactions for a particular ore. The Bureau is currently 
 investigating various solutions that may be injected into the ore prior 
 to acid injection to reduce or eliminate the impact of ion-exchange 
 reactions on the leaching process. 
 
 The leachant chemistry (figs. 13-14) and postleach petrologic 
 analyses indicate that the biotite in this rock undergoes attack by 
 the sulfuric acid leach solutions and contributes aluminum, iron, 
 magnesium, and potassium to the leach solutions. While some 
 aluminum and iron were removed from the clay minerals and 
 limonite, respectively, biotite dissolution and the initial ion-exchange 
 reactions in the clay minerals were the main mechanisms occurring 
 in this rock that consumed acid. However, even though biotite 
 dissolution was occurring, postleach microprobe studies revealed 
 that only a minor amount of the sorbed copper present in the biotite 
 (fig. 9) was removed. Most of the copper in solution was contributed 
 by chrysocolla and altered plagioclase phenocrysts, with a minor 
 amount from limonite (7). The implications of this are significant, 
 
 »V # 
 
 Figure 11.— Quartz monzonite containing disseminated atacamite mineralization; photomicrograph. 
 
28 
 
 as previous studies have found that as much as 15 pet of the total 
 copper in a sample is present in the biotite (7). Also, most of the 
 copper contained in the clay minerals was solubilized during 
 leaching, with little evidence of significant sorbed copper remain- 
 ing in the clay minerals examined in the postleach samples. This 
 again illustrates the value of the information obtained from the 
 leaching and petrologic studies. 
 
 In order to obtain the required information to perform an 
 economic evaluation of in situ mining on a larger scale, it is 
 necessary to determine the consumption of acid per unit of copper 
 produced and the dependence of acid consumption on leach solu- 
 tion concentration. For example, using the general equation for 
 chrysocolla dissolution by sulfuric acid: 
 
 CuSi0 3 -2H 2 + 2H + = Cu +2 + Si0 2 + 3H 2 0, 
 
 the stoichiometric acid consumption is calculated to be 1.54 g 
 sulfuric acid per 1 g of copper removed. By comparing the acid 
 consumption observed in each experiment with the calculated 
 stoichiometric value, one can evaluate the severity of gangue mineral 
 acid consumption and draw correlations between consumption and 
 leach solution concentration. As previously discussed, experiments 
 performed on the granodiorite porphyry ore resulted in an initial 
 very high consumption of acid due to ion-exchange reactions with 
 the clay minerals (fig. 15), which decreased rapidly as the ion- 
 
 exchange reactions neared completion. As copper mineralization 
 becomes depleted, gangue mineral consumption of acid steadily in- 
 creases, and it is proportional to acid concentration. Graphs such 
 as figure 15 illustrate that core-leaching experiments can help to 
 evaluate the reactivity of various types of ores and can also aid in 
 determining the optimum leach solution concentration for specific 
 ore types, as acid consumption will have to be weighed against cop- 
 per loading in solution and copper recovery rate (1). 
 
 The net permeability of the three samples increased during all 
 three experiments, despite the fact that sulfate precipitation occurred. 
 Sulfate concentrations in the leach solutions revealed that sulfate 
 precipitation was occurring early in the experiments, before the con- 
 centrations rebounded to the initial level of the fresh leach solu- 
 tion. These depletions corresponded to the initial elevated calcium 
 concentrations observed in the experiments. Therefore, it is likely 
 that gypsum precipitation occurred during this time, as gypsum was 
 observed in the postleach core samples. Also, the first few leachant 
 samples contained gypsum that had precipitated from solution. 
 However, the dissolution of chrysocolla had a stronger influence 
 on permeability, resulting in the net increases in permeability for 
 the three experiments. These increases ranged from 4 to 1 10 times 
 the initial value, with the large increases resulting from the observed 
 dissolution of fracture-hosted mineralization. It should also be noted 
 that no chloride was present in the leachant because of the absence 
 of chloride mineralization in this ore type. 
 
 3 
 
 u 
 
 10,000 
 
 8,000 
 
 6,000 
 
 4,000 
 
 2,000 
 
 KEY 
 25 g/L H 2 S0 4 
 15 g/L H 2 S0 4 
 5 g/L H 2 S0< 
 
 **> *b ^Oo '-tQo *'OQo <*b <4fc 3 '*Qo 
 
 Time, h 
 
 Figure 12.— Copper concentration in solution versus time for three core-leaching experriments using granodiorite porphyry-hosted chrysocolla ore. 
 
29 
 
 500 1,000 1,500 2,000 2,500 3,000 
 Time, h 
 
 Figure 13.— Aluminum, calcium, and iron concentration in solution versus time for three core- 
 leaching experiments using granodiorite porphyry-hosted chrysocolla ore. 
 
30 
 
 Dim 
 400 
 
 i ■ ■ — i — 
 
 5 
 
 g/L 
 
 i 
 
 H 2 S0 4 
 
 T 
 
 1 
 
 KEY 
 
 Na 
 
 300 
 
 
 
 
 
 
 Mg 
 K 
 
 200 
 
 
 
 
 
 
 - 
 
 100 
 
 -^—- i i 
 
 
 
 
 
 
 - 
 
 n 
 
 
 
 
 500 
 
 
 DO 
 2 
 
 500 1,000 1,500 2,000 2,500 3,000 
 Time, h 
 
 Figure 14.— Magnesium, potassium, and sodium concentration in solution versus time for 
 three core-leaching experiments using granodiorite porphyry-hosted chrysocolla ore. 
 
31 
 
 4oo *Qo 7 <9ao r '<kb *'<>ao **b *'*b 3 '*Qo 
 
 Time, h 
 
 Figure 15. — Acid consumption versus time for three core-leaching experiments using granodiorite porphyry-hosted chrysocolla ore. 
 
 Atacamite Hosted in Quartz Monzonite 
 
 A second type of ore examined was a quartz monzonite, which 
 contained atacamite as the predominant copper mineral. The rock 
 is medium to coarse grained and exhibits a relatively low matrix 
 permeability. Petrologic studies of preleach thin sections determined 
 that the average modal mineralogy of this ore was 40 pet quartz, 
 25 pet potassium-feldspar, 20 pet sericite, 5 pet atacamite, 5 pet 
 limonite, and 5 pet clay mineralization. The copper content of these 
 samples ranged from 3 to 9 wt pet, with the large variability 
 attributable to the substantial fracture density variations from sam- 
 ple to sample. Unlike the granodiorite porphyry -hosted chrysocolla 
 ore previously discussed, nearly all of the copper in this rock is 
 located in fracture-hosted atacamite mineralization (figs. 10, 16), 
 with minor amounts distributed in the clay minerals and limonite 
 (7). These differences in mineralogy and texture result in very dif- 
 ferent leaching characteristics from those displayed in the 
 chrysocolla ore. 
 
 Three experiments were conducted on this ore using three dif- 
 ferent leach solution concentrations: 10, 20, and 40 g/L sulfuric 
 acid (/). These experiments were also conducted in the acrylic reac- 
 tion cells, with the leach solution delivered to the samples with a 
 peristaltic pump. Leachant was not recirculated during these ex- 
 periments; thus, fresh leach solutions were continuously injected 
 into the samples. The samples used in these experiments were split 
 pieces of 5-cm-diameter core approximately 6 cm long. These 
 
 experiments were also conducted at ambient temperature and 
 pressure conditions. 
 
 Because of the high ore grade, the distribution of nearly all 
 ore mineralization in fractures, and the relative absence of acid- 
 consuming gangue minerals, copper loadings in solution were very 
 high in these experiments (fig. 17). The absence of acid-consuming 
 gangue mineralization is indicated by the fluid chemistry from the 
 experiments (figs. 18-19) and was also supported by petrologic 
 studies. Only a small number of ion-exchange reactions with clays 
 occurred, as shown by the rapid decrease in calcium, potassium, 
 and sodium in solution. A relatively small amount of iron was mobi- 
 lized as the result of limonite dissolution. The absence of biotite 
 in these samples resulted in virtually no magnesium mobilization. 
 
 The low reactivity of the gangue minerals in these samples is 
 reflected in the graph of unit acid consumption versus time for the 
 experiments (fig. 20). The average consumption for the three ex- 
 periments ranged from 1 .2 to 1 .5 g sulfuric acid per gram of cop- 
 per. In the dissolution of atacamite, 
 
 Cu 2 (OH) 3 Cl + 4H + = 2Cu +2 + HC1 + 3H 2 0, 
 
 it is seen that hydrochloric acid is generated. Thus, 1.5 mol of 
 sulfuric acid (3 H + ) is consumed in the production of 2 mol of cop- 
 per, resulting in an acid consumption of 1.16 g sulfuric acid per 
 1 g of copper removed. Thus, the acid consumption observed in 
 these experiments is very close to this stoichiometric consumption. 
 
32 
 
 1 6 3 
 
 r 
 
 m 
 
 Figure 16. — Left, backscattered electron image (BED of fracture-hosted atacamite mineralization (brightest areas) in preleach sample of quartz 
 monzonite ore. Right, BEI image with computer-enhanced X-ray overlay maps that show the distribution of copper (red), iron (green), and silicon 
 (blue). Copper is distributed almost entirely in fractures. The dark-blue areas correspond with quartz, and the green areas correspond with limonite. 
 
 40,000 
 
 10,000 
 
 5,000 
 
 1,000 
 
 KEY 
 40 g/L H 2 S0 4 
 20 g/L H 2 S0 4 
 10 g/L H 2 S0 4 
 
 3,000 
 
 4,000 
 
 2,000 
 Time, h 
 
 Figure 17. — Copper concentration in solution versus time for three core-leaching experiments using quartz monzonite-hosted atacamite ore. 
 
33 
 
 300 
 
 200 
 
 100 
 
 
 10 g/L H 2 S0 4 
 
 — i — — — 
 
 KEY 
 
 ■ 
 
 
 Fe 
 
 
 
 Al 
 Mg 
 
 V 
 
 
 
 300 
 
 GO 
 
 a* 
 
 200 
 
 100 
 
 4,000 
 
 Figure 18. — Aluminum, iron, and magnesium concentration in solution versus time for three 
 core-leaching experiments using quartz monzonite- hosted atacamite ore. 
 
34 
 
 uuu 
 900 
 
 r ■■ 1 1 
 
 
 800 
 
 10 g/L H 2 S0 4 
 
 KEY 
 
 700 
 
 - 
 
 Ca 
 
 600 
 
 \ 
 
 K 
 
 500 
 
 i 
 
 ^la~ 
 
 400 
 
 1 
 
 - 
 
 300 
 
 \ 
 
 - 
 
 200 
 
 
 - 
 
 100 
 n 
 
 
 
 4,000 
 
 Figure 19.— Calcium, potassium, and sodium concentration in solution versus time for three 
 core-leaching experiments using quartz monzonite- hosted atacamite ore. 
 
35 
 
 u 
 
 i_ 
 
 Q. 
 60 
 
 c 
 o 
 
 »mm 
 
 E 
 
 3 
 «o 
 
 C 
 
 8 
 
 • MM 
 
 1 
 
 KEY 
 40 g/L H 2 S0 4 
 20 gft. H 2 S0 4 
 10 g/L H 2 S0 4 
 
 1,000 
 
 2,000 
 Time, h 
 
 3,000 
 
 4,000 
 
 Figure 20. — Acid consumption versus time for three core-leaching experiments using quartz monzonite-hosted atacamite ore. 
 
 An important observation is that the acid consumption in these three 
 experiments was relatively independent of acid concentration, unlike 
 acid consumption with the granodiorite porphyry ores. Therefore, 
 a higher acid concentration can be used to achieve greater copper 
 loadings in solution and a greater recovery rate without a prohibitive 
 increase in acid consumption. 
 
 As in the experiments conducted on the chrysocolla ore, 
 permeability also increased in all three of the atacamite experiments. 
 The initial permeabilities of these samples were very low, and nearly 
 all copper mineralization was fracture hosted, which led to extremely 
 large increases in permeability relative to the initial values as 
 atacamite was leached from the extensive fracture network. 
 Permeability increased from 100 to 1,300 times in these samples 
 as leaching progressed. It was initially very difficult to inject the 
 leach solution into the samples, but eventually the flow rate increased 
 as atacamite was removed from the fractures. Thus, in situ mining 
 may be applicable to ores that have low initial permeabilities, 
 because mineral dissolution increases the permeability. A small 
 amount of gypsum was observed on the sericite phenocrysts in the 
 postleach samples. However, little S0 4 " 2 depletion was observed 
 in the leachant, indicating that sulfate precipitation was relatively 
 minor in these experiments. The molar ratio of copper to chlorine 
 was near 2 throughout the experiment, indicating that very little, 
 if any, chloride precipitated during leaching. 
 
 Applicability of Laboratory Experiments to Deposit 
 Evaluations 
 
 Conducting a few core-leaching experiments and combining 
 leaching chemistry data with preleach and postleach petrologic 
 studies of the samples can provide a great deal of information about 
 the applicability of in situ mining for a particular type of ore. Prior 
 to initiating an in situ mining operation, both the types and abun- 
 dances of gangue minerals present in the various ore types should 
 be identified, as well as their reactivities with the leach solution. 
 This identification will also assist in optimizing the leach solution 
 concentration and will help identify and control potentially detrimen- 
 tal secondary reactions that may occur during leaching. For exam- 
 ple, clay minerals can consume large amounts of acid in the ex- 
 change of calcium, resulting in gypsum precipitation. Therefore, 
 it may be necessary to treat the ore body with a solution prior to 
 leaching with acid in order to remove the exchangeable clay ions 
 that can have deleterious effects on in situ mining. Alternatively, 
 it may be necessary to avoid leaching certain parts of the ore body 
 in order to eliminate problems that may affect a larger portion of 
 the well field. It is also important to determine the spatial and mineral 
 distribution of copper in the ores. The solubilities of the various 
 copper-bearing phases and the accessibility of leach solution to them 
 will control the total copper recovery and recovery rates. The six 
 
36 
 
 experiments described above resulted in copper recoveries rang- 
 ing from 57 to 90 pet; leach solution accessibility was the major 
 limiting factor influencing copper recovery. Although copper 
 recovery achieved in the laboratory may be different from that ob- 
 
 tained in the field because of scale differences, relatively high 
 recoveries can be attained if an adequate portion of the ore minerals 
 can be contacted by the leach solution. 
 
 SUMMARY 
 
 The Bureau has developed a comprehensive laboratory research 
 program to study both the chemical and physical influences on in 
 situ mining. The three main areas of this program are core-leaching 
 experiments, monitoring the resulting fluid chemistry from these 
 experiments, and preleach and postleach petrologic studies. Core- 
 leaching experiments more closely simulate the nature of the 
 solution-ore mineral contact in situ than do experiments with crushed 
 rock. Three core-leaching systems are currently being used to con- 
 duct experiments on oxide copper ores. The basic design is similar 
 in all three systems; the main differences are in the materials used 
 to construct the reaction cells and the pumps used to deliver the 
 leach solution to the samples. The necessary features of each type 
 of system include flexibility to accommodate a variety of sample 
 sizes and shapes, nonreactivity with the leach solution, relatively 
 low cost, and ability to recover postleach ore samples for petrologic 
 studies. These systems cover the pressure range of to 1,000 psi 
 and flow rates of 2 mL to several liters per day. These systems 
 can also be utilized to conduct leach solution recirculation 
 experiments to study the effects of ion accumulations in the leach 
 solution on the leaching chemistry during in situ mining. 
 
 The significance of mineralogy and texture on leaching 
 characteristics was evident when results from experiments using 
 
 two distinctly different ore types were compared. A granodiorite 
 porphyry that contained fracture-hosted chrysocolla and 
 disseminated copper in altered plagioclase phenocrysts contained 
 two gangue minerals that contributed significantly to acid consump- 
 tion. Very high initial acid consumption was due to the exchange 
 of H + for Ca +2 and Na + in the clay minerals. Copper mobilization 
 did not commence until nearly all of the exchangeable cations had 
 been depleted from the clay minerals. The other gangue mineral, 
 biotite, showed appreciable attack by the sulfuric acid leach solu- 
 tions, resulting in the mobilization of aluminum, iron, magnesium, 
 and potassium. However, postleach petrologic studies determined 
 that little of the copper contained in the biotite was solubilized. Acid 
 consumption was proportional to leach solution concentration. The 
 second type of ore studied was a quartz monzonite porphyry that 
 contained all copper mineralization as fracture-hosted atacamite. 
 This ore was relatively free of acid-consuming gangue minerals such 
 as biotite and clay minerals, resulting in maximum acid consump- 
 tions between 1.2 and 1.5 g sulfuric acid per gram of copper. In- 
 formation such as this is necessary to properly evaluate various ore 
 types for their suitability for in situ mining. 
 
 REFERENCES 
 
 1. Paulson, S. E. Core Leaching Experiments To Assess Leaching 
 Characteristics During In Situ Mining of Oxide Copper Ores. Paper in 
 Proceedings, In Situ Recovery of Minerals (Eng. Found. Conf., Santa 
 Barbara, CA, Oct. 25-29, 1987). AIME, in press. 
 
 2. Paulson, S. E., L. J. Dahl, and H. L. Kuhlman. In Situ Mining 
 Geologic Characterization Studies: Experimental Design, Apparatus, and 
 Preliminary Results. Miner. & Metall. Process., v. 4, No. 4, 1987, pp. 
 181-189. 
 
 3. Larson, W. C, J. K. Ahlness, and S. E. Paulson. The Bureau of Mines' 
 Role in the Development of True In Situ Copper Mining as a Future 
 Technology. Miner. Resour. Eng., v. 1, No. 2, 1988, pp. 171-180. 
 
 4. Cook, S. S., and S. E. Paulson. Leaching Characteristics of Selected 
 Supergene Copper Ores. Min. Eng. (Littleton, CO), v. 41, No. 1, 1989, 
 pp. 33-39. 
 
 5. Lewis, W. E., and S. Tandanand (eds.). Bureau of Mines Test 
 Procedures for Rocks. BuMines IC 8628, 1974, 223 pp. 
 
 6. Davidson, D. H., R. V. Huff, R. E. Weeks, and J. F. Edwards. 
 Generic In Situ Copper Mine Design Manual (contract J0267001, Science 
 Applications Int. Corp.). Volume II: Draft Generic In Situ Copper Mine 
 Design Manual, 1988, 454 pp.; for inf. contact J. K. Ahlness, TPO, Twin 
 Cities Res. Cent., BuMines, Minneapolis, MN. 
 
 7. Cook, S. S. Petrologic Analysis of Laboratory Core Leaching 
 Experiments. Paper in Proceedings, In Situ Recovery of Minerals (Eng. 
 Found. Conf., Santa Barbara, CA, Oct. 25-29, 1987). AIME, in press. 
 
37 
 
 METHODS FOR DETERMINING THE GEOLOGIC STRUCTURE OF AN ORE 
 BODY AS IT RELATES TO IN SITU MINING 
 
 By Linda J. Dahl 1 
 
 ABSTRACT 
 
 As part of its in situ mining research program, the U.S. Bureau of Mines is studying the 
 geologic structure of the Santa Cruz and Casa Grande West porphyry copper deposits near 
 Casa Grande, AZ. A datum joint set technique was developed to measure joint orientations 
 on unoriented drill core. Based on regional fabric and oriented-core data, a particular joint 
 set is defined as the datum joint set that is oriented in a certain direction. Other fracture 
 orientations are measured relative to this datum joint set. The major joint sets are indentified 
 by plotting lower hemisphere Schmidt equal-area projections. This technique allows the 
 structural data base of a deposit to be extended while keeping the high cost of oriented-core 
 drilling to a minimum. Methods of calculating fracture frequency and fracture density are 
 also discussed. 
 
 The major joint orientations at the Santa Cruz and Casa Grande West deposits strike 
 northeast and northwest. The most prominent joint set strikes northeast and dips 70° to the 
 northwest with a frequency of 0.76 to 0.98 fracture per foot. The frequency of other iden- 
 tified joint sets ranges from 0.19 to 0.41 fracture per foot. 
 
 INTRODUCTION 
 
 A thorough study of the geology of an ore body is impor- 
 tant in planning and operating underground, open pit, or in 
 situ mining operations. Exploratory drilling provides much of 
 the information necessary to evaluate a deposit for its mining 
 potential. The principal uses of drilling data at conventional 
 mining operations are to calculate ore reserves, predict ground 
 support requirements, minimize the amount of bad ground 
 through which headings must be driven, and minimize water 
 inflow to the mine. Drilling data are also important for plan- 
 ning an in situ mining operation. Since fractures are the 
 primary fluid flow channels during in situ mining, determining 
 the primary joint patterns and preferred orientation of frac- 
 tures containing ore minerals is important. Previous studies 
 suggest that a preferred orientation of flow is related to 
 primary fracture orientations (I). 2 This information can be 
 
 useful in designing in situ mining well patterns. Modifying well 
 field design (e.g., geometry and spacing) to take full advantage 
 of preferred flow direction and designing injection wells that 
 intersect an optimum number of mineralized fractures may 
 result in increased metal recoveries. Although this has not yet 
 been demonstrated for in situ mining, it will be investigated in 
 future experiments. 
 
 The geologic structure of an ore deposit is often deter- 
 mined with the aid of oriented drill core. However, drilling 
 costs are approximately doubled when drilling oriented core. 
 Therefore, the use of oriented core is usually kept to a 
 minimum. Thus, as part of its in situ mining research program, 
 the Bureau of Mines developed a technique to extend the struc- 
 tural data base of an ore deposit at a low cost relative to drill- 
 ing oriented core, by using datum joint sets. 
 
 BACKGROUND 
 
 The Bureau is conducting an in situ mining field research 
 project at the Santa Cruz and Casa Grande West porphyry 
 copper deposits near Casa Grande, AZ (figs. 1-2). The Santa 
 
 'Geologist, Twin Cities Research Center, U.S. Bureau of Mines, Minneapolis, 
 MN. 
 
 2 Italic numbers in parentheses refer to items in the list of references at the end 
 of this paper. 
 
 Cruz deposit is owned by ASARCO Santa Cruz Inc. and 
 Freeport Copper Co. The Casa Grande West deposit is owned 
 by Casa Grande Copper Co. The general geology and structure 
 of the Santa Cruz deposit are being studied using exploration 
 drill core. The structural geology of the Casa Grande West 
 deposit has been well documented with oriented drill core (2). 
 To date, there has been no oriented core drilled in the Santa 
 Cruz portion of the deposit. 
 
38 
 
 Figure 1.— Map location of Santa Cruz and Casa Grande West ore deposits and surrounding 
 mountains where outcrop data were obtained. 
 
 Sec. 13 
 
 SC-19A 
 
 Santa Cruz 
 deposit 
 
 (ASARCO-Freeport) 
 
 :?:■. * ■ 
 SC-46A 
 
 ir 
 
 Property line 
 
 Casa Grande West 
 deposit 
 
 : (Casa Grande Copper Co.) 
 
 Sec. 18 
 
 Clayton Rd. 
 
 1 1 A mile 
 
 Township 6 South 
 Range 4 East 
 Pinal County, AZ 
 
 Figure 2.— Santa Cruz and Casa Grande West deposits. 
 
39 
 
 As part of the in situ mining field test program, two new 
 holes were wedged off two previously drilled exploration holes 
 (SC-19 and SC-46) in the Santa Cruz deposit. Unoriented core 
 was obtained near the top of the ore zone in SC-19A (222 ft, 
 from 1,239- to 1,461-ft depth) and SC-46A (275 ft, from 1,410- 
 to 1,685-ft depth). 
 
 Holes SC-19A and SC-46A were drilled in a supergene 
 oxide ore deposit. The primary lithologic units are Oracle 
 Granite, quartz monzonite porphyry, and infrequent diabase 
 dikes. Supergene copper mineralization includes chrysocolla, 
 atacamite, and copper clays. Copper mineralization occurs 
 primarily as fracture fillings but also is disseminated. 
 
 DATUM JOINT SET TECHNIQUE 
 
 HYPOTHESIS 
 
 Based on Heidrick's (5) study of porphyry copper deposits 
 in the Southwestern United States and oriented core data from 
 the Casa Grande West deposit (2), areas that have northeast- 
 trending joints dipping to the northwest, it is hypothesized that 
 the steeply dipping joints observed in the Santa Cruz core 
 samples also trend northeast and dip to the northwest. As will 
 be discussed later, this assumption is fully supported by field 
 measurements acquired near the deposit. 
 
 PROCEDURE 
 
 A joint set dipping 70° to 80° is present in the core. This 
 steeply dipping joint set is hypothesized to strike northeast and 
 dip northwest and is defined as the datum (reference) joint set. 
 At each datum joint, as much surrounding core as possible was 
 reassembled and, where necessary, taped together with mask- 
 ing tape (fig. 3). The drill hole was assumed to be vertical, and 
 dip angles were measured and marked next to each fracture. A 
 mark was made on the datum joint indicating the dip direc- 
 tion. The length of reassembled core was placed in a core tray 
 with the dip direction of the datum joint aligned with the edge 
 of the tray. All the other pieces of core were fit together on 
 either side of the piece containing the datum joint to align 
 them in their original orientation relative to the datum joint. 
 
 The core tray edge was used as a guide to mark an "orientation 
 line" along the length of core (fig. 3). 
 
 A goniometer is used to measure joint orientations on 
 oriented drill core. Since the core used in this study was 
 unoriented, it was necessary to modify the standard 
 goniometer measuring technique. The goniometer was set at 
 the north 45° east orientation, which is an average orientation 
 of the joints in the area. Core was placed in the goniometer 
 with the orientation line on the core aligned with the 
 goniometer orientation line, ensuring that the primary joint set 
 was dipping northwest (fig. 4). Without moving the core, the 
 goniometer was then rotated to measure the relative strikes 
 and dips of all other fractures (fig. 5). 
 
 Many pieces of core were just a few inches in length. To 
 take measurements on these small pieces of core, a 25-cm-long 
 clear plastic extension tube was fabricated and used with wood 
 (25-cm length) and polyvinyl chloride (PVC) spacers (5-, 7-, 
 and 17-cm lengths) to raise and support various sizes of core to 
 the proper height. 
 
 ANALYTICAL PRECISION 
 
 Dip angles were measured to the nearest 5°. Strikes of 
 70°- to 80°-dipping joints that were not identified as datum 
 joints had relative measurements, which varied from the 
 datum joint strike by approximately ±15°. Precision in 
 goniometer measurements was approximately ±5°. 
 
 
 Figure 3.— Marking orientation line on reassembled core. 
 
40 
 
 Figure 4.— Aligning datum joint set orientation line on core with goniometer orientation line. 
 
41 
 
 Figure 5.— Measuring relative orientation of fracture. 
 
42 
 
 PRIMARY JOINT ORIENTATIONS 
 
 A total of 752 fractures, 383 from hole SC-19A and 369 
 from hole SC-46A, were measured. The joint orientation data 
 for both holes were plotted on lower hemisphere Schmidt 
 equal-area projections (4). These plots (fig. 6) show ortho- 
 gonal joint patterns, with the primary joints striking northeast 
 and northwest. Northeast-trending joints dip to the northwest 
 and southeast, and northwest-trending joints dip to the north- 
 east and southwest. 
 
 Strike-frequency rosette diagrams were prepared for all 
 fractures (fig. 7), mineralized fractures (fig. 8), and un- 
 
 mineralized fractures (fig. 9). There was no significant dif- 
 ference seen between the strikes of mineralized and un- 
 mineralized fractures. Approximately 25 pet of all fractures 
 measured contained copper mineralization (atacamite and 
 chrysocolla). The strikes of approximately 60 pet of fractures 
 containing copper mineralization were plotted in the northeast 
 quadrant, and 40 pet of the strikes were plotted in the north- 
 west quadrant. Copper mineralization was observed to be 
 most prevalent in steeply dipping fractures and many near- 
 vertical fractures (fig. 10). 
 
 >3% 
 
 
 KEY 
 
 
 
 1%-2% 
 
 I 
 
 ] 3%-4% 
 
 
 2%-3% 
 
 
 1 >4% 
 
 Figure 6.— Lower hemisphere Schmidt equal-area projections for Santa Cruz joints using datum joint set technique. Percen- 
 tages indicate the number of data points that fall within a 1-pct area of the circle. 
 
43 
 
 20% 15% 10% 5% 
 
 SC-19A 
 
 E W 
 
 5% 10% 15% 20% 20% 15% 10% 5% 5% 10% 15% 20% 
 
 SC-46A 
 
 Figure 7.— Strike-frequency rosettes for all fractures, holes SC-19A and SC-46A. 
 
 20% 15% 10% 5% 5% 10% 15% 20% 20% 15% 10% 5% 5% 10% 15% 20% 
 SC-19A SC-46A 
 
 Figure 8.— Strike-frequency rosettes for mineralized fractures, holes SC-19A and SC-46A. 
 
 20% 15% 10% 5% 5% 10% 15% 20% 20% 15% 10% 5% 5% 10% 15% 20% 
 
 SC-19A SC-46A 
 
 Figure 9.— Strike-frequency rosettes for unmineralized fractures, holes SC-19A and SC-46A. 
 
44 
 
 SC-19A 
 N 
 
 Figure 10.— Lower hemisphere Schmidt equal-area projections for Santa Cruz mineralized fractures. Percentages indicate the 
 number of data points that fall within a 1-pct area of the circle. 
 
 VERIFICATION OF DATUM JOINT SET HYPOTHESIS 
 
 The joint patterns of the Casa Grande West deposit are 
 well documented from previous oriented-core drilling (2). A 
 composite lower hemisphere Schmidt equal-area projection 
 showing the Casa Grande West joint sets (2) is shown in figure 
 11. The four predominant joint sets are (1) a northeast- 
 trending set that dips approximately 50° to the southeast, (2) a 
 northeast-trending set that dips approximately 45° to the 
 northwest, (3) a northwest-trending set that dips approximate- 
 ly 45° to the northeast, and (4) a northwest-trending set that 
 dips approximately 40° to the southwest. White (2) did not 
 report many steeply dipping fractures, postulating that this 
 was due to the undersampling of near-vertical joints by a ver- 
 tical drill hole. This undersampling will be discussed in detail 
 in the next section, "Fracture Frequency." 
 
 Table 1 is a comparison of joints identified using oriented 
 core in the Casa Grande West deposit with the joints identified 
 using the datum joint set technique at the adjacent Santa Cruz 
 deposit. The values are averages for all measurements re- 
 corded. There is good correlation between the data obtained 
 
 Table 1.— Comparison of joint orientation data (average strike- 
 dip) from Santa Cruz and Casa Grande West ore deposits 
 
 Deposit NE-NW NE-SE NW-NE NW-SW 
 
 Casa Grande West (2) 45° 50° 45° 40° 
 
 Santa Cruz: 
 
 SC-19A 70° 45° 50° 50° 
 
 SC-46A 70° 45° 45° 45° 
 
45 
 
 using oriented core and the datum joint set technique. The 
 primary difference observed was that Santa Cruz joints trend- 
 ing northeast and dipping northwest are somewhat steeper 
 than those measured at Casa Grande West. One possible ex- 
 planation for this difference is that the data collected at the 
 Casa Grande West deposit were from greater depths (approx- 
 imately 1,800 to 2,950 ft deep) than at the Santa Cruz deposit 
 (approximately 1,240 to 1,685 ft deep). Many of the faults and 
 fractures in this area are observed to be listric in shape, which 
 could account for more shallowly dipping fractures at greater 
 depths. 
 
 To further verify the structural fabric of the Santa Cruz 
 deposit, joint orientations were measured at nearby outcrops 
 of the Oracle Granite, which is the same rock unit that hosts 
 the ore deposit. The nearest outcrops of the Oracle Granite are 
 located in the Sacaton Mountains to the north of the deposit, 
 Table Top Mountains to the west, and Silver Reef Mountains 
 to the southeast (fig. 1). Over 700 joints were measured on out- 
 crops surrounding the Santa Cruz Basin, using a Breithaupt 
 Kassel 3 geological stratum compass. As seen in the lower 
 hemisphere Schmidt equal-area projections developed from 
 the outcrop data (fig. 12), joint orientations are similar to 
 those observed in the core samples from the Santa Cruz and 
 Casa Grande West deposits. These field measurements verify 
 that the primary joints are steeply dipping and trending north- 
 east and northwest. 
 
 'Reference to specific products does not imply endorsement by the Bureau of 
 Mines. 
 
 Figure 11.— Lower hemisphere Schmidt equal-area projec- 
 tion for Casa Grande West joints. Contour numbers are 
 percentages that indicate the number of data points that fall 
 within a 1-pct area of the circle. (Based on data from White (2).) 
 
 Silver Reef 
 N 
 
 Table Top 
 N 
 
 
 KEY 
 
 j/i£s 
 
 
 KEY 
 
 
 r c * ■* 
 
 1%-4% 
 
 
 10%-1 3% 
 
 
 1%-2% 
 
 ; 
 
 ] 3%-4% 
 
 
 
 
 
 
 
 
 | >4% 
 
 
 4%-7% | >13% 
 
 
 2%-3% 
 
 
 
 
 
 
 
 :■■■■"■■'■ : ..\ 
 
 7%-10% 
 
 
 
 rHP^v\ 
 
 
 
 KEY 
 
 M 
 
 
 1%-3% [ 
 
 
 5%-7% 
 
 
 
 >7% 
 
 3%-5% 
 
 Figure 12.— Lower hemisphere Schmidt equal-area projections for Sacaton, Silver Reef, and Table Top Mountains. Percentages 
 indicate the number of data points that fall within a 1-pct area of the circle. 
 
46 
 
 FRACTURE FREQUENCY 
 
 A common technique used to estimate fracture frequency 
 is simply to count the number of fractures intersected in a drill 
 hole and divide it by the drill-hole length. However, these 
 estimates are dependent on the orientation of the observation 
 line relative to the joints. In many porphyry copper systems in 
 the Southwestern United States, the strikes of fractures are not 
 random (J), and a change in observation line can cause large 
 changes in observed fracture frequencies. Thus, there is a 
 natural bias against intersecting joints as the angle between the 
 joints and the observation line decreases. The actual fracture 
 frequency is greatest in a line normal to the joint set (fig. 13). 
 Therefore, fracture frequencies based on the number of frac- 
 tures intersected in a drill hole are misleading. The following 
 calculations indicate that actual fracture frequencies may be 
 underestimated by as much as two-thirds by the direct observa- 
 tion method. Observed fracture frequency can be converted to 
 actual fracture frequency (number of fractures encountered 
 along a line normal to the dip of the joint set) using the follow- 
 ing equation: 
 
 Actual fracture frequency = (N d /L d )/cos 9, (1) 
 
 where N d = number of fractures observed in a drill hole, 
 
 L d = length of drill hole, 
 and 6 = dip of joint sets. 
 
 Actual fracture frequencies were calculated for each joint 
 set, using the data obtained with the datum joint set technique 
 (table 2). The effect of observation line bias is most 
 dramatically seen in the steeply dipping joint set (striking 
 northeast and dipping 70° northwest), where the observed 
 fracture frequency is only a third of the actual fracture fre- 
 quency. 
 
 Table 2.— Fracture frequencies 
 
 Hole 
 
 SC-19A 
 
 SC-46A 
 
 Orientation 
 
 Fracture 
 
 per foot 
 
 Strike 
 
 Dip 
 
 Observed 
 
 Actua 
 
 NE 
 
 45° SE 
 
 0.29 
 
 0.41 
 
 NE 
 
 70° NW 
 
 .35 
 
 .98 
 
 NW 
 
 50° SW 
 
 .22 
 
 .34 
 
 NW 
 
 50° NE 
 
 .20 
 
 .30 
 
 NE 
 
 45° SE 
 
 .23 
 
 .34 
 
 NE 
 
 70° NW 
 
 .27 
 
 .76 
 
 NW 
 
 45° SW 
 
 .13 
 
 .19 
 
 NW 
 
 45° NE 
 
 .20 
 
 .30 
 
 Normal to 
 joint set 
 
 Joint set 
 
 Drill hole 
 
 KEY 
 L(j Length of drill hole 
 © Joint set dip 
 
 Figure 13.— Undersampling near-vertical joints in a vertical drill hole. 
 
47 
 
 FRACTURE DENSITY 
 
 Fracture frequencies calculated with a count of the 
 number of joints in a particular set that intersect a borehole 
 divided by the total borehole length (even those corrected for 
 observation line bias) do not adequately describe possible 
 fluid-carrying surfaces. Fracture frequencies calculated in this 
 way reflect only one set of fractures and not the influence of 
 all the fractures at particular depths. Another way of 
 calculating fracture frequency in drill core is by utilizing the 
 fracture surface area per core volume. Calculations using frac- 
 ture surface area per core volume will be referred to as fracture 
 densities to differentiate them from fracture frequencies 
 calculated using simple counts of fractures. 
 
 Using drill core, fracture densities can be calculated using 
 the quotient of the fracture surface area (7rrVsin f ) and the 
 volume of length of core (7rr 2 L) (fig. 14) (J). The integrated 
 densities are the sum of these quotients: 
 
 where 
 
 f=i 
 
 7rr 2 /sin f 
 irr 2 L 
 
 E 
 
 f=i 
 
 1 
 
 (sin f )L 
 
 (2) 
 
 and 
 
 n = 
 
 f = 
 t = 
 r = 
 
 f = 
 
 L = 
 
 fracture density, 
 
 each individual fracture, 
 
 total number of fractures, 
 
 radius of drill core, 
 
 angle fracture makes with long axis of core, 
 
 length of drill core. 
 
 This is a simple calculation using only the fracture angle with 
 reference to the long axis of the drill core and the total length 
 of core. A fracture density plot to overlay geologic cross sec- 
 tions can be generated and contoured to give an indication of 
 highly fractured zones, which may be potential high-flow 
 areas. 
 
 Nearly 11,000 fractures were recently measured on drill 
 core to obtain fracture densities for 6 drill holes in the Santa 
 Cruz deposit. Data are currently being compiled, and potential 
 high-flow areas are being contoured. A field test to correlate 
 actual field permeability measurements with those predicted 
 using this method is scheduled for this fiscal year (1989). 
 
 
 fracture area 
 
 
 core volume 
 
 _ 
 
 7T r 2 / sin fy 
 
 
 77-r 2 L 
 
 = 
 
 1 
 
 ( sin f ) L 
 
 where f = each individual fracture, 
 n = fracture density, 
 r = radius of drill core, 
 L = length of drill core, 
 
 and 9t = fracture angle. 
 
 Figure 14.— Fracture density using fracture area per core volume. (Based on data from Haynes (5).) 
 
48 
 
 PERMEABILITY DETERMINATIONS USING FRACTURE DENSITIES 
 
 One of the variables controlling fluid flow rates is flow 
 porosity or permeability (6). Norton (7) suggests that 
 permeability of a fractured plutonic rock may be approx- 
 imated by a simple flow porosity model. From Darcy's law for 
 tluid flow through a porous medium, it can be shown that for 
 fractured rocks, permeability is a function of both the abun- 
 dance of fractures and the cube of the fracture aperture or 
 width (5). 
 
 k.-f- (3) 
 
 where k = permeability, cm 2 , 
 
 n = fracture density, cm -1 , 
 and d = fracture aperture, cm. 
 
 Measuring fracture apertures accurately is difficult. Using ac- 
 tual field-measured permeabilities from holes SC-19A and SC- 
 
 46A, fracture apertures were calculated using equation 3 to be 
 7.5 and 4.3 /*m, respectively. Since these calculated fracture 
 aperture numbers are very small ( < 10 /mi), they have little ef- 
 fect on equation 3. If the fracture aperture (d) is held constant, 
 permeability calculated using equation 3 is most greatly in- 
 fluenced by fracture density. 
 
 Since fractures are the primary fluid flow channels during 
 in situ mining, maximizing the number of mineralized frac- 
 tures contacted by the leach solutions can result in increased 
 recoveries. Increasing the number of fractures a drill hole in- 
 tersects increases observed fracture density and therefore 
 permeability. Two of the ways this can be accomplished are by 
 hydrofracturing and by drilling directional holes. Both tech- 
 niques are costly, but the reward of higher total recovery, 
 greater flow rates, and increased control over solution flow 
 could potentially justify the added expense. 
 
 SUMMARY 
 
 The Santa Cruz deposit is a highly fractured deposit with 
 primary joints striking northeast and northwest. Orthogonal 
 fracture patterns are observed with northeast-trending frac- 
 tures dipping northwest and southeast and northwest-trending 
 fractures dipping northeast and southwest. Copper mineraliza- 
 tion is most prevalent in steeply dipping, northeast-trending 
 fractures. 
 
 The results obtained using the datum joint set hypothesis 
 were consistent with the major joint sets measured using 
 oriented drill core on the Casa Grande West deposit. The 
 results were also consistent with the regional fabric measured 
 on outcrops surrounding the Santa Cruz Basin. The or- 
 thogonal joint patterns observed in the Santa Cruz deposit are 
 also typical of porphyry copper deposits in the Southwestern 
 United States as described by Heidrick (3). The technique of 
 
 using a datum joint set can be effective in extending the struc- 
 tural data base of a deposit without adding the prohibitive 
 costs involved with extensive oriented-core drilling. 
 
 Fracture frequencies based on the number of fractures in- 
 tersected in a drill hole are misleading. Actual fracture fre- 
 quencies may be underestimated by as much as two-thirds by 
 the direct observation method. Fracture frequencies are most 
 accurately described along a line normal to the joint set and 
 can be corrected using a simple calculation. 
 
 Fracture frequencies do not adequately describe possible 
 fluid-carrying surfaces for qualitatively estimating permeabili- 
 ty. Fracture densities calculated using fracture surface area per 
 core volume reflect the effect of all joint sets and allow con- 
 touring of potential high-flow areas. 
 
 REFERENCES 
 
 1. Larson, W. C, and T. H. McCormick. Utilizing Geologic 
 Characterization Techniques To Evaluate an Unsaturated Gold 
 Deposit for In Situ Mining. Paper in Application of Rock 
 Characterization Techniques in Mine Design, ed. by M. Karmis (Proc. 
 Symp., New Orleans, LA). AIME, 1986, ch. 10, pp. 88-97. 
 
 2. White, D. H. Rock Mass Characterization and Preliminary Mine 
 Design Estimates. Casa Grande West Deposit. Casa Grande Copper 
 Co. internal rep., Feb. 1981, 114 pp.; available from L. Dahl, Twin 
 Cities Res. Cent., BuMines, Minneapolis, MN. 
 
 3. Heidrick, T. L., and S. R. Titley. Fracture and Dike Patterns in 
 Laramide Plutons and Their Structural and Tectonic Implications. 
 Ch. in Advances in Geology of the Porphyry Copper Deposits. 
 Southwestern North America, ed. by S. R. Titley. Univ. AZ Press, 
 1982, pp. 73-91. 
 
 4. Billings, M. P. Structural Geology. Prentice-Hall, 1964, pp. 
 107-115. 
 
 5. Haynes, F. M. Vein Densities in Drill Core, Sierrita Porphyry 
 Copper Deposit, Pima County, Arizona. Econ. Geol. and Bull. Soc. 
 Econ. Geol., v. 79, 1984, pp. 755-758. 
 
 6. Haynes, F. M., and S. R. Titley, The Evolution of Fracture- 
 Related Permeability Within the Ruby Star Granodiorite, Sierrita 
 Porphyry Copper Deposit, Pima County, Arizona. Econ. Geol. and 
 Bull. Soc. Econ. Geol., v. 75, 1980, pp. 673-683. 
 
 7. Norton, D., and R. Knapp. Transport Phenomena in 
 Hydrothermal Systems: The Nature of Porosity. Am. J. Sci., v. 277, 
 1977, pp. 937-981. 
 
49 
 
 COMPUTER MODELING APPLICATIONS IN THE CHARACTERIZATION 
 
 OF IN SITU LEACH GEOCHEMISTRY 
 
 By Dianne C. Marozas 1 
 
 ABSTRACT 
 
 This paper discusses U.S. Bureau of Mines research on using computerized models to 
 characterize in situ mining. Analytical results from Bureau leaching tests were input into the 
 U.S. Geological Survey (USGS) computer program NEWPHRQ, which calculated the com- 
 ponent speciation in solution and the stability of mineral phases with respect to the leach 
 solutions. Data from this type of analysis can be used to predict ore solubilities, gangue 
 mineral interferences, and metal mobilities under in situ leach mining conditions. One of the 
 most powerful uses of computer modeling programs, such as NEWPHRQ, is to follow 
 hypothetical reaction paths defined by the user and to calculate the solution compositions 
 that result. Several hypothetical models are presented that demonstrate the potential applica- 
 tions of reaction path modeling to in situ mining. With additional laboratory data on 
 dissolution rates and metastable phase equilibrium, this type of analysis can be used to 
 calculate the metal recoveries expected for a variety of mining situations. 
 
 INTRODUCTION 
 
 Computer models can be valuable tools for predicting the 
 behavior of complicated in situ leach mining systems, where a 
 large number of variable physical and chemical parameters can 
 affect the efficiency of metal recovery. Bureau of Mines 
 laboratory experiments in leaching whole-core copper oxide 
 ore have provided the preliminary data needed to initiate 
 modeling efforts in the characterization of the chemistry of 
 leaching solutions and in projecting the reaction paths of in 
 situ chemical systems. However, individual laboratory ex- 
 periments are constrained by the unique variables operating 
 under the conditions of the experiment. Geochemical com- 
 puter models can extend the predictive usefulness of 
 
 laboratory studies by numerically simulating the effects that 
 changing variables will have on leaching results. 
 
 The objective of this paper is to demonstrate the range of 
 potential applications of geochemical modeling for practical 
 use in in situ leach mining programs by presenting results from 
 several hypothetical models. The models are designed to test 
 the relative variation in projected metal recovery under a varie- 
 ty of conditions that might be encountered in the field. An ac- 
 curate computer model for characterizing in situ phenomena is 
 a more efficient way to estimate the effects of a wide number 
 of variables on metal recovery than time-consuming and costly 
 laboratory and field investigations. 
 
 BACKGROUND 
 
 Computers perform numerical simulations of in situ 
 leaching by solving the set of mathematical equations that de- 
 scribe chemical interactions in a system at equilibrium. 
 Chemical equilibrium models can be used in defining irreversi- 
 ble reaction paths and mass transfer in nonequilibrium 
 chemical systems by assuming that the overall reaction path 
 
 Geologist, Twin Cities Research Center, U.S. Bureau ot Mines, Minneapolis, 
 MN. 
 
 can be represented by a series of partial equilibrium states be- 
 tween product minerals and the aqueous phase as the system 
 evolves toward a true equilibrium state. A reaction path model 
 can predict component redistribution and variations in solu- 
 tion composition that result from chemical reactions occurring 
 in rock-water systems. By comparing the model's prediction 
 with analytical data from laboratory experiments, it is possible 
 to evaluate the validity of the reaction model and estimate its 
 accuracy in describing less well-defined systems. 
 
50 
 
 In recent reviews of the state of the art in geochemical 
 modeling, Nordstrom (l-2) : and Jenne (3) compared the wide 
 variety o\' computer programs currently available and de- 
 scribed the strengths and weaknesses of individual programs in 
 their mathematical construction and in their conceptual ap- 
 proach. Simulation of in situ leaching requires a modeling pro- 
 gram that can accept a wide variety of solution components, 
 including leach target metals and related dissolution byprod- 
 ucts that are not usually considered major constituents in 
 natural waters. The program should solve rock-water interac- 
 tions and simulate irreversible reaction progress in open 
 systems as well as solve for the solution speciation of all com- 
 ponents. Kinetic rate constraints derived from laboratory or 
 field data should be easily incorporated into the algorithm for 
 mineral dissolution or precipitation. In situ leach mining 
 systems also require that the selected program be able to han- 
 dle solutions that range in concentration from initially dilute 
 conditions to highly concentrated conditions. Finally, the 
 computer program should be usable on personal computers in 
 mining offices where mainframe computers may not be 
 available and should allow the user to easily adjust the input so 
 that a number of in situ problems can be modeled. 
 
 The computer program selected for initial Bureau efforts 
 in in situ modeling research is the USGS program NEW- 
 PHRQ, which is an updated version of computer program 
 PHREEQE (4). NEWPHRQ is a FORTRAN computer pro- 
 gram designed to model geochemical reactions based on an 
 ion-pairing model. NEWPHRQ simulates many common 
 hydrogeochemical phenomena such as mixing, rock-water in- 
 teractions, mineral equilibrium, and ion-exchange reactions. 
 The reaction path, system composition, temperature condi- 
 tions, equilibrium conditions, aqueous species, and mineral 
 phases are completely user definable, which presents a power- 
 ful approach to modeling the unique conditions of in situ min- 
 ing. NEWPHRQ is currently running on a Bureau personal 
 computer. 
 
 The PHREEQE family of geochemical models has been 
 successfully applied to a number of hydrogeochemical prob- 
 lems. Plummer (5) used reaction path calculations by PHREE- 
 QE in conjunction with mass balance calculations to identify 
 the effects that incongruent dissolution reactions have on 
 ground water compositions within a carbonate aquifer. 
 Marozas (<5) used techniques similar to Plummer's that com- 
 bined PHREEQE's geochemical reaction path modeling with 
 mass balance calculations to study the migration of base 
 metals along ground water flow paths in desert alluvial basins. 
 
 In order to extend the application of geochemical model- 
 ing to in situ leach mining, several adjustments to the com- 
 puter program are necessary. The current ion-pairing ap- 
 proach used by PHREEQE and NEWPHRQ is based on the 
 Debye-Hiickel mathematical expression for determining activi- 
 ty coefficients of aqueous species appropriate for dilute 
 aqueous solutions. However, Bureau experimental studies 
 show that, when recycling lixiviants, in situ mining will gener- 
 ate highly concentrated solutions, and thus it is important that 
 the program be modified for solutions of high ionic strength. 
 At higher concentrations, the chemical identity of the species 
 begins to play an important role in the thermodynamics of the 
 system. While some of this behavior can be described by em- 
 pirical modifications to the Debye-Hiickel limiting law, such as 
 the Davies equation, which is incorporated in NEWPHRQ, 
 many important contributions are specific to the chemistry of 
 the species. Recently developed phenomenological equations, 
 known as the Pitzer equations, appear to have the required 
 flexibility to describe specific ion interactions and have been 
 successfully applied to a number of geochemical processes 
 (7-8). Pitzer equations can be added to the NEWPHRQ code 
 to accurately model highly concentrated leaching solutions and 
 more closely characterize ore-body interactions with recycled 
 lixiviants. 
 
 POTENTIAL APPLICATIONS OF GEOCHEMICAL MODELING 
 
 Geochemical models can characterize the chemical prop- 
 erties of an in situ leach mining system in terms of its dominant 
 variables — major and minor ions, oxidation-reduction poten- 
 tial, acidity, complexing components, mineral stability, and 
 adsorbing surfaces. Computer analyses of these data provide a 
 systematic approach for defining the relative importance of 
 different variables in determining metal recoveries expected 
 during in situ mining. 
 
 Computer models can calculate the equilibrium speciation 
 of metals in solution if relevant stability constants are known. 
 This approach presents a distinct advantage over laboratory 
 methods, which are presently unable to determine unequivocal 
 identification of solution species. The term "species" refers to 
 the actual molecular form in which an element is present in 
 solution. For example, copper may exist as one of the free ions 
 Cu 1+ or Cu 2 + , or as a complex ion pair, such as Cu(OH) + , 
 Cu(S0 4 ) , or CuCl + . Speciation can be one of the critical fac- 
 tors in estimating metal recovery because the chemical 
 behavior of any element depends on its chemical form in solu- 
 tion. For example, adsorption of metals onto clay minerals 
 and metal hydroxides can be either reduced or enhanced, 
 depending on what metal complex exists in solution. At high 
 
 -Italic numbers in parentheses refer to items in the list of references at the end 
 of this paper. 
 
 metal concentrations, the presence of complexes can affect the 
 solubility of solids. Complex formation in solution among 
 constituents derived from dissolving mineral lattices or with 
 constituents of the lixiviant must be taken into account in ex- 
 ploring mechanisms of metal transport by in situ leach mining 
 solutions. 
 
 Computer models can test the effects that the variation of 
 system parameters will have on metal recoveries. Acidity ef- 
 fects can be modeled by varying the pH input into the model 
 and charting the consequent changes in metal concentration. 
 Computer models can calculate the variations expected in ore 
 mineral solubilities at different temperatures. Models can 
 simulate temperature variations along flow paths to evaluate 
 potential metal loss that is likely to occur during flow from 
 near the bottom to the surface of recovery wells. Interference 
 from gangue cations on metal recovery can be evaluated by 
 changing input solution composition or increasing the amount 
 of gangue reaction occurring along the reaction path. Kinetic 
 controls on metal recovery can be tested by mathematically 
 varying the dissolution rate of minerals into solution. Initially, 
 the predictive capabilities of computer models can be verified 
 by comparison with laboratory experiments and preliminary 
 field tests; then the model can be extended to numerically 
 simulate recoveries for conditions not duplicated in the 
 laboratory or well controlled in the field. 
 
51 
 
 RESULTS 
 
 SPECIATION AND MINERAL STABILITY 
 ANALYSIS 
 
 Data from whole-core leaching experiments with sulfuric 
 acid on oxide copper ores from deposits in Pinal County, AZ 
 (9),'' were used to initiate research in modeling in situ leach 
 chemistry. Analytical data from fluids collected at 48- to 72-h 
 intervals were input into NEWPHRQ for each leach experi- 
 ment. NEWPHRQ was then used to solve for component 
 speciation distribution and for mineral stability during time 
 steps throughout the experiment. The thermodynamic data 
 base used for this analysis is from the compilation of Ball (10). 
 Typical speciation and mineral stability results will be il- 
 lustrated by detailed analysis of one of the Bureau's core- 
 leaching experiments in which 1 -pet-copper chrysocolla ore 
 from Santa Cruz was leached with 50 g/L sulfuric acid. Figure 
 1 shows the pH variation versus the amount of solution re- 
 covered in the experiment. 
 
 Figure 2 shows how the relative concentration of copper 
 species varies as a function of the amount of solution re- 
 covered. The amount of solution recovered increases directly 
 with time as sulfuric acid is added to the system. The data 
 show that the speciation and therefore the chemical behavior 
 of copper vary throughout the experiment. Initially, metal 
 sulfate complexes dominate copper speciation in solution; 
 however, as the acidity of the recovered solution increases and 
 about 100 g of solution is recovered, free copper ion becomes 
 the dominant species. This speciation distribution implies that 
 there is less metal complexing and consequently relatively less 
 metal mobility in solution as acidity increases. The causes 
 behind the loss of sulfate complexing can be investigated by 
 analysis of the speciation behavior of sulfate in the system. 
 
 Figure 3 is a plot of sulfate speciation in the experiment and 
 shows the relative amount of sulfate that is present as metal 
 sulfate complexes, free sulfate ions, and bisulfate ions. Ini- 
 tially, free sulfate anions and metal sulfate complexes domi- 
 nate the sulfate system, but with time, acidity increases and 
 sulfate reacts to bisulfate as reaction A is driven to the right. 
 
 H + +S0 4 =-HS0 4 
 
 (A) 
 
 The fraction of hydrogen sulfate increases at the expense of 
 metal sulfate complexes, and consequently, metal mobility 
 may be reduced. Speciation analyses of particular systems can 
 help determine the lowest acidity required for continued cop- 
 per oxide dissolution yet maximize metal mobility in solution. 
 Further information on copper behavior in solution is de- 
 rived from calculations of the activity or the effective concen- 
 tration of a species in an electrolyte solution. Evaluation of 
 metal activity in solution is important because it is from this 
 measurement that the solubilities of minerals in the leaching 
 environment can be calculated. The activity of solution com- 
 ponents is related to measured molalities by the equation 
 
 where 
 and 
 
 m S 7s = a s> 
 
 m s = molality of species S, 
 
 7 S = activity coefficient of species S, 
 
 a s = activity of species S. 
 
 (1) 
 
 Activity coefficients are determined for each species in 
 NEWPHRQ by the user's choice of either an extended Debye- 
 Hiickel formula or the Davies formula. Both of these formulas 
 
 'The deposits are owned by Cyprus Casa Grande Corp. and by ASARCO San- 
 ta Cruz Inc. and Freeport Copper Co. 
 
 calculate the change in activity coefficients as a function of 
 ionic strength. 
 
 There is a considerable difference between measured con- 
 centrations of total copper and the calculated effective copper 
 
 x 
 a 
 
 200 300 400 500 
 SOLUTION RECOVERED, g 
 
 600 
 
 Figure 1.— The pH of leach solutions from 50 g/L whole-core 
 leach experiments on 1.0-pct-copper chrysocolla ore from 
 Santa Cruz. 
 
 | 0.8 
 o 
 
 o .6 
 E 
 
 </) 
 
 LU 
 
 O .4 
 
 Ld 
 
 Q. 
 
 3 
 O o 
 
 OCu + */total Cu 
 a Cu(S0 4 )/total Cu 
 
 100 
 
 200 
 
 300 
 
 400 
 
 500 
 
 600 
 
 SOLUTION RECOVERED, g 
 
 Figure 2.— Mole fraction of copper in experimental leach 
 solutions present as free metal ion and copper sulfate 
 complex. 
 
 DSO«/total S 
 
 HSOVtotal S 
 
 O Metal(S0 4 )/total S 
 
 100 200 300 400 500 
 SOLUTION RECOVERED, g 
 
 600 
 
 Figure 3.— Mole fraction of sulfate in experimental leach 
 solutions present as sulfate ion, bisulfate, or complexed with 
 cations. 
 
52 
 
 concentration or activity in the leaching solutions from this ex- 
 periment. Figure 4 shows the total concentration of aqueous 
 copper regardless of its form, which includes free or hydrated 
 copper ion and copper sulfate complexes. The molality and the 
 activity of free copper ion are also shown. The data show that 
 there is a wide variation in total copper concentration, relative 
 to the variation in the activity of free copper. A simple model 
 of copper fluctuation, therefore, may apply to in situ systems 
 that need not account for wide variations in measured copper 
 concentration but rather can focus on explaining the less 
 variable activity of copper. Future research goals of the 
 Bureau are to use computer modeling analyses of the variation 
 in copper activity to provide insights into the kinetic and 
 equilibrium mechanisms that control the in situ leaching of 
 copper. 
 
 The activity of copper is emphasized here because it is used 
 for determining the saturation state of mineral phases in the 
 system. The saturation state of an aqueous system, with 
 respect to mineral phases, can be calculated by dividing the ion 
 activity product (IAP) by the solubility product (K sp ) constant. 
 The IAP is the product of the activities of the reaction prod- 
 ucts, each raised to the power indicated by its numerical co- 
 efficient, divided by the product of the activities of the re- 
 actants raised to the power of their numerical coefficients. The 
 log of the IAP/K sp ratio is called the saturation index (SI). If a 
 solution is supersaturated with respect to a particular mineral, 
 the SI is positive and the mineral has a tendency to precipitate. 
 If the solution is undersaturated, the SI is negative and the 
 mineral has a tendency to dissolve. If a mineral is at equi- 
 librium, then the SI is zero. 
 
 Figure 5 is a plot of SI versus solution recovered for three 
 mineral phases in the Santa Cruz experiment. The SI line for 
 amorphous silica is close to zero or equilibrium throughout the 
 experiment. This implies that the leaching system responds to 
 the addition of silica from the dissolution of chrysocolla or 
 clays by precipitation of amorphous silica. The gypsum SI line 
 is close to equilibrium early in the experiment but falls to 
 undersaturated levels later. This implies that gypsum precipi- 
 tation would be a problem only in the early stages of in situ 
 leaching. The chrysocolla SI line indicates that it is at first 
 supersaturated but later becomes undersaturated in the experi- 
 ment. The calculated supersaturation of chrysocolla may have 
 several implications. First, there may be a kinetic inhibition to 
 precipitation during the supersaturation interval, which keeps 
 copper in solution long enough for recovery in the experiment. 
 However, this may not be the case in the field, where time to 
 overcome activation energy barriers to precipitation may be 
 within the residence time of the lixiviants in the host ore. Sec- 
 ond, because the activity of copper remains near steady state, 
 this may indicate that there is another, faster control affecting 
 copper concentration in solution other than chrysocolla 
 dissolution, such as the presence of unidentified metastable 
 phases. Finally, the chrysocolla solubility data used in the 
 calculation may need to be revised to account for the solid 
 solution identified by microprobe studies of chrysocolla in the 
 drill core. 
 
 APPLICATION MODELS 
 
 The most powerful application of computer modeling 
 programs, such as NEWPHRQ, to in situ mining is the pro- 
 gram's ability to follow user-defined hypothetical reaction 
 paths and to calculate the resulting solution compositions. 
 Given a set of realistic variables, a reaction path model can 
 predict expected metal recoveries, gangue mineral in- 
 terferences, and lixiviant losses. Conversely, given unexpected 
 
 metal recoveries, reaction path models can test variables that 
 might be causing anomalies and may suggest a means to max- 
 imize positive variances and minimize negative variances. 
 
 The following sections present several reaction path 
 models that simulate the effect of several variations in system 
 parameters on the leaching of Santa Cruz 1.0-pct-copper 
 chrysocolla ore with 50 g/L sulfuric acid. These models are 
 presented to demonstrate potential applications of computer 
 modeling for in situ mining. 
 
 The copper recoveries specified in figures 6 to 13 are cal- 
 culated for certain theoretical conditions and therefore should 
 not be taken as the copper recovery expected for an actual field 
 situation. Rather, the intent of this study is to show how ac- 
 curate computer models can be integrated into systematic 
 methods of predicting copper recovery under a given set of 
 conditions. 
 
 Effects of Sulfate Complexing on Copper Recovery 
 
 Cations in an aqueous medium react with available bases 
 to improve the stability of electrons in their outer shell. Metal 
 ions, such as copper, can react to achieve such stabilization 
 without the formation of precipitates by forming metal coordi- 
 nation complexes with anions or negatively charged molecules 
 called ligands. It is well known that the presence of foreign 
 
 200 300 400 500 
 
 SOLUTION RECOVERED, g 
 
 600 
 
 Figure 4.— Concentration of total copper and molality and 
 activity of free copper ion in experimental leach solutions. 
 
 a. 
 
 < 
 
 x 
 
 Ul 
 
 a 
 
 O Amorphous SI0 2 
 a Gypsum 
 
 Chrysocolla 
 
 O - 
 
 < 
 cc 
 
 => 
 
 < 
 
 V) 
 
 600 
 
 SOLUTION RECOVERED, g 
 
 Figure 5.— SI of mineral phases with respect to experi- 
 mental leach solutions. Zero line represents equilibrium; 
 positive values indicate supersaturation, and negative values, 
 undersaturation. 
 
53 
 
 ligands increases the solubility of slightly soluble metal salts 
 (11). Sulfate in sulfuric acid lixiviants increases the solubility 
 of chrysocolla during in situ mining. 
 
 It is possible to illustrate by computer modeling the sig- 
 nificance of complex formation in enhancing the solubility of 
 copper solids. In this model, chrysocolla dissolves to equi- 
 librium with solutions of varying acidity, while equilibrium is 
 also maintained with amorphous silica — first, with no sulfate 
 present and, second, with excess sulfate present for complex- 
 ing. Chrysocolla solubility was estimated from 50-g/L sulfuric 
 acid leach experiments. 
 
 Results show (fig. 6) that copper mobility is significantly 
 increased by sulfate complexation in solution. Copper re- 
 coveries are approximately 1 l A times higher when sulfate is 
 present at a pH of about 0.8. 
 
 Effects of Gangue Cations on Copper Recovery 
 
 Gangue cations compete with copper for sulfate ligand in 
 solution, and because alkali-earth sulfate aqueous complexes 
 are more stable than copper sulfate complexes, sulfate ligand 
 will preferentially complex with gangue cations, leaving copper 
 in solution as free copper ion. The presence of sulfate-seeking 
 cations from the dissolution of gangue minerals, in addition to 
 copper, has the effect of forcing copper to its free ion state. If 
 free copper is abundant in solution, copper minerals such as 
 cuprite, chalcanthite, or native copper may approach satura- 
 tion levels and precipitate. High concentrations of gangue cat- 
 ions may hinder the transport of copper by forcing precipita- 
 tion, even at low pH values. 
 
 The effects of gangue cations at low pH on copper re- 
 coveries were simulated by using NEWPHRQ to calculate the 
 total amount of copper that would enter solution during the 
 equilibrium dissolution of chrysocolla in the presence of 
 gangue cations at various pH levels. Theoretical copper re- 
 coveries were calculated for a system with (1) no gangue cat- 
 ions in solution, (2) gangue cations present at concentration 
 levels typical of those found in the late stages of the Santa 
 Cruz experiment (178 ppm calcium, 396 ppm magnesium, 
 1,090 ppm iron, and 1,390 ppm aluminum) and with 49,000 
 ppm sulfate, and (3) twice as many gangue cations in solution 
 as those in the Santa Cruz experiments (356 ppm calcium, 792 
 ppm magnesium, 2,180 ppm iron, and 2,780 ppm aluminum) 
 and 24,500 ppm sulfate. The last model might simulate condi- 
 tions expected during the recycling of in situ leach solutions. 
 
 The results of these simulations (fig. 7) show that gangue 
 cations interfere with copper mobility in solutions during the 
 equilibrium dissolution of chrysocolla in sulfuric acid lixiviant. 
 This model can predict at what level gangue mineral inter- 
 ferences with copper mobility will present an economically 
 unacceptable situation. 
 
 Effects of Ionic Strength on Copper Recovery 
 
 An efficient way to lower mining costs during in situ min- 
 ing is to reuse lixiviants that maintain sufficient acidity to 
 dissolve copper oxide minerals. During this recycling, the con- 
 tinuous dissolution of ore and host minerals increases the con- 
 centration of solutes in solution and thereby raises the ionic 
 strength of recycled solutions. The effective concentration of 
 metals in electrolyte solutions, and therefore the solubilities of 
 ore minerals, are dependent on ionic strength. As ionic 
 strength increases, solute molecules will, on the average, be 
 closer together and interactions in solution will take place 
 more readily. 
 
 Because the activity of copper changes with ionic 
 strength, solubilities will vary with ionic strength. NEWPHRQ 
 was used to investigate the effects of increasing ionic strength 
 on total copper recovery. In order to observe only the effects 
 
 0.7 
 
 .6 
 
 o 
 E 
 
 Ld 
 
 > 
 
 o 
 o 
 
 Ld 
 
 or 
 o 
 
 KEY 
 S0 4 complexing 
 
 OWith 
 □ Without 
 
 0.6 
 
 1.8 
 
 pH 
 
 Figure 6.— Equilibrium concentration of copper in presence 
 and absence of sulfate in solutions saturated with amorphous 
 silica. 
 
 pH 
 
 Figure 7.— Equilibrium concentration of copper in solutions 
 saturated with amorphous silica. A, No gangue cations; 6, 
 gangue cation concentrations are typical of one-pass leach 
 experiments with 0.54 mol/L sulfate; C, gangue cation concen- 
 trations are twice that of one-pass leach experiments with 0.26 
 mol/L sulfate. 
 
54 
 
 of ionic strength variations without gangue cations or ligand 
 concentration effects as previously considered, ionic strength 
 in the model was changed by the addition of potassium and 
 chlorine, which do not interfere with copper or copper com- 
 plexes in solution. Potassium and chlorine were added in con- 
 centrations sufficient to reach the desired level of ionic 
 strength. 
 
 The results (fig. 8) show that, as ionic strength increases at 
 the same pH, copper recovery is decreased. As ionic strength 
 increases, the activity coefficient for copper increases to values 
 greater than one, so that the effective concentration of copper 
 increases. This inhibits the solubility of chrysocolla because 
 saturation is achieved at lower copper concentrations. This 
 model uses the Davies equation for the calculation of activity 
 coefficients, but as was discussed in an earlier section, a 
 specific interaction model would probably describe the solu- 
 tion behavior more accurately at high ionic strengths. 
 
 Effects of Increasing Gangue Cations 
 on Clay Mineral Precipitation 
 
 Metal cations in solution are known to adsorb onto clay 
 minerals in rock-water systems. Solutions concentrated with 
 gangue cations can approach clay mineral saturation levels 
 even at low pH. When clay minerals precipitate in in situ 
 environments, surface adsorption sites are provided for the ad- 
 sorption and consequent loss of metal from solution. 
 
 NEWPHRQ was used to model the saturation state of 
 clay minerals during the concentration of gangue cations. 
 Silica concentrations in solution were held at equilibrium with 
 amorphous silica, and copper concentrations were held con- 
 stant at 5,000 ppm. Gangue cation concentrations were gradu- 
 ally increased at a system pH of 1 .0 and at a system pH of 3.0. 
 
 The saturation states of potassium-illite and of mag- 
 nesium-beidellite minerals in these hypothetical solutions is il- 
 lustrated by plotting the SI of each mineral (fig. 9). The zero 
 line represents equilibrium; when the SI plots above this line, 
 clay minerals precipitate, while at values below this line, clay 
 minerals dissolve. The results show that potassium clay 
 minerals approach saturation as potassium in solution reaches 
 0.05 mol/L, regardless of pH. Magnesium clays do not ap- 
 proach equilibrium even at magnesium concentrations of 1 
 mol/L. Models can predict when gangue cation buildup in re- 
 cycled solutions might begin precipitating clay minerals. 
 
 Effects of Potassium Exchange Reactions 
 on Copper Recovery 
 
 Bureau laboratory experiments indicate that a con- 
 siderable amount of acid is consumed by ion-exchange reac- 
 tions in the early phases of whole-core leaching. This acid is no 
 longer available to react with copper oxide minerals, which 
 decreases the efficiency of in situ leaching. Reduction of this 
 initial acid consumption may be required to economically 
 leach copper. 
 
 NEWPHRQ was used to model the effect of adding 
 potassium to the lixiviant while maintaining a constant number 
 of exchange sites, chrysocolla equilibrium, and amorphous 
 
 0.7 
 
 o 
 
 .5 
 
 E 
 
 
 „ 
 
 
 o 
 
 .4 
 
 Ul 
 
 
 QL 
 
 
 Ul 
 
 
 > 
 
 ..5 
 
 o 
 
 
 o 
 
 
 Ul 
 
 
 az. 
 
 .2 
 
 3 
 
 
 o 
 
 
 .1 - 
 
 KEY 
 
 Ionic strength 
 
 O0.8 
 
 □ 1.0 
 4.3 
 
 0.6 
 
 1.8 
 
 pH 
 
 Figure 8.— Equilibrium concentration of copper in solutions 
 of varying ionic strengths saturated with amorphous silica. 
 Zero line represents equilibrium. 
 
 K-illite 
 Mg-beidellite 
 
 _J l 
 
 -5 -4 -3 -2 -1 
 LOG (PREDOMINANT CATION), mol 
 
 Figure 9.— SI of clay minerals in 50 g/L sulfuric acid, amor- 
 phous silica-saturated solution with varying concentrations of 
 gangue cations. 
 
55 
 
 silica equilibrium. The pH of the system is adjusted as ex- 
 change and dissolution reactions occur. Potassium affects the 
 reaction by limiting the amount of cations exchanging for 
 hydrogen ion at clay exchange sites. Consequently, acid loss 
 from exchange reactions is diminished and chrysocolla dissolu- 
 tion is enhanced. 
 
 Figure 10 is a plot of copper recovered from the leaching 
 of chrysocolla ore as a function of potassium concentration in 
 the lixiviant. Copper recovered without added potassium is 
 shown by the "potassium-free" line. This figure shows that 
 significant differences in copper recoveries are expected only 
 when potassium exceeds 0.01 mol/L in the solution. The eco- 
 nomic feasibility of adding potassium or other outside reagents 
 to lixiviants for copper recovery enhancement can be deter- 
 mined by computer estimations of how much of the reagent 
 will be required to significantly increase copper recoveries. 
 
 Effects of Biotite Dissolution on Copper Recovery 
 
 Characterization and leaching studies on Santa Cruz ore 
 (9) have shown that the presence of biotite in the rock matrix 
 inhibits the leaching of copper. Biotite dissolves in acid, which 
 removes acid from the lixiviant, increases the pH, and pro- 
 motes copper mineral precipitation. Gangue cations released 
 by biotite dissolution compete with copper for sulfate ligand 
 and consequently can inhibit copper mobility in solution. 
 
 Biotite interference in copper leaching can be simulated 
 with computer models. The model simulates the dissolution of 
 biotite into a system in which equilibrium is maintained with 
 chrysocolla and amorphous silica at low pH. The amount of 
 biotite that dissolves is increased throughout the model to 
 gauge the effects of increased biotite reactivity in the system. 
 
 Figure 11 shows that copper recovery decreases by an 
 order of magnitude as 0. 15 mol of biotite dissolve. The extent 
 of biotite mineralization in potential ore bodies can be esti- 
 mated by petrologic examination and the consequent effects of 
 biotite reactivity on copper mobility projected by computer 
 modeling. 
 
 Effects of Copper Dissolution Rates 
 on Copper Recovery 
 
 Characterization work on thin sections of oxide ore 
 deposits (9) has revealed that copper is present in two different 
 mineral phases — chrysocolla and copper-bearing clays. 
 Chrysocolla is principally fracture hosted, while copper- 
 bearing clays are principally disseminated in the rock matrix. 
 The amount of copper released by solubilization of copper ore 
 depends on the dissolution kinetics of the two different phases. 
 Laboratory leaching studies on this type of ore show that cop- 
 per recovery reaches a maximum in the early stages of leaching 
 (fig. 12) and then decreases to a near-constant level for the re- 
 mainder of the experiment. 
 
 Path-dependent reaction progress models are simulated 
 by NEWPHRQ as a series of steps where the resulting solution 
 from one step becomes the starting solution for the next step 
 of the model. NEWPHRQ can simulate expected copper re- 
 covery from two different mineral phases by varying the rate at 
 
 -6 -4-2 2 
 
 LOG (K ADDED TO SOLUTION), mol/L 
 
 Figure 10.— Equilibrium concentration of copper in 50 g/L 
 sulfuric acid, amorphous silica-saturated solution as function 
 of potassium concentration in lixiviant. "Potassium-free" line 
 marks copper concentration level when no potassium is pres- 
 ent. 
 
 0.24 
 
 o 
 
 o 
 
 UJ 
 
 on 
 
 Ld 
 > 
 
 o 
 o 
 
 UJ 
 Q£ 
 
 3 
 O 
 
 0.20 
 
 BIOTITE DISSOLVED INTO SOLUTION, mol 
 
 Figure 11.— Equilibrium concentration of copper in 50 g/L 
 sulfuric acid, amorphous silica-saturated solution as function 
 of increasing biotite reactivity. 
 
56 
 
 which each phase is added to the system, allowing the solution 
 to come to equilibrium with amorphous silica and passing the 
 solution on to the next step. Depletion of fracture-hosted 
 chrysocolla is simulated by stopping the chrysocolla dissolu- 
 tion reaction while maintaining the matrix clay dissolution at 
 the same rate as before. Acid is added at each step to simulate 
 the continual influx of acid to the system. 
 
 The results of the model are shown in figure 12, where a 
 peak similar to that observed in the laboratory experiment is 
 generated. The results of this model suggest that differing 
 dissolution kinetics of copper-bearing phases are potential 
 controls on copper recoveries in the experiment. This example 
 illustrates one way that computer simulations may explain the 
 reactions responsible for copper release and subsequently can 
 predict expected copper recoveries. 
 
 Effects of Temperature Variations 
 on Copper Recovery 
 
 Temperature variation during in situ mining may affect 
 copper recoveries. Temperatures may vary considerably when 
 ore deposits are leached at depth in areas with high thermal 
 gradients. The basic relationships that describe the influence 
 
 0.6 
 
 1 1 1 
 
 Chrysocolla dissolution ceases 
 
 O Model 
 
 D Experiment 
 
 8 
 
 pH 
 
 Figure 12.— Effects of two different rates of ore mineral 
 dissolution on copper concentration. Experimental data are 
 from 50 g/L sulfuric acid leach experiments on 1-pct-copper 
 chrysocolla ore. At the point indicated on the diagram, 
 chrysocolla dissolution ceases while copper-clay minerals 
 continue to dissolve. 
 
 of temperature on the equilibrium constant of a chemical reac- 
 tion or phase equilibrium require knowledge of the enthalpy of 
 reaction, the heat capacity of the reaction, and the variation of 
 the heat capacity with temperature. Heat capacity data on 
 chrysocolla solid solutions are limited or unavailable, and en- 
 thalpy data must be estimated. Until more accurate data are 
 obtained from direct calorimetry and by measurement of equi- 
 librium properties of chemical reactions, predictions of copper 
 recovery by chemical models are necessarily hypothetical. The 
 following is presented only as an example of how models can 
 calculate copper recoveries when thermodynamic values for 
 the mineral phases of interest in the system are assessed. 
 
 Computer simulations are especially useful for modeling 
 changing temperature conditions because of the complexity of 
 the effects of temperature on the entire system. In general, in- 
 creasing temperature favors dehydration reactions, solubiliza- 
 tion of chrysocolla, and ion association reactions, while acid 
 dissociation diminishes. 
 
 The model used to demonstrate temperature variation ef- 
 fects on in situ mining was designed so that chrysocolla dis- 
 solves to equilibrium as the pH varies and the system is main- 
 tained in equilibrium with amorphous silica. Figure 13 il- 
 lustrates how temperature can affect copper recovery when 
 calculated with an estimated chrysocolla enthalpy. Results 
 show that at similar acidity levels more copper is recovered at 
 100° C than at 25° C. 
 
 pH 
 
 Figure 13.— Equilibrium concentration of copper in 50 g/L 
 sulfuric acid, amorphous silica-saturated lixiviant at 25° C and 
 100° C. 
 
57 
 
 SUMMARY 
 
 It has been demonstrated that computer modeling pro- 
 grams have applications for in situ mining through the quanti- 
 fication of factors that affect the leaching chemistry. To esti- 
 mate ore recovery and lixiviant losses, knowledge of ore 
 mineral solubility and in situ fluid-rock interactions is essen- 
 tial. Geochemical models can characterize the chemistry of in 
 situ systems in terms of the speciation of solution components 
 and can identify those mineral phases that should dissolve and 
 those that should precipitate under leach mining conditions. 
 
 Several models have been developed to demonstrate how 
 computer simulations can be used to predict ore recoveries for 
 a variety of conditions not evaluated by laboratory testing. 
 These hypothetical models calculate the effects of different 
 copper dissolution rates on the behavior of copper recovery 
 curves, the effects of preflushing with potassium to diminish 
 hydrogen exchange, and the effects of temperature on chryso- 
 colla solubility. 
 
 Although it is generally known that highly reactive gangue 
 constituents in the host matrix can consume acid and raise 
 mining costs, mathematical treatment of gangue reactivity can 
 project acid consumption and establish at what gangue con- 
 centrations it would no longer be economically feasible to 
 leach-mine. Several different models that address the impact 
 of secondary reactions on copper recovery show the wide 
 variety of approaches to this problem that computer simula- 
 tions make possible. 
 
 The models presented in this paper are initial attempts to 
 use computer simulations for predicting copper behavior dur- 
 ing in situ leaching. Future work by the Bureau will focus on 
 continued refinement of the geochemical models by calibra- 
 tion of the data base with experimental results, by the incor- 
 poration of known kinetic constraints into the program, and 
 by testing of the program under actual field conditions. 
 
 REFERENCES 
 
 1. Nordstrom, D. K., L. N. Plummer, T. M. L. Wigley, T. J. 
 Worley, J. W. Ball, E. A. Jenne, R. L. Bassett, D. A. Crerar, T. ML 
 Florence, B. Frit, M. Hoffman, G. R. Holdren, Jr., G. M. Lafan, S. 
 V. Mattigod, R. E. McDuff, F. Morel, M. M. Reddy, G. Sposito, and 
 J. Thailkill. A Comparison of Computerized Chemical Models for 
 Equilibrium Calculations in Aqueous Systems. Ch. in Chemical 
 Modeling in Aqueous Systems, ed. by E. A. Jenne. ACS Symp. Ser. 
 93, 1979, pp. 857-892. 
 
 2. Nordstrom, D. K., and J. W. Ball. Chemical Models, Computer 
 Programs and Metal Complexation in Natural Waters. Ch. in Com- 
 plexation of Trace Metals in Natural Waters, ed. by C. J. M. Kramer 
 and J. C. Duinker. Nijhoff (The Hague)-Junk Publ., 1984, pp. 
 149-164. 
 
 3. Jenne, E. A. Chemical Modeling — Goals, Problems, Ap- 
 proaches and Priorities. Ch. in Chemical Modeling in Aqueous 
 Systems, ed. by E. A. Jenne. ACS Symp. Ser. 93, 1979, pp. 3-21. 
 
 4. Parkhurst, D. L., D. C. Thorstenson, and L. N. Plummer. 
 PHREEQE — A Computer Program for Geochemical Calculations. 
 U.S. Geol. Surv. WRI 82-14, 1982, 29 pp. 
 
 5. Plummer, L. N., D. L. Parkhurst, and D. C. Thorstenson. 
 Development of Reaction Models for Groundwater Systems. Geo- 
 chim. et Cosmochim. Acta, v. 47, 1983, pp. 665-685. 
 
 6. Marozas, D. C. The Effects of Mineral Reactions on Trace 
 Metal Characteristics of Groundwater in Desert Basins of Southern 
 California. Ph.D. Thesis, Univ. AZ, Tucson, AZ, 1987, 116 pp. 
 
 7. Harvie, C. E., and J. H. Weare. The Prediction of Mineral 
 Solubilities in Natural Waters: The Na-K-Mg-Ca-Cl-S0 4 -H 2 System 
 From Zero to High Concentration at 25° C. Geochim. et Cosmochim. 
 Acta, v. 44, 1980, pp. 981-997. 
 
 8. Harvie, C. E., N. Moller, and J. H. Weare. The Prediction of 
 Mineral Solubilities in Natural Waters: Na-K-Mg-Ca-Cl-S0 4 -OH- 
 CO3-CO2-H2O System to High Ionic Strengths at 25° C. Geochim. et 
 Cosmochim. Acta, v. 48, 1984, pp. 723-751. 
 
 9. Cook, S. S., and S. E. Paulson. Leaching Characteristics of 
 Selected Supergene Copper Ores. Soc. Min. Eng. AIME preprint 
 88-195, 1988, 15 pp. 
 
 10. Ball, J. W., D. K. Nordstrom, and E. A. Jenne. Additional and 
 Revised Thermochemical Data and Computer Code for WATEQ2 — A 
 Computerized Chemical Model for Trace and Major Element Specia- 
 tion and Mineral Equilibria of Natural Waters. U.S. Geol. Surv. WRI 
 78-116, 1980, 109 pp. 
 
 11. Stumm, S., and J. J. Morgan. Aquatic Chemistry. Wiley, 1981, 
 780 pp. 
 
58 
 
 MODELING INFILTRATION TO UNDERGOUND MINE WORKINGS 
 DURING BLOCK-CAVE LEACHING 
 
 By Robert D. Schmidt 1 
 
 ABSTRACT 
 
 The U.S. Bureau of Mines has modeled the infiltration of leach solution to an 
 underground mine void during block-cave copper leaching. The underground leachant col- 
 lection area is represented as a distributed hydrologic sink, referred to as an "area sink." The 
 area sink solution, which was developed previously using the analytic element method, is 
 described briefly. The area sink model is calibrated against field data from well "drop" tests 
 and an in-mine flow survey; it is then used to estimate the size and shape of the free saturated 
 plume of leachant that exists above the mine void during leaching. An inverse problem is 
 posed, using area sink and well solutions to estimate the leakage coefficient and permeability 
 of ore material above the mine during ongoing leaching operations. 
 
 INTRODUCTION 
 
 Understanding and controlling the distribution of leach 
 solutions in unsaturated ore material is essential for con- 
 ducting safe and efficient leaching operations above the am- 
 bient water table. The development of a distributed area sink 
 solution for modeling infiltration to underground mine work- 
 ings during block-cave leaching is part of a Bureau of Mines 
 effort to provide the industry with modeling tools that 
 facilitate understanding of in situ leaching hydrology in both 
 saturated and unsaturated flow settings. 
 
 A distributed area sink solution, developed using the 
 method of analytic elements and previously used to model 
 localized aquifer infiltration from surface sources such as 
 ponds, rivers, and lakes, has been adapted for modeling in- 
 filtration to irregularly shaped underground mine workings. 2 
 The solution, which is highly efficient, has been implemented 
 in a Bureau hydrology computer program, MINEFLO. This 
 paper describes the application of the area sink solution to the 
 hydrologic setting of a block-cave leaching operation near San 
 Manuel, AZ. 
 
 METHOD OF BLOCK-CAVE COPPER LEACHING 
 
 Block-cave copper leaching is a method by which copper 
 is leached in situ from low-grade ore that has been at least par- 
 tially rubbled by prior conventional block-cave mining ac- 
 tivities. In one variation, shallow wells are constructed on the 
 surface of a rubbled oxide ore mass and an acidic leach solu- 
 
 tion is injected into the wells. The leach solution flows 
 downward through the rubbled mass until it reaches an ex- 
 isting underground mine drift. The mineral-bearing solution is 
 then collected in the mine workings, pumped to the surface, 
 and processed to extract metal values. 
 
 'Hydrologist, Twin Cities Research Center, U.S. Bureau of Mines, Min- 
 neapolis, MN. 
 
 2 Strack, O. D. L. Groundwater Mechanics. Prentice-Hall, 1987, 643 pp. 
 
59 
 
 Figure 1 is a schematic cross section showing the setting of 
 a block-cave copper-leaching operation near San Manuel, AZ. 
 The entire setting is situated above the local ambient water 
 table. The rubbled zone in figure 1 is low-grade oxide ore that 
 was left behind following conventional block-cave mining of 
 an underlying sulfide ore. The thickness of the rubbled ore 
 mass in figure 1 averages 1,626 ft. Injection wells are drilled 
 from the surface to an average depth of 910 ft. The bottom 200 
 ft of each well is slotted pipe. The underground mine workings 
 in figure 1 consist of an undercut drift at elevation 1,125 ft 
 mean sea level (msl) and a haulage drift 65 ft below it, at eleva- 
 tion 1,060 ft msl. The haulage level and the undercut level are 
 
 connected by vertical transfer raises. The underground work- 
 ings that serve as a collection drift for injected leach solution 
 are on the haulage level. 
 
 The cascade area in figure 1 is a zone of partially rubbled 
 ore, which does not directly overlie the undercut workings. 
 Ore material in the cascade area is characterized by mining- 
 induced fractures or crack lines, which run parallel to the 
 boundary of the undercut workings and dip at an angle of 
 about 50°. There is no evidence that these fractures intercept 
 the haulage workings. Thus, there are almost no permeable 
 connections between the cascade area and the underlying 
 mine. 
 
 Figure 1.— Schematic cross section of block-cave leaching. 
 
60 
 
 AVAILABLE HYDROLOGIC FIELD DATA 
 
 At the time of this study, block-cave leaching operations 
 had been ongoing at this site for about 72 weeks. Field data 
 that describe the hydrologic setting during a 16-week interval 
 of the mining operation, from weeks 26 to 41, consist of the 
 following: (1) a daily history of the total solution injection rate 
 in wells and the total solution recovery rate inside the mine, (2) 
 an underground survey of mine inflow conducted during week 
 41, and (3) the results of steady-state "drop" tests conducted in 
 three wells during week 26. These data are used to calibrate a 
 model of saturated conditions at the site. 
 
 INJECTION AND RECOVERY HISTORY 
 
 The total daily rates of solution injection and recovery 
 during the first 72 weeks of leaching operations are plotted in 
 figure 2. Variation in total injection rate during this period was 
 due mainly to variation in the number of operating wells. The 
 operational life span of injection wells averaged about 24 
 weeks. Well losses occur as a result of subsidence and plug- 
 ging. 
 
 Fourteen wells were in operation in the interval from 
 weeks 26 to 41. As figure 2 indicates, the injection and 
 recovery rates during this period were relatively stable; injec- 
 tion averaged 0.85 million gal/d, and recovery averaged 0.4 
 million gal/d. During this period, about 0.45 million gal/d of 
 injected solution was going into phreatic (free-surface) 
 storage; thus, the saturated plume was expanding. 
 
 WELL DROP TESTS 
 
 Also shown in figure 3 are the locations of three wells in 
 which drop tests and permeability slug tests were conducted. 
 During these tests, wells 38, 42, and 45 were shut down one at 
 a time, so that 13 injection wells remained in operation. 3 The 
 decline in solution level was monitored in each well for about 
 30 h. Steady-state levels were reached in all three wells within 
 about 2 h after the start of the tests. Table 1 gives the steady- 
 state solution level and a local permeability estimate (from slug 
 tests) for each well. The steady-state elevations in table 1 are 
 evidence that a stable piezometric surface exists from 150 to 
 650 ft below the rubbled surface, in the immediate vicinity of 
 operating injection wells. 
 
 'Erskine, C. Hydrologic Drop Test In Situ Leach Operations. Magma Mining 
 Co. internal rep., Mar. 1987, 5 pp.; for inf., contact C. Erskine, Pinto Valley 
 Mining Co., Miami, AZ. 
 
 Table 1.— Drop test results in three wells 
 
 Time to reach 
 
 steady state, 
 
 min 
 
 110 
 
 50 
 
 130 
 
 'Above the undercut level. 
 
 2 From slug test conducted in a nearby well. 
 
 
 Well 
 
 Time to reach 
 
 steady state, 
 
 min 
 
 Steady-state 
 
 elevation 
 
 ft msl 
 
 Effective 
 
 permeability 
 
 ft/s 
 
 Saturated 
 
 thickness, 1 
 
 ft 
 
 38 
 42 
 
 45 
 
 
 110 
 
 50 
 
 130 
 
 2,590 
 2,153 
 2,575 
 
 2 6 x 10" 7 
 
 6 x 10~ 5 
 
 7 x 10- 6 
 
 1,465 
 1,028 
 1,450 
 
 MINE INFILTRATION SURVEY 
 
 An in-mine survey to locate the transfer raises where leach 
 solution was either flowing or dripping into the mine was con- 
 ducted on the haulage level during week 41. Figure 3 is a plan 
 view of the mine workings, showing the results of this survey. 
 The outlines of the undercut workings and the haulage work- 
 ings are shown in this figure in relation to the 14 overlying in- 
 jection wells. The locations of 94 flowing and dripping transfer 
 raises are represented on this figure as black rectangles. Dry 
 transfer raises are not marked. The surveyed area of inflow in 
 figure 3 is called the in-mine collection area. Because of the 
 low permeability of undisturbed ore, there was no measurable 
 infiltration to haulage workings located beneath the cascade 
 area, northwest of the undercut drift in figure 3. 
 
 The steady infiltration of leach solution observed in the 94 
 transfer raises indicates that saturated conditions had 
 developed in a substantial portion of the overlying rubbled 
 material owing to 41 weeks of continuous solution injection. 
 Further, the locations of flowing and dripping transfer raises 
 south and east of the injection wells in figure 3 suggest the ap- 
 proximate location of the saturated boundary as it exists above 
 the undercut workings, during week 41. 
 
 KEY 
 
 Injection 
 Recovery 
 
 i i i i i i i i 
 
 10 20 30 40 50 80 70 80 
 
 WEEK 
 
 Figure 2.— Injection and recovery history through week 72. 
 
61 
 
 800 
 
 600- 
 
 LU 
 
 U 
 
 < 
 
 r- 
 
 — 
 a 
 
 400- 
 
 200- 
 
 LEGEND 
 
 Injection well collar 
 
 h Dripping transfer raise 
 
 1 Flowing transfer raise 
 
 ii Haulage drift 
 II (elev 1,064 ft) 
 
 I — | Undercut drift 
 1 — ' (elev 1,124 ft) 
 
 t r 
 
 200 400 600 
 
 DISTANCE, ft 
 
 800 
 
 Figure 3.— Plan-view injection well locations and collection area. 
 
 MODELING MINE INFILTRATION USING DISTRIBUTED AREA SINKS 
 
 In two-dimensional flow problems, an area sink is a 
 distribution of sinks over an area. Steady-state infiltration 
 from a saturated aquifer into an overlying or underlying 
 underground mine void can be modeled, in plan view, by one 
 or more distributed area sinks, provided the size and shape of 
 the area sinks can be chosen arbitrarily, to conform to the 
 distribution of infiltration into the mine. In such a model, the 
 rate of infiltration per unit area of mine workings, y, is given 
 by 
 
 y=*^®-, (1) 
 
 where 0* - </>(z) is the difference in piezometric elevation be- 
 tween the mine void and the overlying (or underlying) aquifer 
 at a location z; z is a complex variable used to represent a two- 
 dimensional plan view coordinate location; and c is the 
 resistance to vertical flow through a layer of saturated material 
 located above (or below) the mine. By application of Darcy's 
 law in equation 1, 
 
 c = -j- 
 
 (2) 
 
 where h is the thickness of the resistance layer and k is its ver- 
 tical permeability. The use of a resistance parameter in equa- 
 tion 1 to describe both the permeability and the thickness of a 
 saturated layer implies that flow through the resistance layer is 
 assumed to be one-dimensional. 
 
 A distributed area sink solution that satisfies equation 1 at 
 every point z in an aquifer above (or below) a mine void can be 
 obtained by integrating the solution for a point sink over an 
 area. This integration is cumbersome, however, and may be 
 carried out analytically for only a few, relatively simple 
 shapes. Alternatively, a solution developed using the method 
 of analytic elements can be used to approximate the distribu- 
 tion of piezometric head, <j>(z), at every point z in the aquifer. 
 The advantage of the analytic element methods is in the ease 
 with which an approximate solution can be obtained for flow 
 problems involving arbitrarily configured mine voids. 
 
62 
 
 DISTRIBUTED AREA SINK SOLUTION USING ANALYTIC ELEMENTS 
 
 A full derivation of the distributed area sink solution 
 using the method of analytic elements, together with a discus- 
 sion of its use in representing infiltration sources such as 
 ponds, rivers, and lakes, is given by Strack. 4 For com- 
 pleteness, a brief outline of the method is presented here, with 
 particular reference to conditions of underground mine in- 
 filtration. 
 
 Making use of the Dupuit-Forchheimer assumption, an 
 area sink potential, <£ = <t>(0,h,k), can be defined as a func- 
 tion of piezometric head, <$> = </>(z), saturated thickness, h, 
 and permeability, k, that is continuous in both confined and 
 unconfined mine inflow settings. 
 
 Considering a contour C enclosing a domain D + to be 
 the arbitrary plan-view boundary of an underground mine 
 void and using 7 to denote a constant distribution of inflow to 
 (or outflow from) the mine, then following Strack, an area 
 sink solution for representing the mine void consists of deter- 
 mining an area sink potential, <i> = $(z), at a location z, which 
 fulfills the differential equation 
 
 V 2 <J> 
 
 for any z in D + , 
 
 and the differential equation 
 
 = 7 
 
 V 2 <I> = 
 
 (3) 
 
 (4) 
 
 for any z in D", 
 
 where D" is outside of C. 
 
 The problem is solved by splitting the potential <I> into two 
 parts: 
 
 d> = <t>i + <i>% 
 
 (5) 
 
 where 4>' is any function that satisfies V 2 <i>' = 7 for z in D*, and 
 where <£' = in D". By itself, <£' is discontinuous across the 
 boundary C and therefore violates the conditions for continui- 
 ty of both pressure and flow; 4> e is used to eliminate these 
 discontinuities. In order for the potential <J> (and therefore the 
 pressure head 0) to be continuous on the boundary, <i> e must 
 jump by an amount 
 
 <|>e-_<i>e += _$i ( 6 ) 
 
 across the boundary, where the superscript ( + ) and ( - ) refer 
 here to the value of $ e inside and outside of C, respectively. 
 Additionally, if Q' represents the component of discharge due 
 to $' that is normal to the mine boundary and Q e represents 
 the normal component due to <i> e , then in order for flow to be 
 continuous, Q e also must jump by an amount 
 
 Q'+-Qe-= -Q< 
 
 (7) 
 
 across the mine boundary. 
 
 A line integral that is a distribution of dipoles oriented 
 normal to a line element is referred to as a "line doublet" ele- 
 ment. The line doublet has the property that the potential $ is 
 
 discontinuous across the element. Line doublets are placed 
 along the boundary C in order to generate the necessary 
 discontinuity in potential <t> e . In a similar fashion, the 
 necessary discontinuity in the normal discharge vector, Q e , is 
 produced by placing line sink elements (line integrals that are 
 distributions of sinks) along the boundary C. 
 
 Thus the function 3> e , which can also be interpreted as the 
 real part of a complex potential fl e = <£ e 4- i* e , may be written 
 as a complex contour integral consisting of doublet and line 
 sink elements. The function ft e is the complex potential that is 
 valid at z locations outside the area sink; the potential valid at 
 z locations inside the area sink is obtained by adding <$' to the 
 real part of ft e . 
 
 The complex potential function ft e is given as 
 
 ft e = 
 
 n(6) d<5 
 
 2tt 
 
 In(z-Zi), 
 
 (8) 
 
 where the first term is the complex potential for a line doublet, 
 the second term is the complex potential for a line dipole, and 
 the second and third terms together constitute the potential for 
 a line sink. The condition of continuity of <t> and ¥ along C is 
 expressed in equation 8 in terms of density distributions for 
 line doublets, X, and line dipoles, /i, respectively. The total 
 discharge from the area sink, which is not necessarily known in 
 advance is Qo. The integration in equation 8 is performed 
 analytically on an area sink boundary C that has been 
 discretized as a polygon. The starting point of integration on 
 the boundary is zi. 
 
 For a constant infiltration rate 7, the potential function $' 
 is given as 
 
 $' = l/2 7 £ 2 
 
 (9) 
 
 -Work cited in footnote 2. 
 
 where £ is a unit length along the boundary of the area sink. 
 
 An area sink may be infiltration-rate specified or head 
 specified. In cases were the area sink infiltration rate is known, 
 the complex potential (ft, where ft = ft' inside an area sink and 
 ft = ft e outside an area sink) can be calculated directly. In 
 cases where area sink head <f>* and resistance c are known, 
 equation 1 is used to determine the infiltration rate per unit 
 area 7. 
 
 The system of equations that results when multiple area 
 sink solutions are superimposed, or when area sink solutions 
 are superimposed on other analytic element solutions, such as 
 wells, is solved directly in MINEFLO using Gauss elimination. 
 In an unconfined flow setting, the aquifer thickness is itself the 
 piezometric head; thus, the function relating potential, <£, to 
 piezometric head, <j>, is nonlinear. In this case, the infiltration 
 rate for an area sink that is head and resistance specified is 
 calculated iteratively by MINEFLO. 
 
63 
 
 DROP TEST SIMULATION 
 
 Modeling the distribution of saturated conditions during 
 weeks 26 to 41 of block-cave leaching operations consists, 
 essentially, of superimposing 3 distributed area sink solutions 
 and 15 steady-state well (point source) solutions in order to 
 simulate the drop tests conducted in wells 38, 42, and 45. In 
 the simulation, just as in the field test, 13 of 14 injection wells 
 are specified on the basis of a known fixed head (2,850 ft msl), 
 and a condition of zero injection is individually specified for 
 each of wells 38, 42, and 45. 
 
 REPRESENTING THE DROP TEST SETTING 
 
 Figures 4 and 5 describe the setting of simulated drop tests 
 in plan view and in cross section. The plan-view boundaries of 
 the three area sinks used to represent collections of dripping 
 and flowing transfer raises during drop tests are shown in 
 
 figure 4. The area sinks are at the undercut level (1,125 ft msl). 
 Their boundaries are chosen so that each sink underlies one of 
 the three wells (38, 42, and 45). Thus, the three area sinks are 
 referred to by these numbers also. 
 
 Figure 5 is a schematic cross section showing components 
 of an area sink used to simulate the well 45 drop test. The zone 
 of saturated rubbled material above the area sink in figure 5 is 
 assumed to be composed of a resistance layer and a source 
 aquifer. Flow within the resistance layer, which is a 716-ft 
 thickness of rubbled material located between the bottom of 
 the injection well slotted interval (1,840 ft msl) and the under- 
 cut level (1,125 ft msl), is assumed to be downward only. The 
 source aquifer in figure 5 is interpreted as a saturated zone 
 located between the top of the resistance layer (1,840 ft msl) 
 and the local piezometric surface (2,575 ft msl at well 45). Flow 
 in the source aquifer is modeled as two-dimensional and un- 
 con fined. 
 
 LU 
 
 u 
 < 
 
 r- 
 00 
 
 Q 
 
 200 
 
 ouun 
 
 
 
 600- 
 
 Cascade 
 
 area \\ \\ \ ^^^\f^ 
 
 400- 
 
 
 V^ 38 \\ \\ % \ 
 
 200- 
 
 N V 1 
 
 1 x 
 
 
 n 
 
 
 
 u 
 
 1 
 
 i il 
 
 LEGEND 
 
 Injection well collar 
 
 Dripping transfer raise 
 
 Flowing transfer raise 
 
 Haulage drift 
 (elev 1.064 ft) 
 
 Undercut drift 
 (elev 1.124 ft) 
 
 Distributed area sinks 
 at undercut level: 
 
 Area sink 38 
 
 Area sink 45 
 
 Area sink 42 
 
 □ 
 
 
 400 
 DISTANCE, ft 
 
 600 
 
 800 
 
 Figure 4.— Area sink representation of mine infiltration. 
 
64 
 
 Figure 5.— Schematic cross section of distributed area sink. 
 
 AREA SINK CALIBRATION 
 
 Calibration consists of determining the infiltration rate 
 for each area sink, such that the calculated heads at the loca- 
 tions of wells 38, 42, and 45 match the measured heads given in 
 table 1. 
 
 Rearranging equation 1 to incorporate the definition of 
 the resistance parameter c given in equation 2 and writing the 
 infiltration rate per unit area y as Q c /A c , where A c is the plan- 
 view area of a distributed area sink, gives 
 
 Qc h 
 
 0(Z) = 
 
 (10) 
 
 It follows that for area sinks that are head and resistance 
 specified, a calibrated distribution of infiltration is one in 
 which the three Q c values chosen (one for each area sink) 
 satisfy equation 10 at the locations of the three test wells. 
 Because of the uncertainty associated with the permeability 
 values obtained from slug tests, a limited amount of manipula- 
 tion of resistance values c is permitted during MINEFLO 
 calibration runs. By contrast, the value of 0* - 0(z) at the 
 location of the test wells is known with much greater certainty. 
 The calibration also takes advantage of the fact that the total 
 well injection rate and the total mine outflow rate are known 
 precisely during the drop tests. 
 
 In equation 10, estimates of resistance layer permeability, k, 
 and thickness, h, are available from the drop test results in 
 table 1 and from the definitions in figure 5. The plan view area 
 A c for each of the three area sinks is laid out in figure 4, and <£* 
 is simply the elevation of the undercut level. The difference in 
 piezometric elevation, </>*-0(z), between the source aquifer 
 and the area sink is available at just one location z above each 
 area sink, i.e., at the locations of wells 38, 42, and 45 (see table 
 1). 
 
 DISTRIBUTION OF SATURATED CONDITIONS 
 
 The results of calibration are presented in tables 2 and 3. 
 The actual and simulated </>* - <t>(z) values at the locations, z, 
 of wells 38, 42, and 45 are given in table 2. The discrepancy be- 
 tween the two sets of values is small, relative to the total 
 thickness of saturated material, and thus equation 8 is con- 
 sidered to have been satisfied. In addition, table 2 shows the 
 total well injection rate, actual and simulated. Table 3 gives 
 
65 
 
 the proportional distribution of (total) mine infiltration, i.e., 
 367,500 gal/d, used to obtain these results. In the calibrated 
 model, area sinks 42 and 45 account for about 95 pet of total 
 solution recovery, while area sink 38 accounts for only 5 pet. 
 The distribution is consistent with other estimates of plume 
 size and shape 5 and with the in-mine survey results of figure 3, 
 which show a dense cluster of flowing transfer raises in the 
 vicinity of wells 42 and 45 and mainly dripping transfer raises 
 spread out in the vicinity of well 38. 
 
 In comparing tables 2 and 3, it is seen that a high infiltra- 
 tion rate per unit area of mine workings is associated with an 
 overlying test well that has a low-steady-state piezometric 
 head; compare, for instance, well 42 and area sink 42. Con- 
 versely, a low mine-infiltration rate is associated with a test 
 well that has a high piezometric head; compare well 38 and 
 area sink 38. This apparent contradiction between piezometric 
 condition and mine infiltration is resolved when it is recalled 
 that the operational injection wells are head specified in 
 MINEFLO, not flow specified. Thus, given the same concen- 
 tration of injection wells and the same injection well head 
 specification (2,850 ft msl), the observation of a low head in a 
 test well above the mine implies the existence of greater 
 permeability in the vicinity of the well and therefore lower 
 resistance to the downward flow of leach solution. The obser- 
 vation of a high head in a test well implies lower permeability 
 and greater resistance to downward flow. 
 
 In the case of well and area sink 45, a well head that is 
 comparable to well 38 is associated with an area sink infiltra- 
 
 'Swineford, G. Volume of Ore Wetted by In Situ Leaching. Magma Mining 
 Co. internal rep., July 1987, 16 pp.; for inf., contact M. Miller, Magma Mining 
 Co., San Manuel, AZ. 
 
 tion rate that is about half that of area sink 42. The elevated 
 head in well 45 is due to a greater concentration of head- 
 specified injection wells (7 of 14 injection wells overlie area 
 sink 45). As with all three area sinks, the infiltration rate per 
 unit area of area sink 45 is proportional to the overlying 
 permeability (see table 1). 
 
 The saturated plume that surrounds the injection wells in 
 the rubbled ore mass and in the cascade area can be described 
 using the distribution of mine infiltration in table 3. The 
 simulated plume is shown in figure 6. The plume boundary is a 
 
 Table 2.— Area sink calibration results, week 41 
 
 Condition Simulated 
 
 Piezometric elevation above the 
 undercut level, ft: 
 
 Well 38 1,440 
 
 Well 42 1,004 
 
 Well 45 1,520 
 
 Total injection rate from all 
 
 wells gal/d.. 876,000 
 
 Actual 
 
 1,465 
 1,028 
 1,450 
 
 875,000 
 
 Table 3.— Area sink outflow distribution 
 
 Area sink Area, Outflow rate, Total outflow 
 
 location' ft* (gal/d)ft 2 rate, gal/d 
 
 38 54,137 0.36 19,438 
 
 42 21,872 7.72 168,849 
 
 45 49,912 3.59 179,216 
 
 Total simu- 
 
 lated outflow 367,503 
 
 'In reference to the overlying test well. 
 
 Percent of 
 total 
 
 5 
 46 
 49 
 
 100 
 
 2,000 
 
 1,500 
 
 LU 
 
 CJ 
 
 < 
 
 r- 
 oo 
 
 Q 
 
 1,000 
 
 500- 
 
 ( 
 
 
 
 LEGEND 
 
 
 Steady-state well 
 
 ( 
 
 Saturated zone 
 
 piezometric contours 
 
 
 (interval = 100 ft) 
 
 m 
 
 In place ore 
 
 o 
 
 Rubbled material 
 
 
 Distributed area sinks 
 
 
 at undercut level: 
 
 LJ 
 
 Area sink 38 
 
 ■1 
 
 Area sink 45 
 
 k. 
 
 Area sink 42 
 
 500 1,000 1,500 
 
 DISTANCE, ft 
 
 Figure 6.— Simulated solution plume, week 41. 
 
66 
 
 zero-head condition in the source aquifer; in the resistance 
 layer, it is a vertical free surface extending from the bottom of 
 the source aquifer (1,840 ft msl) to the undercut level (1,125 ft 
 msl). It is not modeled as a zero flux condition, since figure 2 
 indicates that the plume boundary is expanding during almost 
 
 all of the first 41 weeks of leaching operations. The presence of 
 a large portion of the plume in the cascade area results from 
 the asssumption that there is no permeable connection be- 
 tween the cascade area and the underlying haulage drifts. 
 
 PERMEABILITY ESTIMATION USING AREA SINKS 
 
 Consider the inverse of the previous calibration problem. 
 The ore permeability is not known in advance, but the 
 distribution of infiltration within the mine has been deter- 
 mined in an in-mine survey, which precisely measured the 
 discharge from individual sump areas on the haulage level or 
 possibly even from individual transfer raises. Rearranging 
 terms in equation 10, the unknown permeability, k, can be 
 written as 
 
 _Q^ 
 
 A c (0*-<Kz)). 
 
 (11) 
 
 Equation 11 is identical to the expression used for 
 estimating aquifer recharge due to vertical leakage through a 
 
 deposit, 6 where k/h (the inverse of resistance) is called the 
 leakage coefficient, and A c is the area over which vertical 
 leakage is diverted to a pumping center. 
 
 If, in addition to the above in-mine survey, steady-state 
 drop tests are conducted at well locations carefully chosen to 
 provide representative measures of the head distribution in the 
 overlying solution plume, then the MINEFLO computer pro- 
 gram and the distributed area sink solution can be applied to 
 determine a leakage coefficient k/h and a vertical permeability 
 k, which satisfies equation 11. The method is appropriate for 
 an ongoing block-cave leach setting in which it can be assumed 
 that [<t>* - 0(z)]/h measures the average vertical gradient in 
 head above the area sink and that <j>* - <j>(z) and Q c /A c are 
 constant with respect to time. 
 
 SUMMARY AND CONCLUSIONS 
 
 Distributed area sink solutions developed using the 
 analytic element method have been shown to be useful for 
 modeling irregular underground mine infiltration areas. 
 Together with superimposed analytic element solutions for 
 wells and other flow features, area sinks can be used to 
 describe the saturated plume that develops above an 
 underground mine collection area during block-cave leaching. 
 
 The area sink model aids in describing two important 
 hydrologic variables that influence the size and shape of the 
 saturated plume: the placement of injection wells above the 
 underground solution-collection area and the occurrence of 
 permeability layering in the rubbled material overlying the col- 
 lection area. 
 
 The area sink and well solutions can be applied via an in- 
 verse problem to estimate the leakage coefficient and 
 
 permeability of saturated material overlying the in-mine collec- 
 tion area. For instance, the entire in-mine collection area can 
 be represented by a single distributed area sink, a single test 
 well can be used to estimate the average head in the saturated 
 zone above the area sink, and then a single average leakage 
 coefficient and permeability can be calculated for the entire 
 saturated, rubbled zone. Alternatively, the in-mine collection 
 area can be divided into several area sinks based on differences 
 in infiltration rate. Steady-state head measurements can be ob- 
 tained at one or more locations in the saturated material above 
 each area sink, and the permeability of each portion can be 
 estimated individually. 
 
 'Walton, W. C. Groundwater Resource Evaluation. McGraw-Hill, 1970, 664 
 
 pp. 
 
67 
 
 HYDROLOGIC: AN INTELLIGENT INTERFACE 
 FOR THE MINEFLO HYDROLOGY MODEL 
 
 By Michael E. Salovich 1 
 
 ABSTRACT 
 
 The U.S. Bureau of Mines has developed an expert system to provide assistance to 
 researchers using the Bureau's hydrology computer program MINEFLO. This expert system, 
 HydroLogic, is a multifaceted computer program that addresses the various ways in which 
 MINEFLO may be used, including the design and analysis of in situ hydrologic operations. 
 With tutorials, templates, and diagnostic questions, HydroLogic combines educational 
 resources with diagnostic reasoning to create an environment supporting the MINEFLO 
 hydrology program. This added support reduces the level of technical difficulty in using 
 MINEFLO and thereby opens the model to a wider range of possible users. As a conse- 
 quence, MINEFLO's effectiveness as a hydrologic tool for the mining industry is enhanced. 
 
 INTRODUCTION 
 
 The design and analysis of in situ systems require an 
 understanding of the various factors that interact and define 
 the hydrology of the system. This is a complicated task 
 because, over time, hydrologic parameters, such as aquifer 
 permeability, solution injection rates, well locations, and solu- 
 tion recovery rates, interact and change. An impermeable layer 
 may form, the leachant solution may alter the permeability of 
 the constituent ore deposit, or the occurrence of fracture fault 
 lines may increase as subsidence occurs during block-cave 
 leaching. These changes, in turn, call for a response in mining 
 operations. Injection rates may have to be raised, wells may 
 have to be repositioned and drilled deeper, the chemistry of the 
 leachant solution may have to be remixed, or the in situ opera- 
 tion itself may have to be reevaluated and discontinued. In 
 other words, this type of analysis of in situ systems can be ap- 
 proached as a study of parameters — hydrologic parameters 
 that are identified and manipulated so that their interaction 
 and effects are understood. In situ mining is a dynamic proc- 
 ess. Understanding these factors and being able to anticipate 
 their interaction is vital for efficient in situ operation. In light 
 of this, the Bureau of Mines has developed two computer pro- 
 grams that work together in the analysis of these hydrologic 
 factors: MINEFLO and HydroLogic. These programs have 
 
 'Computer programmer analyst, Twin Cities Research Center, U.S. Bureau of 
 Mines, Minneapolis, MN. 
 
 been developed to assist the mining industry in the planning 
 and development of in situ mining operations, as part of the 
 Bureau's effort to increase minerals extraction efficiency. 
 
 MINEFLO is a hydrologic simulator that provides de- 
 tailed and general analyses of hydrologic systems. It can pro- 
 vide information about aquifer permeability, well discharge 
 and collection rates, leachant streamline direction and veloci- 
 ty, changing head pressures, and the effects of fractures and 
 impermeable zones. Although this hydrologic simulator is 
 robust and versatile, prior understanding of the hydrologic 
 terms and methods it employs is required. With the goal of 
 making MINEFLO a useful tool for the mining industry, the 
 Bureau developed HydroLogic — a companion computer pro- 
 gram designed to help researchers use MINEFLO. 
 
 HydroLogic, the focus of this paper, is a multifaceted 
 computer program that addresses the various ways in which 
 MINEFLO can be used. With diagnostic questions, tutorials, 
 and templates, HydroLogic combines educational resources 
 with diagnostic reasoning to create an "expert support environ- 
 ment" for the MINEFLO hydrology model. In other words, 
 HydroLogic functions as an intelligent interface for 
 MINEFLO. In the sense that expert systems are computer pro- 
 grams that attempt to emulate the thoughts and actions of an 
 expert in a particular field of knowledge, HydroLogic is an ex- 
 pert system that provides expert guidance to those using 
 MINEFLO. 
 
68 
 
 ENVIRONMENT 
 
 HydroLogic was developed in HyperTalk, the script 
 language of HyperCard, for use on Apple Macintosh com- 
 puters. 2 HyperCard provides an excellent environment for the 
 implementation of HydroLogic because it offers "hyper- 
 media" capabilities (i.e., interactive graphics combined with 
 text) and the flexibility to be linked to other computer pro- 
 grams and data bases. HyperCard's advanced graphics tools 
 enable the HydroLogic expert system to present different com- 
 binations of pictures, diagrams, and text with "cursor 
 sensitive" areas for simple communication. Figure 1 shows the 
 main menu of HydroLogic. Each rectangle in this picture is 
 designed to be a "graphic button," sensitive to the cursor. 
 When a button is selected, the system interprets it as a com- 
 mand and responds accordingly. With this method, users can 
 navigate HydroLogic's menus efficiently by communicating 
 commands on a graphic level. 
 
 HyperCard also has the ability to act as a central control 
 point for accessing other computer files and applications. 
 With its "open" command, it can initiate other computer pro- 
 
 reference to specific products does not imply endorsement by the Bureau of 
 Mines. 
 
 grams, and with its "read" and "write" commands, it can ac- 
 cess and manipulate any other computer files residing in 
 memory. These commands allow HydroLogic to read and 
 manipulate MINEFLO's data sets or to start the program 
 itself. This creates a workstation environment that eliminates 
 the awkwardness of having to switch back and forth between 
 the consulting program and the intended application. 
 
 Unlike many computer programs, HydroLogic is not an 
 isolated system; it provides a dynamic environment that can 
 interact with any resource on the computer. This is important 
 because HyperTalk's control structures are limited in creating 
 systems that involve complex recursion or symbol manipula- 
 tion. Future development of HydroLogic, then, may be ex- 
 tended to include additional artificial intelligence languages or 
 systems, such as PROLOG or LISP. 
 
 Versions of the MINEFLO hydrology model exist on the 
 MicroVax, Harris, and Macintosh computers. At this time, 
 versions of the Hydrologic system operate only on the Macin- 
 tosh II, Macintosh SE, and Macintosh Plus. Because of the 
 size of the HyperCard code and the possibility of creating large 
 data sets with MINEFLO, a hard disk is necessary for op- 
 timum performance. 
 
 rfliihnrihiiiflMniin 
 
 HydroLogic 
 
 1:57 | l 
 
 Quit I 
 
 ^jj|||w 
 
 
 Mineflo 
 
 
 Tutorials 
 
 Templates 
 
 Diagnostics 
 
 
 mmmmmmmmm 
 
 Figure 1.— HydroLogic's main menu, composed of graphic buttons. 
 
69 
 
 DESIGN 
 
 The HydroLogic system is composed of three major sec- 
 tions: Diagnostics, Tutorials, and Templates. Only the 
 Diagnostics section employs the diagnostic queries familiar in 
 most expert systems. The Tutorials and Templates sections are 
 educational resources that provide important information 
 about the methodology and terminology used by the 
 MINEFLO program. This combination of educational 
 resources with diagnostic analysis enables HydroLogic to pro- 
 vide a broad base of support to those using the MINEFLO 
 hydrology model. 
 
 DIAGNOSTICS 
 
 Most diagnostic systems require a small problem domain 
 because of the way they capture and apply knowledge. This is 
 done commonly through production rules — simple "IF . . . 
 THEN" statements. These rules are carefully designed and 
 structured so that they recreate the various steps of reasoning 
 an expert may take in reaching a decision. Since all the mental 
 steps involved in analyzing a problem must be explicitly 
 represented in production rules, a problem must be sufficiently 
 limited in its complexity to be addressed effectively by a 
 diagnostic system. Otherwise, the diagnostic system would be 
 too complex to design properly. 
 
 Additionally, diagnostic systems defined by production 
 rules tend to be stiff and brittle. This is because production 
 rules are predefined and their coherency can easily break down 
 in the face of unforeseen situations. Since most of the produc- 
 tion rules are interrelated, it is very difficult to make even the 
 smallest changes to a system without a total reconfiguration. 
 Therefore, the problem domain addressed by a diagnostic 
 system should be sufficiently limited in complexity for reasons 
 of effectiveness and efficiency. 
 
 The task of HydroLogic, as stated before, is to provide 
 assistance to people using MINEFLO. The ideal expert, the 
 model for the system, is someone who knows how to operate 
 and apply MINEFLO in various types of situations. This, 
 however, is a formidable task. The different combinations of 
 MINEFLO input errors, user questions, and modeling prob- 
 lems are numerous and create a large problem space. This task 
 is too large and complex to be accurately and efficiently 
 represented by a single structure of production rules. 
 However, subsets of this problem domain (i.e., specific topics 
 concerning input parameters, error diagnosis of streamline 
 generation, etc.) are less complex and therefore are viable sub- 
 jects for smaller diagnostic systems. The large problem domain 
 confronting HydroLogic could therefore be reduced to a col- 
 lection of smaller and relatively simple subset units. 
 
 This translation of a large problem space into a collection 
 of smaller units was the key for the development of 
 HydroLogic. This method of "distributed problem spaces," 
 developed during the research, opened the way for creating ex- 
 pert systems for large or inexact problem domains. It is based 
 upon the observation that a large or partially defined problem 
 space can be mapped to a collection of smaller well-defined 
 subsets. These smaller subset units are logically independent 
 and highly specialized and therefore can be easily represented 
 through production rules. Thus, an expert system, composed 
 of these diagnostic subsets, can be created for a large or par- 
 tially defined problem space. 
 
 HydroLogic, then, can be viewed as a collection of small 
 diagnostic systems. These subsets are highly specialized, and 
 some border on being trivial, but they are related under the 
 common theme of being a resource for the MINEFLO 
 
 hydrology model. Figure 2 shows how the HydroLogic system 
 enables users to select one of the diagnostic subtopics for con- 
 sultation. Figure 3 shows a partial decision tree, depicting the 
 typical scope and complexity of a diagnostic subtopic. These 
 diagnostic subtopics use forward chaining to proceed from an 
 initial (output problem) state to a goal (input revision) state. 
 The initial problem state shown in figure 3 is the condition 
 where the user of MINEFLO cannot start MINEFLO's flow 
 streamlines. The rules in this case consist of all the required in- 
 put specifications for generating streamlines using MINEFLO. 
 The goal (revision) state is a condition in which the input error 
 that prevented streamlines from starting is identified for the 
 user. 
 
 If the logical structure formed by a set of production rules 
 can be viewed as a decision tree, these subsets can be con- 
 sidered branches of a larger or, perhaps, partially defined tree. 
 Therefore, the method of distributed problem spaces is a 
 means of adding modularity to the design of a diagnostic 
 system. An expert system, then, designed in this manner en- 
 joys the same benefits as other modular programs: ease of 
 design, maintenance, and modification. Small subsets are less 
 complex and therefore are easier to create and manage. 
 Together the subsets can create an extremely robust and flexi- 
 ble system. In addition, because the main problem is expressed 
 in subtopics, users enjoy the benefit of having greater control 
 over the system. One of the difficulties with expert systems is 
 that they tend to interrogate users with a litany of seemingly 
 trivial or inappropriate questions. With a modular program, 
 users may select the topic of inquiry and thereby constrain the 
 flow of questions. 
 
 As stated before, HydroLogic incorporates educational 
 resources with the Diagnostics section. This is done because 
 certain types of information are more easily and effectively 
 communicated through references rather than through 
 diagnostic analysis. For example, if a researcher needed to 
 locate a menu of commands inside the MINEFLO program, it 
 is apparent that viewing a structured diagram of MINEFLO's 
 menu hierarchy would be more efficient than submitting to a 
 diagnostic question-answer session. Therefore, for reasons of 
 efficiency, the HydroLogic system was designed to include, in 
 addition to its Diagnostics section, sections containing educa- 
 tional tutorials and templates. 
 
 TUTORIALS 
 
 HydroLogic's Tutorials section is a collection of interac- 
 tive help files containing explanations and definitions to assist 
 MINEFLO users with basic definitions and terminology 
 associated with in situ hydrology. For example, tutorials have 
 been developed to describe the analytic element methodology 
 of the MINEFLO program. Specific tutorials exist for each of 
 the analytic element solutions used in the program for model- 
 ing hydrologic features such as wells, permeability zones, bar- 
 riers, fractures, etc. They contain information ranging from 
 MINEFLO's mathematics to the various steps needed for the 
 creation of a three-dimensional profile of an aquifer. The 
 tutorials are interactive and easy to use, employing graphic 
 buttons and other hypermedia features. 
 
 TEMPLATES 
 
 HydroLogic's Templates section contains well-defined ex- 
 amples of MINEFLO data sets. (MINEFLO uses data sets as a 
 permanent means of recording and storing information about 
 
70 
 
 Diagnostics 
 
 HydroLogic 
 
 CCQ0@[^[J)QQ(BDQ fflDB" [iQD[p n 
 
 discharge rates are too high 
 
 O 
 
 discharge rates are too loin 
 
 
 head ualues are too high 
 
 
 head ualues are too low 
 
 
 streamlines are jagged 
 
 
 streamlines stop prematurely 
 
 
 streamlines don't start 
 
 
 streamlines are uery slow 
 
 
 streamlines stop because head is zero 
 
 O 
 
 Figure 2.— HydroLogic's diagnostic subsets displayed in scroll box. 
 
 Rre you starting 
 streamlines interactiuely? 
 
 Ves 
 
 flduice No. 1 
 Continue? 
 
 No 
 
 No 
 
 Return 
 
 Ves 
 
 Haue you specified 
 any starting points? 
 
 v No 
 flduice No. 2, Return 
 
 Do you know the streamline 
 termination condition? 
 
 Ves 
 
 No 
 
 Figure 3.— Partial decision tree for diagnostic subtopic 
 'streamlines don't start." 
 
 a particular hydrologic setting.) Each template is designed to 
 convey important information about how to apply MINEFLO 
 to a particular type of problem. Essentially, the templates are a 
 way of imparting to MINEFLO users the Bureau's experience 
 and expertise in identifying hydrologic concerns that are com- 
 monly associated with in situ mining activity. In this sense, 
 they serve as starting points for researchers creating new data 
 sets for the MINEFLO program. Users profit by running the 
 MINEFLO program using these data sets and by manipulating 
 the various input parameters. As researchers become more 
 familiar with the program, they can build on the data sets and 
 customize them, thus progressively generating their own in- 
 sights and understandings about their particular hydrologic 
 setting. In addition, the templates provide concrete examples 
 of correct and acceptable input parameter definitions for the 
 MINEFLO program. 
 
 Figure 4 shows an example of a template title card. Each 
 card contains a title, a key of important features, and a 
 graphic image generated by the MINEFLO hydrology model. 
 The template shown in figure 4 is a fully defined data set used 
 for modeling a single leachant injection well drilled above an 
 underground mine void. It contains the basic hydrologic 
 features characteristic of a method of in situ leaching called 
 block-cave leaching. The key features of the data set are iden- 
 tified on the template, and users have the option of looking at 
 additional notes, running the MINEFLO program with this 
 data set, or copying and modifying it by adding more wells, 
 changing the shape and location of the underlying mine, or 
 adding other features. 
 
71 
 
 Figure 4.— Example of a template title card. 
 
 DISCUSSION 
 
 The primary purpose of HydroLogic and MINEFLO is to 
 provide the mining industry with an effective tool for the 
 design and analysis of in situ hydrologic systems. The 
 MINEFLO hydrology model offers the capability to analyze 
 the complex interaction of several types of hydrologic 
 elements. HydroLogic functions as a companion program that 
 provides information and support regarding MINEFLO's ter- 
 minology and methodology. The measure of success of 
 HydroLogic is its usefulness in providing support to 
 MINEFLO. Although the basic steps have been taken in 
 establishing diagnostic modules, tutorials, and templates for 
 fundamental topics concerning MINEFLO, it is recognized 
 
 that further additions are necessary for the Hydrologic system 
 to be considered comprehensive. As the Bureau's experience 
 with the MINEFLO hydrology model grows, further develop- 
 ment is planned for the HydroLogic system. It is also recog- 
 nized that there are different types of computers used in the 
 mining industry and that versions of MINEFLO and 
 HydroLogic that operate on these computers should be con- 
 sidered. It is planned that, as other computer systems develop 
 the necessary graphic and hypermedia capabilities required by 
 these programs, versions of MINEFLO and HydroLogic will 
 be developed for these systems. 
 
 SUMMARY AND CONCLUSIONS 
 
 In summary, HydroLogic contains three sections: 
 Diagnostics, Tutorials, and Templates. Together they create a 
 support environment for the MINEFLO hydrology computer 
 program. Although they do not form a complete resource, 
 they provide important information and perspective about 
 MINEFLO's capabilities and operation. With the additional 
 
 support of HydroLogic, the task of learning how to operate 
 MINEFLO is simplified. This added simplification, in turn, 
 translates into greater productivity because it opens 
 MINEFLO to a larger number of possible users and adds to 
 the depth and quality of research performed with the model. 
 
72 
 
 BIBLIOGRAPHY 
 
 Dreyfus, H. L., and S. E. Dreyfus. Mind Over Machine. Free Press, Rich, E. Artificial Intelligence. McGraw-Hill, 1983, 436 pp. 
 
 1986, 231 pp. Sowa, J. F. Conceptual Structures. Addison-Wesley, 1984, 481 pp. 
 
 Nilsson, N. J. Principles of Artificial Intelligence. Morgan Kauf- Waterman, D. A. A Guide to Expert Systems. Addison-Wesley, 
 
 mann, 1980, 476 pp. 1986, 419 pp. 
 
 Reboh, R., J. Reiter, and J. Gaschnig. Development of a Winston, P. H. Artificial Intelligence. Addison-Wesley, 1984, 524 
 
 Knowledge-Based Interface to a Hydrologic Simulation Program. pp. 
 SRI, 1982, 110 pp. 
 
73 
 
 PREDICTING AND MONITORING LEACH SOLUTION FLOW 
 WITH GEOPHYSICAL TECHNIQUES 
 
 By Daryl R. Tweeton, 1 Calvin L. Cumerlato, 2 Jay C. Hanson, 2 and Harland L. Kuhlman 3 
 
 ABSTRACT 
 
 The U.S. Bureau of Mines is conducting research to develop improved methods for 
 predicting and monitoring the flow of leach solution during in situ mining. Potential benefits 
 include more reliable assessment of leaching feasibility, higher metal recovery through better 
 leach solution distribution, and greater confidence by regulatory agencies that leach solution 
 will not escape from the mine. 
 
 The ability of seismic cross-hole tomography to detect fractured zones and saturated 
 areas is being field tested for applications in predicting flow patterns and in monitoring leach 
 solution injected above the water table. A tomographic analysis computer program, BOM- 
 TOM, was written with special options for geophysical applications. It was first used with 
 seismic refraction data for detecting fractured zones in a quarry. Tomographic analysis of 
 seismic cross-hole travel-time data detected a mound in the water table made by injecting 
 water between the source and receiver boreholes. Research to improve the reliability of 
 tomographic reconstructions is continuing. Electromagnetic methods for determining where 
 high-conductivity leach solution has replaced ground water are also being evaluated. 
 Preliminary computer simulations indicate that surface-to-borehole time-domain elec- 
 tromagnetic induction is promising. 
 
 INTRODUCTION 
 
 IMPORTANCE OF PREDICTING AND 
 MONITORING FLOW 
 
 Reliably predicting and monitoring the flow of leach solu- 
 tion is important in planning and operating an in situ mine. 
 Predicting flow patterns is critical in assessing the feasibility of 
 mining; if most of the flow is in a few large fractures, much of 
 the ore may remain unleached. If mining is feasible, predicting 
 and verifying flow patterns can aid in planning well spacings, 
 selecting depths at which to perforate casings, and choosing 
 pumping rates to provide the best leaching coverage of the ore. 
 Monitoring the flow in a pilot-scale study can help in deciding 
 if well spacing should be changed in a commercial operation. 
 Better distribution of leach solution will result in its contacting 
 more of the ore, which will increase recovery of the ore 
 mineral. 
 
 Safeguarding the ground water resources near an in situ 
 leach mining operation is a responsibility shared by the 
 operator and regulatory agencies. Obtaining permits from 
 regulatory agencies is one of the critical steps in starting an in 
 situ mine. Improved methods of predicting and monitoring 
 
 'Research physicist. 
 
 2 Geophysicist. 
 
 'Engineering technician. 
 
 Twin Cities Research Center, U.S. Bureau of Mines, Minneapolis, MN. 
 
 flow could help mining companies control leaching solutions 
 and assure regulatory agencies that these solutions will not 
 escape from the well field. 
 
 ADVANTAGES OF GEOPHYSICAL METHODS 
 
 Bureau of Mines researchers believe that certain cross- 
 hole or surface-to-borehole geophysical techniques could pro- 
 vide important advantages over present methods of predicting 
 and monitoring flow. The present method for finding prob- 
 able flow paths, such as fractured zones, is a combination of 
 borehole geophysics and examination of cores. The volume in- 
 vestigated is limited to the region close to the boreholes. Im- 
 portant geologic features could be missed unless the boreholes 
 are closely spaced. The present method of monitoring is to 
 sample monitor wells. However, the flow of leach solution 
 through an ore body is not uniform, especially in fractured 
 hard-rock deposits, such as copper and most precious-metal 
 ores. The inhomogeneous flow pattern can decrease the 
 reliability of monitor wells for locating leach solution. Placing 
 monitor wells sufficiently close together so that regulatory 
 agencies are confident that leach solution cannot escape 
 undetected can be expensive. At the 600-m depth at which in 
 situ copper mining can occur, a monitor well costs several tens 
 of thousands of dollars. 
 
74 
 
 Cross-hole geophysical methods with tomographic data 
 analysis and surface-to-borehole methods can examine the 
 region between boreholes. Thus, an important geologic feature 
 is less likely to be missed. Used before injection, these methods 
 could indicate probable flow paths. Comparing measurements 
 made before and after injection could locate leach solution. 
 An additional economic advantage of using geophysical 
 systems over coring and sampling monitor wells is that the 
 systems can be reused, which decreases the overall cost over 
 the life of the mine. 
 
 PRIOR RESEARCH 
 
 Bureau contract research in 1979 (I) 4 indicated that 
 neither surface four-terminal resistivity nor controlled-source 
 audio-magnetotellurics (CSAMT) reliably detected carbonate 
 leach solution at a depth of 80 m at an in situ uranium mine in 
 Wyoming. The surface resistivity measurements were 
 hampered by the thinness of the leached zone compared with 
 its depth from the surface. The CSAMT system suffered from 
 interference, such as that from powerlines. However, 
 technology has improved since that time, so that failure of the 
 surface resistivity and CSAMT systems does not necessarily 
 mean that those techniques would fail with modern 
 geophysical equipment. 
 
 Tomography (explained in the section "Seismic Cross- 
 Hole Tomography") has been applied to geophysical investiga- 
 tions by a number of researchers. Ramirez (2) reported the ap- 
 plication of tomography to electromagnetic (5- to 40-MHz) at- 
 
 tenuation cross-hole data with borehole separations up to 30 
 m. Dines (3) analyzed electromagnetic (50-MHz) cross-hole 
 data, measured both attenuation and velocity, and found at- 
 tenuation to be more diagnostic. Applications to seismic 
 reflection data were made by Chiu (4) and Bishop (5). 
 Analyses of seismic cross-hole data were performed by 
 Albright (6), Gustavsson (7), Peterson (8), and Peterson (9). 
 Foss (10) applied tomography to detecting coal mine hazards 
 with ground-penetrating radar (GPR) but reported that 
 tomography did not conclusively show the clay vein of in- 
 terest. Werniuk (11) described research being coordinated by 
 Paul Young of Queen's University, applying tomographic im- 
 aging to mining-induced seismicity and rock burst phenomena. 
 A helpful summary of the advantages and disadvantages of the 
 various tomographic mathematical methods was given by 
 Worthington (12). 
 
 RECENT AND CURRENT BUREAU RESEARCH 
 
 The Bureau's current emphasis is on seismic cross-hole 
 tomography to predict and monitor the flow of leach solution. 
 Preliminary experiments have been conducted to test the abili- 
 ty of this method to locate a curved water surface, simulating 
 locating leach solution injected above the water table. In addi- 
 tion, seismic tomography has been used for detecting fractured 
 zones in a quarry and in building stones. The Bureau is also 
 evaluating electromagnetic methods to monitor leach solu- 
 tions, which are described in more detail in a subsequent sec- 
 tion. 
 
 SEISMIC CROSS-HOLE TOMOGRAPHY 
 
 BACKGROUND 
 
 The seismic cross-hole method consists of transmitting 
 seismic waves from one borehole to another and measuring the 
 corresponding velocity and/or attenuation. Measuring 
 velocities is easier and the interpretation is usually easier. The 
 seismic velocity is lower and the attenuation is greater in zones 
 of fractures and broken rock than in solid rock. The seismic 
 compressional wave velocity increases with saturation. In 
 Bureau laboratory experiments using competent cores from 
 various rock types, the velocity was typically about 10 pet 
 greater in saturated rock than in dry rock (13). Bureau re- 
 searchers expect that there will not be a measurable difference 
 between the velocity in rock saturated with leach solution and 
 that in rock saturated with ground water. 
 
 The general procedures involved in tomography have 
 been described elsewhere (3, 8-9, 12, 14-15). Only a brief 
 discussion, as it relates to this special application, will be 
 given. Tomographic reconstruction as applied to cross-hole 
 data is a mathematical process for constructing a two- 
 dimensional representation of a field, such as seismic velocity 
 between two boreholes. Figure 1 shows a simplified diagram of 
 the source (transmitter) and receiver locations. In practice, 
 there would be many source and receiver locations, so the 
 region between the boreholes would be crossed by many ray 
 paths. One method of collecting tomographic data consists of 
 setting the source at one position, stepping the receiver 
 through its positions, then moving the source to its next posi- 
 tion and repeating the series of receiver positions. This pro- 
 
 "Italic numbers in parentheses refer to items in the list of references at the end 
 of this paper. 
 
 cedure is repeated until all of the desired ray paths have been 
 generated. Alternatively, the receiver position can be fixed, 
 while the source position is varied. If the borehole is filled with 
 fluid so that coupling is adequate without clamping the source 
 to the borehole wall, the source can be moved continuously 
 and fired at timed intervals. 
 
 Tomographic analysis of cross-hole seismic travel times 
 provides a two-dimensional picture of the distribution of 
 seismic velocities between the source and receiver boreholes. 
 The positions of zones of fractures and leach solution can then 
 be inferred from the seismic velocities. Albright (6) showed 
 that seismic cross-hole tomography can indicate geologic 
 structure, including zones of fractures. Research is being con- 
 ducted to determine the reliability of using the method to show 
 the location of leach solution. The method appears promising 
 for detecting leach solution when the ore is unsaturated before 
 injection. Thus, the seismic method may be applicable for 
 detecting likely flow paths above or below the water table and 
 for locating leach solution injected above the water table. 
 
 Compared with surface techniques, seismic cross-hole 
 methods offer higher resolution, higher sensitivity to features 
 in the area of interest, and lower sensitivity to features outside 
 the area of interest. However, boreholes are expensive; to be 
 practical, the cross-hole methods must use boreholes that are 
 needed for other purposes, such as injection or monitor wells. 
 
 Compared with other cross-hole techniques, the seismic 
 method has several advantages. Seismic waves can easily 
 penetrate several hundred meters and propagate through cas- 
 ings. The equipment is relatively inexpensive. For example, an 
 air gun source costs about $12,000. A disadvantage is that this 
 method cannot distinguish between ground water and leach 
 solution. 
 
75 
 
 Source 
 well 
 
 Injection 
 wells 
 
 Receiver 
 well 
 
 ''....V-lV-'-/— -:* 
 
 ^ / / \ V 
 
 ^ \ 
 
 2 v 
 
 7 
 
 
 Figure 1.— Source (transmitter) and receiver locations in boreholes. 
 
76 
 
 TOMOGRAPHIC ANALYSIS COMPUTER 
 PROGRAM 
 
 Bureau researchers wrote the tomographic analysis com- 
 puter program BOMTOM (Bureau of Mines tomography) 5 
 because available computer programs did not contain options 
 needed for the most effective analysis of cross-hole data. To 
 perform tomographic analysis, the region between the 
 boreholes is mathematically divided into small cells called pix- 
 els. The tomographic reconstruction assigns a velocity to each 
 pixel to provide the best fit to the data. Tomographic analysis 
 of cross-hole data without constraints does not yield a unique 
 reconstruction, even when the number of measurements ex- 
 ceeds the number of pixels. Thus, a unique reconstruction can- 
 not be obtained by increasing the number of cross-hole 
 measurements. Cross-hole data can be fit equally well by many 
 different reconstructions, so appropriate mathematical con- 
 straints must be applied to help select the correct one. A 
 tomographic program providing a test for nonuniqueness and 
 constraints to counteract nonuniqueness was needed. 
 Available programs did not provide such a test or constraints 
 appropriate for in situ mining applications. 
 
 BOMTOM provides a test for uniqueness and constraints 
 to help obtain unique reconstructions. It is easy to run on per- 
 sonal computers commonly used by mining and geophysical 
 companies. Options can be selected interactively. BOMTOM 
 uses the simultaneous iterative reconstruction technique 
 (SIRT) with straight-ray paths. The adequacy of straight-ray 
 paths for in situ mining applications will be examined in future 
 research. A curved-ray program with suitable constraints is be- 
 ing developed for use in situations in which velocity contrasts 
 are large and bending of rays is significant. 
 
 BOMTOM allows the user to test uniqueness by varying 
 the pattern of the initial velocity guesses used to start the 
 iterative solution procedure. A unique reconstruction is in- 
 dependent of the initial velocity guesses. BOMTOM provides 
 the following options for constraints: 
 
 1. Maximum and minimum calculated velocities; 
 
 2. Known, fixed velocities in boreholes, as from sonic 
 logs; 
 
 3. Horizontal layers near the top and/or bottom of the in- 
 vestigated area, in which seismic velocity does not vary with 
 horizontal position; 
 
 4. Smoothing, in which the calculated velocity in a pixel is 
 influenced by the velocities in neighboring pixels. 
 
 A Bureau report (16) describes BOMTOM and results of 
 testing the effectiveness of various combinations of constraints 
 in obtaining unique reconstructions when simulating in situ 
 mining. Calculations were performed with synthetic data of 
 the type expected from in situ mining above the water table, 
 with a dip in the leach solution level between injection wells. 
 The seismic velocity distribution was modeled by attributing a 
 velocity increase of 10 pet to the leach solution. The simula- 
 tions generated a high-velocity artifact near the dip between 
 the wells. (A tomographic artifact is an error in the calculated 
 velocity for the corresponding pixels.) The artifact was re- 
 duced by the constraint that velocity was independent of 
 horizontal position near the top and bottom of the pixel grid. 
 This can be a realistic constraint when the ore at the top of the 
 grid is dry and the ore at the bottom is saturated, and there are 
 no other factors (such as nonhorizontal bedding) causing a 
 significant change with horizontal position near the top and 
 bottom of the grid. Applying the additional constraint of fix- 
 ing the velocities in the boreholes at known values reduced the 
 artifact slightly more. The reconstruction improved signifi- 
 cantly when the upper velocity limit was close to the highest 
 velocity in the model. However, the appropriate upper limit 
 may not be well known for field data. The simulations 
 demonstrated the need to consider nonuniqueness when per- 
 forming such calculations and the desirability of surrounding 
 the investigated region by an area of known or constrained 
 velocities. One way to impose constraints at the sides of the 
 region is to measure the seismic velocities at intervals in the 
 boreholes with sonic logs and fix the corresponding velocities 
 to those values during the tomographic data fitting. 
 
 EQUIPMENT DEVELOPMENT 
 
 The Bureau developed an improved well-locking 
 geophone receiver system. It provides a better signal-to-noise 
 ratio than commercial seismic receiver units that the Bureau 
 has leased. The improved signal-to-noise ratio allows a lower 
 strength source to be used, thereby reducing the risk of damag- 
 ing the well casing. It is smaller, easier to use, and less expen- 
 sive than most commercial units. It will be described in a later 
 publication after clarification of patent status. 
 
 A truck-mounted field system is being prepared for re- 
 cording field data. It includes a generator, an air compressor, 
 winches, and all equipment needed to make field ar- 
 rangements. Field tests will be conducted in the type of rock 
 and at the depths typical of in situ copper mining. 
 
 DETECTING FRACTURES WITH SEISMIC TOMOGRAPHY 
 
 DETECTING FRACTURED ZONES IN A QUARRY 
 
 BOMTOM was used for detecting fractured zones in the 
 Rockwell lime quarry near Rockwood, WI, by analyzing 
 seismic refraction data (77). The general principles behind 
 detecting fractures in quarries and detecting fractures at an in 
 situ mine are identical; namely, the seismic velocity decreases 
 in fractured zones (18). Thus, examining the application of 
 tomography to seismic refraction data can indicate its promise 
 for analyzing seismic cross-hole data to detect fractured zones 
 in in situ mines. 
 
 'Source and executable codes in FORTRAN on 5'/>- and 3'/2-in IBM- 
 compatible diskettes are available upon request from D. R. Tweeton, Twin Cities 
 Research Center, Bureau of Mines, Minneapolis, MN 55417. 
 
 Bureau researchers needed a fracture detection technique 
 to evaluate the success of blast design changes in reducing pit- 
 wall damage. Standard seismic methods were inadequate, so 
 seismic refraction tomography was employed. Each survey 
 consisted of a series of refraction fan shots in which a line of 
 geophones was arranged parallel to and offset from a line of 
 shot points (fig. 2). Refraction tomographic surveys consisted 
 of 24 shot points and 24 geophone stations, yielding 576 ray 
 paths per survey. Seismic travel time data were used for imag- 
 ing the seismic velocity distribution on the refracting rock 
 layer. 
 
 Tomographic reconstructions were performed with BOM- 
 TOM. Travel times were accurate to ±0.05 ms. Velocity con- 
 straints consisted of high and low limits and the use of a 
 uniform initial velocity guess. Other optional parameters in- 
 cluded the use of a delay time to correct for seismic wave travel 
 
77 
 
 Planes of travel paths 
 through surface layer 
 
 Refracting layer 
 
 G Geophone station 
 S Shot point 
 
 Figure 2.— Conceptual drawing of ray-path geometry for series of four refraction fan shots. 
 
 time through the surface layer and the use of a weighted 
 smoothing routine after BOMTOM arrived at a final solution. 
 
 A production blast design was modified so that half of the 
 last row of shot holes were loaded with the quarry's standard 
 explosive load, while the remaining holes were loaded with a 
 reduced-diameter explosive column in the stemming zone (fig. 
 3). This was done to reduce cratering while promoting shearing 
 of the rock. 
 
 The results of the preblast and postblast surveys are 
 shown in figures 3 and 4, respectively. The size of each pixel is 
 0.5 m. The most obvious feature in the preblast tomogram is a 
 low-velocity trend from the lower right to near the upper left 
 corner. Since a predominant joint set was found in the quarry 
 with an average strike of N 26° W, it appears reasonable that 
 this feature is the seismic velocity signature of a similar joint 
 set intersecting the shallow refracting layer. The low-velocity 
 zone in the upper right corner may be the expression of 
 another joint set or it may be a zone of fracturing from a 
 previous blast. 
 
 Comparing the preblast and postblast tomograms (fig. 5) 
 shows that the velocity field of the refracting layer has 
 changed. In the upper left and upper center areas, closest to 
 the standard blast design, zones of relatively high velocity 
 before the blast now appear as low-velocity zones. The low- 
 velocity feature interpreted as jointing in the preblast 
 tomogram is also more pronounced, especially in the lower 
 right area. In addition, the high-velocity area nearest the 
 shotholes with the modified blast design has remained about 
 the same. 
 
 It appears that the changes made in the blast design in the 
 five right-hand shotholes have contributed to a reduction of 
 fracturing in the rock left standing, while the rock nearest the 
 
 shotholes with the standard load has been extensively dam- 
 aged. It is possible, however, that much of the energy pro- 
 duced by the entire blast propagated into or was channeled 
 through the preexisting jointed area, causing some additional 
 fracturing, which makes this zone more pronounced in the 
 postblast tomogram. In either case, the results of the seismic 
 refraction tomographic analysis are consistent with known 
 features of the rock. 
 
 There are two possible sources of error that users of this 
 method should consider. One source of error is the uniform 
 delay time for all ray paths, since lateral velocity variations 
 were present in the surface layer as a result of the blast under 
 investigation and subdrilling from the previous bench. The 
 other source of error is the use of straight-ray paths. The 
 velocity contrast in quarries is large enough to make curved- 
 ray analysis desirable. Work is in progress to develop a curved- 
 ray program with BOMTOM's features. 
 
 The methods applied during this study were successful in 
 identifying major velocity trends in shallow refracting rock 
 layers. The trends can be related with confidence to jointing 
 and zones of intense fracturing caused by mine blasting. Fur- 
 thermore, refraction tomography may hold great potential for 
 future widespread application as a powerful and economic 
 tool for mine planning and blasting design. 
 
 DETECTING FRACTURES IN BUILDING STONES 
 
 BOMTOM has been used by a private company for ex- 
 amining building stones by analyzing seismic travel times for 
 two pairs of faces. Fractured zones that were not visible from 
 the surface were detected with seismic tomography. 
 
78 
 
 m 
 
 Receiver 
 positions 
 
 -21 m 
 Source 
 positions 
 
 -I 
 
 KEY 
 
 3.00-3.25 km/s 
 3.25-3.50 km/s 
 3.50-3.75 km/s 
 3.75-4.00 km/s 
 4.00-4.25 km/s 
 4.25-4.50 km/s 
 4.50-4.75 km/s 
 4.75-5.00 km/s 
 
 standard blasthole load 
 modified blasthole load 
 
 Figure 3.— Preblast tomogram and location of blast design variations. 
 
 m 
 
 
 Receiver 
 positions 
 
 21 m 
 
 Source 
 positions 
 
 N 
 
 KEY 
 
 3.00-3.25 km/s 
 3.25-3.50 km/s 
 3.50-3.75 km/s 
 3.75-4.00 km/s 
 4.00-4.25 km/s 
 4.25-4.50 km/s 
 4.50-4.75 km/s 
 4.75-5.00 km/s 
 
 standard blasthole load 
 modified blasthole load 
 
 Figure 4.— Postblast tomogram. 
 
79 
 
 -0 m 
 
 N 1 
 
 KEY 
 
 Receiver 
 positions 
 
 | -2.0 to -1.5 km/s 
 
 I -1.5 to -1.0 km/s 
 
 | -1.0 to -0.5 km/s 
 
 -0.5 to 0.0 km/s 
 
 0.0 to 0.5 km/s 
 
 0.5 to 1.0 km/s 
 
 11.0 to 1.5 km/s 
 1.5 to 2.0 km/s 
 
 • standard blasthote load 
 
 • modified blasthole load 
 
 21 m 
 
 Source 
 positions 
 
 Figure 5. — Preblast velocities minus postblast velocities. 
 
 LOCATING CURVED WATER SURFACE WITH SEISMIC TOMOGRAPHY 
 
 PRELIMINARY TESTS AT BUREAU'S SITE 
 
 Preliminary experiments to evaluate the use of seismic 
 travel times to locate leach solution injected above the water 
 table were conducted at the Bureau's on-site geophysical test 
 facility. Water was injected instead of leach solution because 
 the change in travel times depends on physically wetting the 
 rock rather than on the chemical properties of the solution. 
 The experiment was conducted to help answer two questions: 
 
 1. Does the increase in the velocity of seismic waves in 
 saturated rock observed in the laboratory (13) require the 
 thorough drying and saturating used in those experiments, or 
 will it also occur under field conditions? 
 
 2. If travel times do decrease as the rock becomes 
 saturated, how well can tomographic analysis locate the 
 regions where the rock changed from dry to wet? 
 
 Tests were conducted by transmitting a seismic signal be- 
 tween boreholes 4.2 m apart in fractured limestone. A mound 
 in the water table was created by injecting water into a 
 borehole between the source and receiver boreholes. The 
 seismic source was an air gun. The receiver was a wall-locking 
 triaxial borehole geophone. The source and receiver positions 
 were 3.7 to 9.1 m below the surface, in steps of 0.3 m. 
 
 Data were first collected when no water was being in- 
 jected. The water level in all boreholes was 6.1 m below the 
 surface. After a tomographic data set was collected, water was 
 injected into the middle well at about 40 L/min. This injection 
 raised the water levels in the source, injection, and receiver 
 
 wells to 4.9, 4.2, and 5.3 m, respectively. A curved water sur- 
 face was desired to test the ability of the tomographic 
 reconstruction to show the shape of the top of the water sur- 
 face between the source and receiver wells. A problem 
 prevented collection of a complete tomographic data set. The 
 air gun repeatedly became stuck in the uncased limestone 
 borehole, so about one-fourth of the desired ray paths near the 
 bottom of the grid could not be used. 
 
 The data 6 were analyzed with BOMTOM, using 
 smoothing. The reconstruction using preinjection data showed 
 high-velocity zones corresponding to water at the expected 
 level, but there were also high-velocity zones above the water. 
 The reconstruction using data taken during injection showed 
 the mound in the water where it should be, but mathematical 
 artifacts and effects not related to the mound were present. 
 There were at least two problems interfering with the 
 tomographic analysis. First, the coverage was not complete. 
 Second, the maximum calculated velocity was three times as 
 great as the minimum calculated velocity, so the use of straight 
 rays could produce artifacts in the reconstruction. Considering 
 those problems, the presence of the mound in the reconstruc- 
 tion was encouraging. However, the results demonstrated the 
 need to obtain data from a site that would allow more com- 
 plete coverage. 
 
 'Data and tomograms are available upon request from D. R. Tweeton, Twin 
 Cities Research Center, Bureau of Mines, Minneapolis, MN. 
 
80 
 
 FIELD TESTS IN LIMESTONE QUARRY 
 Experimental Method 
 
 Because an adequate data set was not obtained from the 
 Bureau's on-site test facility, Bureau researchers gathered a 
 more complete set of travel times at the Basic Materials, Inc., 
 limestone quarry near Waterloo, IA. The test site, shown in 
 figure 6, is mostly competent limestone, with some fractured 
 zones. Inspection of the nearby pit wall allowed the Bureau to 
 select borehole locations between major vertical fractures. 
 
 The source borehole, near the bush on the left side of 
 figure 6, was 18.2 m from the receiver borehole, by the seated 
 person near the right of the figure. The water injection 
 borehole was between them, 6.1 m from the receiver borehole. 
 All three boreholes were 18 m deep and 8.9 cm in diameter. 
 The water levels before injection were 16.3, 12.5, and 12.7 m 
 in the source, injection, and receiver boreholes, respectively. 
 The source and receiver positions were from 2.4 to 12.2 m 
 below the surface, all above the water table. There were 17 
 source positions and 17 receiver positions, directly across from 
 each other. Position intervals were 0.61 m. 
 
 The source, shown in figure 7, was an air gun with a 
 diameter of 6.4 cm and a 330-cm 3 chamber run at 10.3 MPa 
 (1,500 psi). This source provided adequate seismic energy, 
 even in these dry boreholes. The receiver was the wall-locking 
 geophone receiver system developed at the Bureau. The fre- 
 quency transmitted through the limestone was centered at 200 
 Hz. The waves were recorded on a 24-channel seismograph. 
 Each waveform was recorded with 959 samples (959 dots on 
 the seismograph display) at a sampling frequency of 20,000 
 Hz. 
 
 Data were first collected when no water was being in- 
 jected. After the preinjection tomographic data set was col- 
 lected, water was injected into the middle borehole to raise its 
 water level to 5.2 m below the surface. The water was main- 
 tained at that level by injecting about 2 L/h. Such a low flow 
 rate indicates the low permeability of the rock. The water level 
 in the other wells did not rise. The lack of a rise in the receiver 
 borehole water level did not mean that the limestone near that 
 borehole did not become saturated, only that the flow rate was 
 not large enough to cause the water level to rise. After the 
 water level in the middle borehole had stabilized overnight, a 
 second set of tomographic data was collected using the same 
 source and receiver positions as for the preinjection set. 
 
 The precision (measurement error) in the data 7 was 0.05 
 ms, based on the sampling frequency. The accuracy was 
 estimated to be 0.1 ms, based on repeatability. 
 
 Analysis and Results 
 
 The data were analyzed with a curved-ray tomographic 
 program. The ray-tracing portion of the program, RAYPS, 
 was written by Michael Weber at the University of Frankfurt 
 
 (currently at Seismologisches Zentralobservatorium, 
 Grafenberg, Federal Republic of Germany). 8 Chris Calnan of 
 the University of Utah added algorithms for calculating the 
 travel times for individual receivers. The Bureau added 
 algorithms for tomographically calculating the distribution of 
 seismic velocities that best fit the travel times. 
 
 The curved-ray tomographic program adjusts the seismic 
 velocities at nodes and then calculates the derivatives of veloci- 
 ty in triangles with those nodes as corners. The derivatives of 
 seismic velocities are used for calculating the bending that oc- 
 curs in each triangle. A continuous velocity at the triangle 
 boundaries was assumed. There were 16 nodes horizontally 
 and 21 nodes vertically. The positions of source and receiver 
 were centered between nodes at the edge of the grid. The node 
 region was extended below the positions of the source and 
 receiver because the paths tended to curve downward toward 
 the high-velocity zones. In the bottom row of the tomograms, 
 nodes for which there were no nearby paths were left un- 
 colored. The only constraint applied in the analyses was to set 
 the upper velocity limit to 4.2 km/s. Without that constraint, 
 the velocity at some nodes became unreasonably high. 
 
 Five iterations were required to obtain the reconstructions 
 shown in figures 8 and 9. There was little qualitative change 
 from iteration 4 to iteration 5, although average velocities rose 
 slightly as paths were bent more to higher velocity zones. Ad- 
 ditional work will be required to improve the stability of the 
 curved-ray tomographic program. Ideally, the program should 
 be able to run for an unlimited number of iterations without 
 departing from a solution. However, when more than 10 itera- 
 tions were allowed, the program began to give unreasonably 
 high velocities when it was not constrained. 
 
 The reconstruction using preinjection data (fig. 8) showed 
 a high-velocity zone sloping upward from near the bottom on 
 the source side to the receiver side, with a low-velocity zone 
 below the high-velocity zone on the receiver side at a depth of 
 10.4 to 11.6 m. The indicated presence of a low-velocity zone, 
 implying fractures, is consistent with the tendency of the 
 geophone receiver to become stuck at 10 to 1 1 m depth. Unfor- 
 tunately, there were no cores or driller's logs for these 
 boreholes. 
 
 The reconstruction using data taken during injection is 
 shown in figure 9. To facilitate comparing velocities before 
 and during injection, figure 10 displays the velocities during in- 
 jection minus the velocities before injection. The mound in the 
 water level is where it should be. The former low-velocity zone 
 near the bottom on the receiver side acquired a higher velocity. 
 This change is consistent with the flow of water from the injec- 
 tion well into the fractured zone, increasing the seismic veloci- 
 ty. An artifact appeared near the top of the grid, where the 
 calculated velocity increased, although no water was injected 
 there. Also, there were some nodes where the calculated veloci- 
 ty decreased. That decrease is probably an artifact of the 
 
 "Data are available upon request from D. R. Tweeton, Twin Cities Research 
 Center, Bureau of Mines, Minneapolis, MN. 
 
 •The ray tracing was based on Weber's 1986 Ph.D. thesis, "Die Gauss-Beam 
 Methode zur Berechnung Theoretischer Seismogramme in Absorbierenden In- 
 homogenen Medien: Test and Anwendung" ("The Gaussian Beam Method for 
 Calculating Theoretical Seismograms in Absorbing Inhomogeneous Media: In- 
 vestigation and Application"). 
 
J ■-**■ 
 
 
 ^ i^fii-V 
 
 ST^-1 
 
 
 33ET***- "^^"Sfc.-:*-^ 
 
 , ,«-^'^? 
 
 BJj r-H--; J. ; .- r h r'"l 
 
 "J. 
 
 
 '.^ — i , j^~' y - *-. 
 
 AH ^s 
 
 
 
 
 Figure 6.— Limestone quarry field test site. 
 
 81 
 
 Figure 7.— Air gun seismic source. 
 
82 
 
 reconstruction. Thus, modification of the program to allow 
 the application of appropriate constraints will be required. 
 The use of sonic borehole logs to determine the seismic 
 velocities at the sides of the ray-path region could help to 
 reduce artifacts. 
 
 The average velocity was 2.92 km/s before injection and 
 3.24 km/s during injection, a 10-pct difference. That change is 
 consistent with previous laboratory experiments. It was less 
 than the decrease observed in the Bureau site, probably 
 because the quarry rock was less fractured. The travel times 
 decreased by an average of 5 pet. The reason that average 
 velocities increased by a greater percentage than travel times 
 
 decreased is that paths became longer as they bent. Paths bent 
 more during injection because of higher velocity contrasts. 
 
 The results were encouraging. The mound was located in 
 the right position, and most of the pattern of changing 
 velocities was consistent with what was known about the site. 
 The data from this experiment and from others demonstrate 
 that the seismic velocity does increase when dry rock becomes 
 wet under field conditions. Extreme drying or saturating, as in 
 the laboratory, was not necessary to observe the effect, and the 
 increase in velocity was large enough to measure with commer- 
 cial seismic equipment. 
 
 Source 
 positions 
 
 Injection 
 borehole 
 
 Receiver 
 positions 
 
 2.4 m 
 
 12.2 m 
 
 KEY 
 
 1.6 to 2.0 km s 
 
 2.0 to 2.3 km s 
 
 2.3 to 2.6 km s 
 
 2.6 to 2.9 km/s 
 
 2.9 to 3.2 km/s 
 
 3.2 to 3.5 km s 
 
 3.5 to 3.8 km s 
 
 3.8 to 4.2 km/s 
 
 Figure 8.— Seismic velocities before water injection. 
 
 Source 
 positions 
 
 Injection 
 borehole 
 
 Receiver 
 positions 
 
 2.4 m 
 
 
 12.2 m 
 
 KEY 
 
 1.6 to 2.0 km/s 
 2.0 to 2.3 km/s 
 2.3 to 2.6 km/s 
 2.6 to 2.9 km/s 
 2.9 to 3.2 km/s 
 3.2 to 3.5 km/s 
 3.5 to 3.8 km/s 
 3.8 to 4.2 km/s 
 
 Figure 9.— Seismic velocities during water injection. 
 
83 
 
 Source 
 positions 
 
 Injection 
 borehole 
 
 Receiver 
 positions 
 
 2.4 m 
 
 12.2 m 
 
 KEY 
 
 1.4 to 
 
 -0.8 km/s 
 
 0.8 to 
 
 -0.2 km/s 
 
 0.2 to 
 
 0.2 km/s 
 
 0.2 to 
 
 0.5 km/s 
 
 0.5 to 
 
 0.8 km/s 
 
 0.8 to 
 
 1.2 km/s 
 
 1.2 to 
 
 1.6 km/s 
 
 1.6 to 
 
 2.0 km/s 
 
 Figure 10.— Seismic velocities during injection minus velocities before injection. 
 
 ELECTRICAL AND ELECTROMAGNETIC METHODS 
 
 The seismic method is not expected to be able to 
 distinguish between leach solution and ground water. 
 Therefore, leach solution injected below the water table will 
 not be located by seismic methods. A method that responds to 
 a different physical property, such as electrical conductivity, is 
 necessary. The conductivity of acidic solutions will usually be 
 higher than the conductivity of the surrounding ground water, 
 thus providing a good target for electrical and electromagnetic 
 methods. Electrical and electromagnetic methods under con- 
 sideration include magnetotellurics (MT), galvanic resistivity, 
 ground-penetrating radar (GPR), and frequency- and time- 
 domain electromagnetics (FEM and TEM). All of these 
 methods have been used in locating and monitoring con- 
 taminated ground water, chemical spills, and brine plumes and 
 are believed to be applicable to leach solution detection as 
 well. These methods are discussed in many references, such as 
 Telford (19) and the recent comprehensive publication by the 
 Society of Exploration Geophysicists (20). 9 
 
 One of the most formidable problems associated with the 
 detection of leach solution is the great depth at which in situ 
 mining can occur. Some porphyry copper deposits in the desert 
 southwest are over 600 m deep, and many geophysical systems 
 cannot penetrate that far. In addition, complications arise 
 from the presence of conductive overburden, such as saturated 
 clays, which greatly attenuate electrical signals. Thick conduc- 
 tive overburden blankets various parts of the desert southwest 
 where in situ copper mining is being considered. Therefore, 
 geophysical systems having deep-penetrating capabilities are 
 desirable for this application. 
 
 CSAMT is a modification of earlier MT and audio- 
 magnetotelluric (AMT) methods. MT and AMT are natural 
 source systems relying on electrical energy provided by 
 thunderstorms and related natural activity. Although these 
 natural source systems have great depth-penetrating 
 capabilities, they suffer from high noise levels caused by un- 
 
 'A second volume, to be published by the Society of Exploration Geologists, 
 will give a method-by-method treatment of principal electromagnetic techniques. 
 
 predictable electric fields. Consequently, natural source 
 methods would not have the resolution necessary to detect a 
 small, deeply buried leach zone. 
 
 CSAMT, on the other hand, uses an artificial or con- 
 trolled current source, a transmitter with a long grounded 
 dipole or ungrounded loop. Current may be injected at any of 
 several frequencies, depending on the depth desired. This 
 eliminates the dependence on unpredictable electric fields. 
 Although CSAMT does not have the penetration capabilities 
 of its natural source counterparts, it does give more accurate 
 information. 
 
 Like any electrical or electromagnetic method, CSAMT 
 suffers from noise introduced by geologic inhomogeneities, 
 conductive overburden, and multiple or nonuniform targets. 
 A leach zone, whose conductivity, shape, size, and distribution 
 may vary considerably within the ore deposit, may be difficult 
 to recognize in one area and easy to recognize in another. Use 
 of several loop or dipole sizes and frequencies, as well as use of 
 other methods, may be necessary to delineate subsurface solu- 
 tions. Further studies will be necessary to determine the 
 feasibility of CSAMT for this application, despite its earlier 
 failure in the detection of carbonate leach solution (/). 
 
 Galvanic resistivity, or simple resistivity, is a direct- 
 current system in which a voltage is measured as the result of 
 an applied current. This voltage, which is a function of the 
 subsurface resistivity, can be used to map changes in lithology, 
 structure, and soil or ground water conductivity. Resistivity 
 electrode arrays can be configured in several different ways, 
 depending on the survey objective, and can be used in both 
 surface and borehole modes. 
 
 Conventional resistivity methods, which depend on a 
 point source (single electrode) of current, have limited depth 
 of penetration. A modified method, known as focused 
 resistivity, holds more promise, since its depth of penetration 
 is much greater. 
 
 Focused resistivity relies on three equidistant collinear 
 point sources, rather than a single source. Current flowing 
 from the outside sources tend to constrain, or focus, the center 
 
84 
 
 current source. This "focused current" is directed vertically 
 downward and propagates a great distance before diffusing, 
 giving it deep-penetrating capabilities. This system has been 
 used successfully by Southwest Research Institute for detecting 
 subsurface cavities or voids 10 and could be applied to deep 
 conductive bodies, such as leachate plumes, as well. Focused 
 resistivity techniques may also be used in a surface or borehole 
 configuration. 
 
 Theoretically, the borehole configuration would be the 
 most advantageous, since operation takes place closer to the 
 target, but in practice, the borehole configuration is less 
 desirable. Cased holes and conductive leachate fluids in the 
 holes severely reduce the usefulness of borehole techniques. 
 Prior to surveying under such conditions, the casing may have 
 to be pulled out or, more likely, the cased portion left un- 
 surveyed and fluids pumped out of the hole. Some research 
 agencies are investigating three-dimensional resistivity 
 measurements using sources that may penetrate through cas- 
 ings. 
 
 GPR operates by transmitting a high-frequency pulse into 
 the surrounding medium. If this pulse impinges upon an elec- 
 trical conductor, such as a leach solution plume, part of the 
 pulse will be reflected back to the receiver. In a surface con- 
 figuration, GPR can be used to map the lateral extent of the 
 plume. In a borehole configuration, a radial picture of media 
 surrounding the borehole is obtained. 
 
 Unfortunately, GPR, like all high-frequency systems, suf- 
 fers from wave attenuation when used in conductive ground. 
 Wave attenuation reduces signal amplitude to such an extent 
 that deep exploration with such systems becomes impossible. 
 One way to reduce the problem is to decrease the frequency, 
 but this tends to accentuate dispersion and distortion effects. 
 In addition, power consumption is higher and resolution 
 becomes poorer. GPR may be useful in shallow or short-range 
 detection and monitoring applications. 
 
 Electromagnetic methods use an alternating current in a 
 coil to produce a primary magnetic field. Secondary fields are 
 
 generated when eddy currents swirl inside a subsurface con- 
 ductor. These secondary fields produce voltages in receiving 
 coils, which are then recorded by the operator. Measurements 
 can be made in either time or frequency domains. TEM 
 measurements are made by shutting off the primary field, then 
 recording the voltage as a function of time. FEM 
 measurements are made while the primary field is on. The 
 secondary field is then measured as a function of frequency. 
 The receiver may be on the surface or in a borehole. 
 
 TEM is better suited for deep detection of leach solution 
 plumes than is FEM, especially if conductive overburden 
 blankets the area. Conductive overburden tends to diminish 
 the signals going to and coming from the target. To measure 
 the small signals generated by a conductor under a thick con- 
 ductive overburden, it is desirable to increase the signal-to- 
 noise ratio. This may be accomplished in several ways. First, 
 TEM measures the secondary field in the absence of the 
 primary field, thus eliminating the need for exact transmitter- 
 receiver-coil separation and orientation. Exact coil separation 
 is necessary in FEM surveys to compensate for the primary 
 field strength, allowing the much weaker secondary field to be 
 measured. Second, TEM measurements facilitate stacking, a 
 process by which a measurement is repeated many times in a 
 short period of time, which helps to reduce random back- 
 ground noise. Finally, TEM systems employ large transmitter 
 coils, sometimes several hundred meters across, resulting in 
 enormous magnetic dipole moments and hence a large signal. 
 Large dipole moments become extremely important when in- 
 vestigating conductive bodies buried under thick conductive 
 cover. 
 
 Computer simulations are being conducted to investigate 
 whether this method has sufficient sensitivity. Preliminary 
 calculations are encouraging, and Bureau researchers believe 
 that TEM is the most amenable system for locating conductive 
 leach solution plumes. After computer simulations are com- 
 pleted, the most promising methods will be selected for ex- 
 perimental evaluation. 
 
 SUMMARY 
 
 Bureau researchers believe that certain geophysical tech- 
 niques are promising as an aid to detecting probable flow 
 paths and monitoring leach solution during in situ mining. The 
 most promising techniques are seismic cross-hole tomography 
 and electromagnetic induction. Seismic cross-hole tomography 
 has the potential for being a practical procedure because of the 
 relatively low cost of the equipment and the large penetration 
 
 ;0 Vv'ork done under Bureau contract H0245005. 
 
 distance. The Bureau has developed components of a seismic 
 cross-hole tomography system. Analyses of experimental data 
 indicate that the method is promising, but more research will 
 be required to minimize mathematical artifacts unrelated to 
 the presence of leach solutions. Electromagnetic induction is 
 promising in principle because of its ability to detect a good 
 conductor in the presence of poor conductors. GPR may be 
 useful for detecting fracture zones and leach solutions near the 
 borehole, but its limited penetration range will prevent it from 
 being used for monitoring. 
 
85 
 
 REFERENCES 
 
 1. Kehrman, R. F. Detection of Lixiviant Excursions With 
 Geophysical Resistance Measurements During In Situ Uranium 
 Leaching (contract J01 88080, Westinghouse Electric Corp.)- BuMines 
 OFR 5-81, 1979, 157 pp.; NTIS PB 81-171324. 
 
 2. Ramirez, A. L. Recent Experiments Using Geophysical 
 Tomography in Fractured Granite. Proc. IEEE, v. 74, No. 2, 1986, 
 pp. 347-352. 
 
 3. Dines, K. A., and R. J. Lytle. Computerized Geophysical 
 Tomography. Proc. IEEE, v. 67, No. 7, 1979, pp. 1065-1073. 
 
 4. Chiu, S. K. L., E. R. Kanasewich, and S. Phadke. Three- 
 Dimensional Determination of Structure and Velocity by Seismic 
 Tomography. Geophysics, v. 51, No. 8, 1986, pp. 1559-1571. 
 
 5. Bishop, T. N., K. P. Bube, R. T. Cutler, R. T. Langan, P. L. 
 Love, J. R. Resnick, R. T. Shuey, D. A. Spindler, and H. W. Wyld. 
 Tomographic Determination of Velocity and Depth in Laterally Vary- 
 ing Media. Geophysics, v. 50, No. 6, 1985, pp. 903-923. 
 
 6. Albright, J. N., P. A. Johnson, W. S. Phillips, C. R. Bradley, 
 and J. T. Rutledge. The Crosswell Acoustic Surveying Project. Los 
 Alamos Natl. Lab. Rep. LA-11157-MS, Mar. 1988, 121 pp. 
 
 7. Gustavsson, M., S. Ivansson, and J. Pihl. Seismic Borehole 
 Tomography — Measurement System and Field Studies. Proc. IEEE, 
 v. 74, No. 2, 1986, pp. 339-346. 
 
 8. Peterson, J. E., B. N. P. Paulson, and T. V. McEvilly. Applica- 
 tions of Algebraic Reconstruction Techniques to Crosshole Seismic 
 Data. Geophysics, v. 50, No. 10, 1985, pp. 1566-1580. 
 
 9. Peterson, J. E., Jr. The Application of Algebraic Reconstruc- 
 tion Techniques to Geophysical Problems. Ph.D. Thesis, Univ. CA, 
 Berkeley, CA, LBL-21498, 1986, 188 pp. 
 
 10. Foss, M. M., and R. J. Leckenby. Coal Mine Hazard Detection 
 Using In-Seam Ground-Penetrating-Radar Transillumination. 
 BuMines RI 9062, 1987, 27 pp. 
 
 11. Werniuk, J. Tomographic Vision Into Solid Rock. Can. Min. 
 J., v. 109, No. 7, 1988, pp. 24-25. 
 
 12. Worthington, M. H. An Introduction to Geophysical 
 Tomography. First Break, v. 2, pt. 11, Nov. 1984, pp. 20-26. 
 
 13. Thill, R. E., and J. A. Jessop. Effects of Water Saturation on 
 Acoustic Wave Velocity. Soc. Min. Eng. AIME preprint 86-148, 1986, 
 13 pp. 
 
 14. Ivansson, S. Seismic Borehole Tomography — Theory and 
 Computational Methods. Proc. IEEE, v. 74, No. 2, 1986, pp. 
 328-338. 
 
 15. Wattrus, N. J. Two-Dimensional Velocity Anomaly 
 Reconstruction by Seismic Tomography. Ph.D. Thesis, Univ. MN, 
 Minneapolis, MN, 1984, 245 pp. 
 
 16. Tweeton, D. R. A Tomographic Computer Program With Con- 
 straints To Improve Reconstructions for Monitoring In Situ Mining 
 Leachate. BuMines RI 9159, 1988, 70 pp. 
 
 17. Cumerlato, C. L., V. J. Stachura, and D. R. Tweeton. Applica- 
 tion of Refraction Tomography To Map the Extent of Blast-Induced 
 Fracturing. Paper in Key Questions in Rock Mechanics: Proceedings 
 of 29th U.S. Symposium (Minneapolis, MN, June 13-15, 1988). A. A. 
 Balkema, 1988, pp. 691-698. 
 
 18. Sjogren, B. Seismic Classification of Rock Mass Qualities. 
 Geophys. Prospect., v. 27, 1979, pp. 409-442. 
 
 19. Telford, W. M., L. P. Geldart, R. E. Sheriff, and D. A. Keys. 
 Applied Geophysics. Cambridge Univ. Press, 1976, 860 pp. 
 
 20. Nabighian, M. N. (ed.). Electromagnetic Methods in Applied 
 Geophysics, Volume I, Theory, v. 3 in Series Investigations in 
 Geophysics. Soc. Exploration Geophysicists, 1988, 513 pp. 
 
86 
 
 ENHANCED WELL-DRILLING PERFORMANCE WITH 
 CHEMICAL DRILLING-FLUID ADDITIVES 
 
 By Patrick A. Tuzinski, 1 John E. Pahlman, 2 Pamela J. Watson, 3 and William H. Engelmann 4 
 
 ABSTRACT 
 
 The U.S. Bureau of Mines has investigated the use of chemical additives to enhance 
 drilling performance, for which conventional hard-rock drilling and in situ borehole 
 development are target applications. Laboratory diamond core drilling tests were performed 
 on Sioux Quartzite, Westerly Granite, Minnesota taconite, and Tennessee marble, using 
 solutions of cationic additives, nonionic polymer, and acid-base buffered solutions as drill- 
 ing fluids. Penetration improvements ranged from 88 to over 650 pet and bit life im- 
 provements from 56 to over 400 pet with chemical additive solution concentrations that pro- 
 duced a zero surface charge (ZSC) on the rock. The ZSC condition was found to be the most 
 important factor in improving drilling performance. The nonionic polymer was found to be 
 the best additive in the laboratory drilling tests and the most appropriate for field use 
 because the ZSC condition could be obtained over a wide range of solution concentrations 
 (10 to 125 ppm). Field validation tests of rotary tricone drilling with polymer on Minnesota 
 taconite increased penetration rate by about 69 pet and bit life by about 26 pet. Application 
 of this enhanced drilling phenomenon to well drilling for in situ mining should be pursued 
 because of the potential for time and cost savings. 
 
 INTRODUCTION 
 
 As part of an overall program to increase minerals extrac- 
 tion efficiency, the Bureau of Mines has investigated the use of 
 chemical additives in drilling fluids to enhance drilling perfor- 
 mance in hard rock. Watson (7) 5 surveyed and summarized 60 
 yr of literature-reported results of using chemical additives in 
 rock fragmentation processes. In that 60-yr period, some 
 researchers reported no improvements when drilling with ad- 
 ditives in the drilling fluid while others reported significant 
 benefits, especially when drilling with additive solutions that 
 create an electrochemical point of zero charge (PZC) or zero 
 surface charge (ZSC) on the surface of the rock being drilled. 
 Some noteworthy successes include Rehbinder's work (2), in 
 which drilling efficiency increases of up to 60 pet were 
 reported; Shepherd's work (3), in which slightly lower in- 
 creases in drilling efficiencies were obtained; and Westwood's 
 
 'Research geochemist. 
 Supervisory physical scientist. 
 'Mining engineer. 
 ■■Research chemist. 
 
 Twin Cities Research Center, U.S. Bureau of Mines, Minneapolis, MN. 
 'Italic numbers in parentheses refer to items in the list of references at the end 
 of this paper. 
 
 work (4), which indicated that drilling efficiency could be im- 
 proved well over 100 pet. The survey showed that, while the 
 use of chemical additives in the drilling fluid was sometimes 
 beneficial, little scientific reasoning was provided in the 
 literature to explain either the presence or absence of beneficial 
 effects when these chemical additives were used. 
 
 Intrigued by these reports, the Bureau designed and con- 
 ducted laboratory drilling studies to sort out the disparities of 
 earlier research and to establish and understand the necessary 
 boundary conditions for drilling performance improvement 
 under ZSC conditions. This paper describes the laboratory 
 research conducted to determine the boundary conditions of 
 the enhanced ZSC-controlled drilling phenomenon and to 
 determine the applicability of the phenomenon with respect to 
 (1) rock type, (2) bit type, and (3) surface charge (zeta poten- 
 tial) modifier. Also included are results from several initial 
 field validation tests of ZSC-controlled drilling, which suggest 
 that drilling with the recommended additive could produce 
 significant drilling performance improvements. One potential 
 application for the use of additive-assisted drilling is in the 
 development of in situ mining deposits, for which considerable 
 time and cost savings could be realized. 
 
87 
 
 ROCK MATERIALS USED IN TESTS 
 
 All rock samples for laboratory drilling tests were wire 
 sawed into 15-cm cubes. Rock fragments of the same samples 
 were ground to minus 149 ^m for zeta potential measurements 
 and chemical analyses. Previous Bureau reports have given 
 detailed descriptions of Sioux Quartzite and Westerly Granite, 
 including their physical properties (5), and Minnesota taconite 
 and Tennessee marble (6). Briefly, Sioux Quartzite is a 
 homogeneous, fine-grained, metamorphosed sandstone that 
 has a relatively fracture-free structure, is comprised mainly of 
 quartz grains, and averages 98.5 pet silica. Samples were ob- 
 tained from a quarry in southwestern Minnesota. Westerly 
 Granite is a light-gray, fine-grained, equigranular 
 granodiorite, containing about 65 pet feldspars, 25 pet quartz, 
 9 pet micas, and about 1 to 2 pet accessory minerals. The 
 Westerly Granite used in this research was obtained as a single 
 1,500-lb slab from an active quarry in Bradford, RI. Min- 
 
 nesota taconite is a dark green to gray, fine-grained, 
 metasedimentary rock consisting of chert, magnetite, 
 hematite, siderite, and the following silicates: minnesotaite, 
 greenalite, stilpnomelane, and amphibole. The taconite has 
 mildly undulating black bands of magnetite that occur in ir- 
 regular layers, which account for the overall 27 to 30 pet iron 
 content. The tactonite was provided by Erie Mining Co., 
 located near Hoyt Lakes, MN. Tennessee marble (quarry trade 
 name) used in this investigation came from an active mine in 
 the Holston Limestone formation in the Great Valley of east 
 Tennessee (7). The formation is an essentially unmetamor- 
 phosed, coarsely crystalline limestone of Middle Ordovician 
 age. The formation shows a range of colors from light gray to 
 pinks and red to dark brown. Chemically, the rock is about 
 97.4 pet calcium carbonate, with trace amounts of other alkali 
 minerals. 
 
 WATERS USED IN TESTS 
 
 Four waters were used in the drilling tests: distilled, 
 deionized water (DDIW), tap water, mine pond water, and 
 mine well water. Chemical analyses of these waters are given in 
 a previous report (6). 
 
 DDIW was prepared by distilling Minneapolis tap water 
 in a high-capacity 200-L still and then passing it through stand- 
 ard and ultrapure ion-exchange cartridges. The pH of the 
 DDIW was in the range 5.3 to 6.0, and the conductivity 
 measured from 0.3 to 0.5 ^mho/cm. 
 
 The Minneapolis tap water used in these tests had an 
 average pH value of 7.3 to 7.7, and the conductivity was about 
 200 ^mho/cm. 
 
 Both the mine pond water and mine well water were ob- 
 tained from Erie Mining Co. in northern Minnesota. The mine 
 pond water, used as a source of water at Erie Mining during 
 the spring, summer, and fall, is obtained from a pond at the 
 bottom of one of the company's open pits. The mine well 
 water, used as a source of water during the winter, is pumped 
 from a well on the property. The pH of the pond water ranged 
 from 7.0 to 7.5, while the pH of the well water was 7.9 to 8.0. 
 Both had conductivities from 500 to 700 ^mho/cm. 
 
 SURFACE CHARGE CONSIDERATIONS 
 
 The surface charge of most rocks in water around neutral 
 pH is negative. If sufficient amounts of surface charge 
 modifiers are added to the drilling fluid, such as inorganic 
 salts, cationic surfactants, cationic polymers, and a special 
 nonionic polymer, the surface charge can be neutralized, 
 
 resulting in the ZSC condition. The zeta potential gives a 
 measure of rock surface charge. Relationships between zeta 
 potential, double-layer theory, and surface charge have been 
 discussed in a previous report (5). 
 
 ZETA POTENTIAL MEASUREMENTS AND ZSC CONCENTRATIONS 
 
 The procedure used for measuring zeta potentials and 
 determining the ZSC concentration is described in detail 
 elsewhere (5). Using a commercial zeta reader, zeta potentials 
 are determined for the rock particles in water (DDIW, tap, or 
 mine water) while concentrations of additive solutions are in- 
 creased. 
 
 The zeta reader operates on the principle of elec- 
 trophoretic mobility, whereby the speed of a particle in an 
 electric field is proportional to its surface charge. The ap- 
 paratus employs a video display to monitor particle movement 
 in the electric field. When the speed of a moving grid line on 
 the video display is matched to that of the particle, the zeta 
 potential of that particle is shown on a digital readout. 
 
 For the drilling experiments tested, zeta potentials were 
 determined for each rock type in water alone and in water with 
 a series of concentrations of each additive. This is done to 
 generate the range of negative and positive zeta potential 
 values needed to graphically determine the concentration at 
 which the zeta potential or surface charge of the rock particles 
 is zero. It is important to determine each water-rock combina- 
 tion because ZSC concentrations were found to be substantial- 
 ly different for each rock tested and depend both on the 
 chemical additive and type of water used. 
 
88 
 
 DRILLING SYSTEM 
 
 The drilling apparatus employed in this investigation is 
 shown in figure 1. A detailed description of the mechanical 
 and electronic components of the drilling apparatus and the 
 drilling procedure are given in a previous report (5). Briefly, 
 drilling tests were performed on 15-cm rock cubes using a 
 1.12-kW drill press, fitted with a water swivel, and either a 
 16-mm-OD (10-mm ID) diamond-impregnated coring bit or a 
 16-mm tungsten carbide water-cooled spade bit. 
 
 The diamond-impregnated coring bits were rotated at a 
 speed of 100 rpm under a thrust of 150 kg; the spade bits were 
 also rotated at a speed of 100 rpm but under a lower thrust of 
 60 kg. Drilling fluid was flushed through either drill at a rate of 
 150 mL/min. Drilling was done perpendicular to the bedding 
 plane of rocks where applicable. 
 
 The coring bit matrix was 100 pet cobalt (powder) sintered 
 to a Rockwell C hardness of 18 to 21. The diamonds in the 
 matrix were quoted by the manufacturer to be minus 425 fim 
 plus 300 /xm, although the size range of the diamonds was ac- 
 tually minus 1,000 plus 250 ^m. The tungsten carbide spade 
 bits were used only to drill the Tennessee marble. These bits 
 featured replaceable, resharpenable blades with two water- 
 ways. Assembled, they resemble a machinist's straight-fluted- 
 type drill bit of 16 mm diameter. The tip had a 125° included 
 angle, a 125° chisel angle, and a 5° to 8° negative axial rake 
 angle. 
 
 The diamond coring drill bits were sharpened by briefly 
 drilling into a superduty fireclay brick (53 pet silica, 42 pet 
 alumina) to produce a sharpness level corresponding to an ini- 
 tial average penetration rate for one hole of 4.5 mm/min in the 
 test rock sample. The tungsten carbide blades were factory 
 sharpened to a uniform degree; therefore, they were not 
 resharpened in the laboratory. Drilling continued for a se- 
 quence of several holes until the average penetration rate for a 
 hole had dropped to 2.0 mm/min in the test rock sample. The 
 diamond coring drill bits were resharpened before another test 
 began; the tungsten carbide blades were simply changed for 
 the next text. 
 
 Data on total penetration and elapsed time of drilling in 
 progressing from the sharp-bit state of 4.5 mm/min to the 
 dull-bit state of 2.0 mm/min were recorded, added, and then 
 used for test comparisons. The enhanced penetration perform- 
 ance of the additive, as compared with the appropriate 
 baseline water, was calculated by 
 
 The bit life effect of the additive, as compared with the 
 baseline water, was calculated by 
 
 E p = [C p - W p )/W p ] • 100, 
 
 (1) 
 
 where E 
 
 and 
 
 - penetration effect for a test, pet, 
 
 C p = total penetration for drilling with a given ad- 
 ditive, mm, 
 
 W p = average total penetration for drilling with water 
 alone, mm. 
 
 E, = [(C, - W,)/W ( ] • 100, 
 
 (2) 
 
 where E t 
 
 C, 
 and W, 
 
 bit life effect for a test, pet, 
 
 total time for drilling with a given additive, min, 
 
 average total time for drilling with water alone, 
 
 min. 
 
 The results of penetration and bit life effects for additive con- 
 centrations compared with their respective baseline water tests 
 are described in the following section. 
 
 Figure 1.— Drilling apparatus. 
 
LABORATORY DRILLING TEST RESULTS AND DISCUSSION 
 
 Results of laboratory drilling tests are presented in table 1 and figures 2 through 1 1 . 
 
 Table 1.— Drilling results for all additives, percent improvement 
 
 89 
 
 Type of water, additive, 
 and additive concentration 
 
 Penetration 
 
 Bit life 
 
 Type of water, additive, 
 and additive concentration 
 
 Penetration 
 
 Bit life 
 
 SIOUX QUARTZITE 
 
 SIOUX QUARTZITE— Continued 
 
 DDIW 
 1.0 
 1.0 
 
 5.0 
 
 1 6.0 
 7.0 
 8.0 
 1.0 
 3.0 
 5.0 
 7.0 
 1.0 
 5.0 
 1.0 
 DDIW 
 1.0 
 1.0 
 1.0 
 5.0 
 7.0 
 9.0 
 
 1 1.1 
 2.0 
 3.0 
 4.0 
 5.0 
 7.0 
 1.0 
 DDIW 
 9.0 
 1.0 
 
 '2.0 
 3.0 
 5.0 
 DDIW 
 1.0 
 5.0 
 9.0 
 
 '1.4 
 5.0 
 1.0 
 1.0 
 DDIW 
 1.0 
 3.0 
 
 1 4.3 
 5.0 
 1.0 
 
 CaCI 2 , mol/L: 
 
 with AICI 3 , mol/L: 
 
 x10" 8 
 
 x10" 7 
 
 x10~ 7 
 
 x10" 7 
 
 x10~ 7 
 
 x10" 7 
 
 x10" 6 
 
 x10" 6 
 
 x10" 6 
 
 x10" 6 
 
 x10" 5 
 
 x10" 5 
 
 x10" 4 
 
 with 
 
 x10" 
 
 x10" 
 
 x10" 
 
 x10~ 
 
 x10" 
 
 x10" 
 
 x10" 
 
 x10~ 
 
 x10" 
 
 x10" 
 
 x10" 
 
 x10" 
 
 x10~ 
 
 with 
 
 x10" 
 
 x10" 
 
 x10" 
 
 x10" 
 
 x10" 
 
 with 
 
 x10" 
 
 x10" 7 
 
 x10" 7 
 
 x10" 6 
 
 x10" 6 
 
 x10" 5 
 
 x10" 4 
 
 with AI(N0 3 ) 3 , mol/L: 
 
 x10" 7 
 
 x10" 7 
 
 x10" 7 
 
 x10" 7 
 
 x10" 6 
 
 NaCI, mol/L: 
 -2 
 
 ZrCI 4 , mol/L: 
 -7 
 
 - 33.53 
 40.12 
 98.50 
 
 103.83 
 62.87 
 61.99 
 28.60 
 17.21 
 52.47 
 35.81 
 79.64 
 60.32 
 50.40 
 
 .87 
 
 53.23 
 
 28.32 
 
 87.32 
 
 32.60 
 
 68.12 
 
 96.53 
 
 -14.18 
 
 -7.46 
 
 1.42 
 
 -17.18 
 
 -3.70 
 
 1.56 
 
 16.52 
 
 14.86 
 
 115.03 
 
 -3.50 
 
 2.22 
 
 33.32 
 -9.57 
 39.54 
 96.29 
 4.48 
 43.37 
 
 - 46.33 
 
 -14.34 
 84.46 
 98.09 
 
 - 24.32 
 11.38 
 
 -33.76 
 44.69 
 98.71 
 85.12 
 35.02 
 40.96 
 23.13 
 22.28 
 50.31 
 34.17 
 73.23 
 44.36 
 54.55 
 
 13.81 
 36.01 
 24.83 
 73.23 
 
 5.30 
 52.85 
 72.67 
 
 5.74 
 -1.49 
 
 5.30 
 -6.59 
 10.39 
 20.58 
 
 5.30 
 
 5.30 
 
 76.35 
 
 -6.59 
 
 15.49 
 
 40.12 
 -1.49 
 
 35.87 
 
 67.71 
 -8.29 
 
 30.77 
 -36.31 
 
 -5.01 
 55.16 
 63.50 
 -16.35 
 18.82 
 
 DDIW with DTAB, mol/L: 
 
 3.0x10" 4 
 
 ^xlO" 4 
 
 1.5x10" 3 
 
 DDIW with TTAB, mol/L: 
 
 5.0x10" 6 
 
 9.0x10" 6 
 
 1.7x10~ 5 
 
 1.85x10" 5 
 
 2.0x10" 
 
 2.2x10" 
 
 2.8x10" 
 
 5.0x10" 
 
 -5 
 -5 
 
 >-5 
 -5 
 
 -6 
 
 ,-6 
 
 1 7.23x10 -5 
 
 9.0X10" 5 
 
 DDIW with HTAB, mol/L: 
 
 5.0x10 -7 
 
 9.0x10 -7 
 
 1 1.6x10 -6 
 
 3.0x10" 
 
 5.0x10" 
 Tap water with PAA, ppm: 
 
 0.10 
 
 0.15 
 
 0.20 
 
 1 0.25 
 
 0.50 
 
 1.0 
 
 Tap water with PEO, ppm: 
 
 1.0 
 
 1 3.0 
 
 7.5 
 
 12.5 
 
 125.0 
 
 Tap water at pH of 3.8 
 
 with AICI3 
 7.0x10"' 
 1.0x10"' 
 1.15x10" 
 1.18x10" 
 1.6x10" 
 1.7x10"' 
 
 1 1.8x10"' 
 1.9x10"' 
 2.0x10"' 
 3.0x10"' 
 
 mol/L: 
 
 - 32.92 
 118.18 
 111.36 
 
 - 34.79 
 -5.82 
 
 -.19 
 
 1.08 
 
 13.31 
 
 -17.47 
 
 -45.20 
 
 -35.11 
 
 87.84 
 
 -6.86 
 
 9.27 
 71.67 
 87.70 
 
 - 10.22 
 -3.94 
 
 - 16.54 
 
 47.99 
 
 209.17 
 
 334.03 
 
 174.83 
 
 26.82 
 
 66.16 
 429.65 
 348.56 
 363.65 
 387.60 
 
 45.37 
 46.20 
 46.27 
 80.51 
 80.74 
 95.82 
 
 113.63 
 97.02 
 33.06 
 
 - 26.82 
 
 - 27.39 
 99.56 
 
 108.41 
 
 -33.11 
 -7.83 
 
 -4.75 
 
 -4.31 
 
 -3.14 
 
 -11.64 
 
 - 45.60 
 -33.51 
 
 87.02 
 22.60 
 
 1.26 
 
 54.57 
 
 55.94 
 
 -10.80 
 
 -3.27 
 
 -15.52 
 
 29.18 
 
 183.14 
 
 187.22 
 
 124.49 
 
 22.02 
 
 45.55 
 270.58 
 235.69 
 235.90 
 330.05 
 
 16.07 
 49.78 
 20.23 
 48.97 
 66.31 
 34.06 
 59.59 
 76.94 
 6.47 
 29.30 
 
 See footnotes at end of table. 
 
90 
 
 Table 1.— Drilling results for all additives, percent improvement— Continued 
 
 Type of water, additive, 
 and additive concentration 
 
 Penetration 
 
 Bit life 
 
 Type of water, additive, 
 and additive concentration 
 
 Penetration 
 
 Bit life 
 
 WESTERLY GRANITE 
 
 MINNESOTA TACONITE- 
 
 -Continued 
 
 
 DDIW with AICI3, 
 1.0x 10~ 7 . . . 
 3. Ox 10~ 7 . . . 
 
 mol/L: 
 
 18.66 
 
 62.05 
 
 133.75 
 
 104.50 
 
 0.83 
 
 67.28 
 
 92.72 
 
 73.92 
 
 109.00 
 
 -8.46 
 
 -22.70 
 
 136.23 
 
 - 29.23 
 
 -22.07 
 
 -52.01 
 
 -16.92 
 
 Mine well water with 
 Percol 402, ppm: 
 0.40 
 
 105.66 
 
 174.98 
 
 -7.73 
 
 9.90 
 
 31.00 
 138.95 
 278.63 
 
 19.43 
 
 107.80 
 246.76 
 647.52 
 661.94 
 
 92.30 
 
 5. Ox 10~ 7 . . . 
 
 1 0.64 
 
 141.40 
 
 6.0 x 10" 7 
 
 1.0 
 
 -3.48 
 
 1 7 Ox 10~ 7 . . . 
 
 
 154.66 
 
 6.4 
 
 22.34 
 
 I.Ox 10" 6 . . . 
 1 5x 10" 6 . . . 
 
 
 -17.45 
 
 -16.81 
 
 164.92 
 
 Buffered pH solutions, pH units: 
 
 7.0 
 
 54.70 
 
 '2.0x 10~ 6 . . . 
 
 6.0 
 
 129.15 
 
 3.0x 10~ 6 . . . 
 
 
 - 36.42 
 
 -19.79 
 
 66.72 
 
 -16.31 
 
 1 5.5 
 
 218.14 
 
 5.0x 10" 6 . . . 
 
 3.5 
 
 49.29 
 
 7.0x10" 6 . . . 
 1 Ox 10" 5 
 
 Mine pond water with PEO, ppm: 
 
 3.0 
 
 130.66 
 
 
 7.5 
 
 
 
 MINNESOTA TACONITE 
 
 
 220.14 
 
 
 1 12 5 
 
 421.28 
 603.67 
 
 DDIW with AICI3, 
 
 mol/L: 
 
 
 42.68 
 
 28.25 
 
 71.15 
 
 104.46 
 
 104.07 
 
 6.14 
 
 37.50 
 
 125.0 
 
 
 
 
 9.0x10" 7 . .. 
 
 
 53.93 
 
 TENNESSEE MARBLE 
 
 1 Ox 10" 6 . . . 
 
 
 70.20 
 
 DDIW with AICI3, mol/L: 
 
 4.0x10" 7 
 
 15.94 
 25.07 
 50.09 
 84.25 
 
 
 1 1.2x10~ 6 . . 
 
 
 143.45 
 
 39.65 
 
 1.4x10 -6 ... 
 1.7x10" 6 . . . 
 2.0x 10" 6 . . . 
 
 
 135.55 
 
 22.48 
 
 34.17 
 
 1.6X10 -6 
 
 1 1.9x1(T 6 
 
 2.3x10" 6 
 
 50.18 
 
 98.18 
 
 146.09 
 
 
 
 
 ' Point of zero charge concentration. 
 
 2 Compared with baseline water of same pH 3.8 level. 
 
91 
 
 6x10 7 mol/L 
 
 g 125 
 
 a. 100 
 | 75 
 
 UJ 
 
 Q- 50 
 O 25 
 t 
 t-25h 
 
 UJ 
 
 uj-50 - 
 > 10 4 
 
 Q 
 Q 
 
 < 
 
 C, NaCI 
 
 '2xl0~' mol/L 1 
 
 Water baseline 
 L 
 
 10 
 
 ,-2 
 
 10" 
 
 I0" 4 I0" 5 
 
 125 
 100 
 
 75 
 
 50 
 
 25 
 
 -25 h 
 -50 
 
 - f.AKNOsh, _ 4 . 33x|0 -7 mo|/L 
 
 1 
 
 B, CaCI 2 
 
 1 1 
 1 .13 x I0 2 mol/L-*, 
 
 
 
 - 
 
 o f\ 
 
 
 - 
 
 - s' 
 
 ^^**^^>*^ o 
 
 
 
 1 
 
 / 
 
 Water baseline^ 
 
 l I 
 
 V o 
 
 
 A ZrCI 4 
 
 v* 
 
 Water baseline 
 
 _L 
 
 10" 
 
 10 
 
 ,-6 
 
 10" 
 
 10" 
 
 ADDITIVE CONCENTRATION, mol/L 
 Figure 2.— ZSC control of drilling penetration in Sioux Quartzite with inorganic salts in DDIW. 
 
92 
 
 Water baseline- 
 I 
 
 125 
 
 100 
 
 75 
 
 50 
 
 25 
 
 
 
 -25 
 
 > -50 
 t 10 
 
 o 
 
 Q 
 
 < 
 
 -4 
 
 10" 
 
 125 
 100 
 
 75 
 
 50 
 
 25 
 
 -25 h 
 -50 
 
 E, AI(N0 3 ) 3 
 
 10" 
 
 10" 
 
 10" 
 
 1 
 
 C, NaCI 
 
 1 
 2 x |o" 
 
 1 
 mol/L-»j 
 
 A 
 
 ? 
 
 1 
 
 Water basel 
 
 l 
 
 ne^ 
 
 1 
 
 - 
 
 1 
 _ B, CaCI 2 
 
 1 1 
 
 - 
 
 - 
 
 1.13x1 0~ 2 mol/L-^j 
 
 - 
 
 - ^^-" 
 
 
 \'S* 
 
 1 
 
 Water baseline 
 
 1 1 
 
 er o 
 
 10" 
 
 10" 
 
 10 
 
 D, ZrCI 4 
 
 Water baseline 
 J I 
 
 I0 U 10 
 
 10 
 
 10" 
 
 10" 
 
 4.33xl0 7 mol/L 
 
 Water baseline- 
 _| L 
 
 10 
 
 10" 
 
 10 
 
 ,-4 
 
 ADDITIVE CONCENTRATION, mol/L 
 Figure 3.— ZSC control of bit life in drilling Sioux Quartzite with inorganic salts in DDIW. 
 
 10 
 
93 
 
 175 
 
 150 
 
 125 
 
 100 
 75 
 50 - 
 25 - 
 
 o 
 
 CL 
 
 
 
 h -25 
 
 u 
 
 UJ 
 
 u- -50 
 u_ 
 
 /4, Penetration 
 
 '7xlO" 7 mol/L 7 
 
 2x|0~ 6 mol/L' 
 
 Water baseline 
 
 > 
 
 Q 
 O 
 < 
 
 175 
 
 150 - 
 
 125 - 
 
 100 - 
 
 75 
 
 50 - 
 
 25 - 
 
 
 
 -25 h 
 -50 
 
 B, Bit life 
 
 2*10 mol/L 
 
 7x|0 mol/L 
 
 Water baseline 
 l__ 
 
 _L 
 
 10 
 
 10 10"° 10" 
 
 ADDITIVE CONCENTRATION, mol/L 
 
 10 
 
 Figure 4.— ZSC control of drilling penetration and bit life in Westerly Granite with aluminum chloride in DDIW. 
 
94 
 
 o 
 
 Q. 
 
 o 
 
 UJ 
 U_ 
 U_ 
 UJ 
 
 180 
 
 160 
 
 140 
 
 120 
 
 100 
 
 80 
 
 60 
 
 40 
 
 20 
 
 
 
 -20 
 
 -40 
 
 -60 
 
 J. AICI 3 inDD'lW 
 
 l.2x|0" 6 — 
 mol/L 
 
 Water baseline- 
 
 .-6 
 
 10 10 w 10 
 
 ADDITIVE CONCENTRATION, mol/L 
 
 300 
 
 280 
 
 260 
 
 240 
 
 220 
 
 200 
 
 180 
 
 160 
 
 140 
 
 120 
 
 100 
 
 80 
 
 60 
 
 40 
 
 20 
 
 
 
 ■ B, Buffered n< — p h 5.5 
 •pHinDDIW 
 
 I 
 
 7 
 
 II 13 
 
 ADDITIVE CONCENTRATION, pH 
 
 < 
 or 
 
 h- 
 
 LJ 
 
 Ld 
 Q_ 
 
 180 
 160 
 140 
 120 
 
 
 100 
 
 
 80 
 
 
 60 
 
 
 40 
 
 
 20 
 
 
 
 
 
 -20 
 
 
 -40 
 
 0. 
 
 
 ADD 
 
 C;Percol403 
 
 in mine 
 
 well 
 
 water 
 
 Water baseline 
 
 0.64 ppm_ 
 
 I 0.3 0.6 1.5 4.0 
 ITIVE CONCENTRATION, ppm 
 
 800 
 700 
 600 
 500 
 400 
 300 
 200 
 100 
 
 
 ZSConset- 
 between 
 -7.5-I2 ppm 
 
 -100 
 
 D, PEO in mine water 
 
 Water baseline 
 
 1.0 2.5 7.0 16.0 40.0 100.0 150.0 
 ADDITIVE CONCENTRATION, ppm 
 
 Figure 5.— ZSC control of drilling penetration in Minnesota taconite with aluminum chloride, acid, Percol 402, and PEO in 
 various waters. 
 
95 
 
 o 
 
 Q. 
 
 (J 
 
 UJ 
 Ll_ 
 U_ 
 UJ 
 
 UJ 
 Li- 
 
 on 
 
 180 
 
 160 
 
 140 
 
 120 
 
 100 
 
 80 
 
 60 
 
 40 
 
 20 
 
 
 
 -20 
 
 -40 
 
 -60 
 
 AAICL in DDIW 
 
 1.2x10 
 mol/L 
 
 Water baseline 
 
 ^-7 
 
 >-6 
 
 10' I0 V 10 
 
 ADDITIVE CONCENTRATION, mol/L 
 
 180 
 
 160 
 
 140 
 
 120 
 
 100 
 
 80 
 
 80 
 
 40 
 
 20 
 
 
 
 -20 
 
 -40 
 
 
 
 ADD 
 
 i 1 1 
 
 C, Percol 402 in mine wel L 
 
 water _ -.-„ 
 
 —0.64 ppm 
 
 Water baseline 
 
 .1 0.3 0.6 1.6 4.0 
 ITIVE CONCENTRATION, ppm 
 
 I 3 5 7 9 II 13 
 
 ADDITIVE CONCENTRATION, pH 
 
 800 
 
 700 - 
 
 600 - 
 
 500 - 
 
 400 - 
 
 300 - 
 
 200 
 
 100 
 
 0* 
 
 -100 
 
 — i i 1 r 
 
 D, PEO in mine pond water 
 
 ZSC onset between 
 7.5- 12 ppm 
 
 Water baseline 
 1 i 
 
 1.0 2.5 7.0 16.0 40.0 100.0 150.0 
 ADDITIVE CONCENTRATION, ppm 
 
 Figure 6.— ZSC control of bit life in drilling Minnesota taconite with aluminum chloride, acid, Percol 402, and PEO in various 
 waters. 
 
96 
 
 80 
 
 o 
 
 Q. 
 
 160 
 
 H 
 
 140 
 
 O 
 
 UJ 
 
 120 
 
 Li_ 
 
 100 
 
 LU 
 
 80 
 
 o 
 
 60 
 
 h- 
 
 40 
 
 < 
 
 20 
 
 Ul 
 
 
 
 -z. 
 
 UJ 
 
 -20 
 
 Q_ 
 
 -40 
 
 1 
 
 
 180 
 
 
 160 
 
 o 
 
 Q. 
 
 140 
 
 H 
 
 120 
 
 O 
 
 UJ 
 
 100 
 
 u_ 
 u. 
 
 80 
 
 Ul 
 
 60 
 
 UJ 
 
 u. 
 
 40 
 
 Ij 
 
 20 
 
 h- 
 
 
 
 CD 
 
 -20 
 
 
 -40 
 
 L2.3x|0~ 6 mol/L 
 
 Water baseline 
 
 0' 
 
 -7 
 
 10 
 
 r6 
 
 2.3x|0~ 6 mol/L 
 
 Water baseline 
 
 <-i 
 
 n-6 
 
 
 
 r5 
 
 0' 10 10 
 
 ADDITIVE CONCENTRATION, mol/L 
 
 Figure 7.— ZSC control of drilling penetration and bit life in Tennessee 
 marble with aluminum chloride in DDIW. 
 
97 
 
 o 
 
 Q. 
 
 O 
 
 LlI 
 
 u_ 
 u_ 
 
 UJ 
 
 r- 
 
 < 
 or 
 
 t- 
 
 UJ 
 
 UJ 
 Q_ 
 
 180 
 
 160 
 
 140 
 
 120 
 
 100 
 
 80 
 
 60 
 
 40 
 
 20 
 
 
 
 -20 h 
 
 -40 
 
 A, DTAB in DDIW 
 
 9.8x10 
 
 mol/L 
 
 
 
 -4 
 
 Water baseline- 
 
 -#,TTAB inDDIW 
 
 
 
 -3 
 
 o" 2 I0" 6 
 
 1.6 10 mol/L ^ 
 
 180 
 
 160 \C, HTAB in DDIW 
 
 140 
 
 120 
 
 100 
 
 80 
 
 60 
 
 40 
 
 20 
 
 -20 r 
 -40 
 
 Water baseline 
 
 -7 
 
 .-6 
 
 7.2x10 mol/L 
 
 10 
 
 -5 
 
 10 10 10 
 
 ADDITIVE CONCENTRATION, mol/L 
 
 Figure 8.— ZSC control of drilling penetration in Sioux Quartzite with cationic surfactants in DDIW. 
 
 
 
 .-4 
 
98 
 
 o 
 
 Q. 
 
 O 
 
 LJ 
 L_ 
 U_ 
 UJ 
 
 CD 
 
 180 
 160 
 140 
 120 
 100 
 
 80 
 
 60 
 
 40 
 
 20 
 
 -20 h 
 -40 
 
 A, DTAB in DDIW 
 
 9.8*10 
 mol/L 
 
 
 
 -4 
 
 Water baseline 
 
 B, TTAB in DDIW 
 
 7.2x10 mol/L 
 
 Water baseline 
 
 10 
 
 -3 
 
 10 
 
 -2 
 
 10 
 
 -6 
 
 10 
 
 -5 
 
 180 
 
 160 |- C, HTAB in DDIW 
 
 140 
 
 120 
 
 100 
 
 80 
 
 60 
 
 40 
 
 20 
 
 -20 \ 
 -40 
 
 l.6x|0 mol/L 
 
 Water baseline 
 
 .-7 
 
 ^-6 
 
 10 10" 10 
 
 ADDITIVE CONCENTRATION, mol/L 
 
 v-5 
 
 10 
 
 Figure 9.— ZSC control of bit life in drilling Sioux Quartzite with cationic surfactants in DDIW. 
 
99 
 
 o 
 
 Q. 
 
 o 
 
 Ld 
 Q_ 
 U_ 
 UJ 
 
 O 
 
 < 
 or 
 
 \- 
 
 UJ 
 
 -z. 
 
 UJ 
 Q_ 
 
 .8x|0 4 mol/L- 
 
 180 
 
 160 |- A AICI 3 in pH-adjusted tap water . 
 
 140 
 120 
 100 
 
 80 
 
 60 
 
 40 
 
 20 
 
 -20 h 
 -40 
 
 Water baseline 
 
 .8, PAA in tap water 
 —0.25 ppm 
 
 -C, PEO in tap water 
 
 a 
 
 10 10 10 
 
 ADDITIVE CONCENTRATION, mol/L 
 
 450 
 
 400 
 
 350 
 
 300 
 
 250 
 
 200 
 
 150 
 
 100 
 
 50 
 
 
 
 v-3 
 
 0.3 0.6 1.6 4.0 
 TIVE CONCENTRATION, ppm 
 
 -50 
 
 ZSC onset between 
 3-7.5 ppm 
 
 Water baseline 
 
 1.0 2.5 7.0 16.0 40.0 100.0 150.0 
 ADDITIVE CONCENTRATION, ppm 
 
 Figure 10.— ZSC control of drilling penetration in Sioux Quartzite with aluminum chloride, PAA, and PEO in various waters. 
 
100 
 
 o 
 Q. 
 
 O 
 
 L±J 
 
 UJ 
 Ld 
 
 0D 
 
 180 
 
 160 
 
 140 
 
 120 
 
 100 
 
 80 
 
 60 
 
 40 
 
 20 
 
 
 
 -20 
 
 -40 
 I 
 
 _ A, AICI3 in pH-adjusted tap water _ 
 
 1.8X10" mol/L 
 
 >4 
 
 _ Water baseline 
 
 V 
 
 .-5 
 
 .-4 
 
 0~ 10 10 
 
 ADDITIVE CONCENTRATION, mol/L 
 
 450 
 400 
 350 
 300 
 250 
 200 
 
 150 
 
 100 
 
 50 
 
 
 
 -50 
 
 v-3 
 
 450 
 
 400 
 
 350 
 
 300 
 
 250 
 
 200 
 
 150 
 
 100 
 
 50 
 
 
 
 -50 
 0. 
 ADDI 
 
 _B, PAA in tap water 
 
 0.25 ppm 
 
 Water baseline^^ 
 
 1 1 L 
 
 I 0.3 0.6 1.6 4.0 
 TIVE CONCENTRATION, ppm 
 
 ,C, PEO 
 
 1 
 in tap 
 
 1 
 water 
 
 1 
 
 1 
 
 - 
 
 
 
 
 □ 
 
 
 
 
 
 
 
 
 a c 
 
 
 
 - / 
 
 
 ^ZSC onset bet 
 3-7.5 ppm 
 
 ween 
 
 Water 
 
 1 
 
 basel 
 
 ine— ^ 
 
 1 
 
 1 
 
 1 
 
 1.0 2.5 7.0 16.0 40.0 100.0 
 ADDITIVE CONCENTRATION, ppm 
 
 150.0 
 
 Figure 11.— ZSC control of bit life in drilling Sioux Quartzite with aluminum chloride, PAA, and PEO in various waters. 
 
 EFFECT OF ROCK TYPE ON ZSC-CONTROLLED 
 DRILLING 
 
 Determination of the rock's response to ZSC-controlled 
 drilling involved testing of rocks of high-, medium-, and no- 
 silicate content. Sioux Quartzite and Westerly Granite were 
 chosen as representative high-silicate rock types, Minnesota 
 taconite was chosen as the representative medium-silicate rock 
 type, and Tennessee marble (Holston Limestone) was chosen 
 as the representative nonsilicate rock type. 
 
 Sioux Quartzite 
 
 Penetration and bit life effects for drilling Sioux Quartzite 
 with various concentrations of aluminum chloride (AlCl 3 ) 
 solutions, compared with drilling using DDIW, are given in 
 
 table 1 . Drilling with aluminum chloride at the ZSC concentra- 
 tion of 6 x 10" 7 mol/L produced a maximum increase in 
 penetration of 104 pet and a near maximum in bit life of 85 pet 
 (figs. 2A-3A). 
 
 Westerly Granite 
 
 Figure 4 shows plots of penetration and bit life effects for 
 drilling Westerly Granite as a function of aluminum chloride 
 concentration, while table 1 lists the penetration and bit life ef- 
 fects for each aluminum chloride solution concentration 
 tested. Maximums in penetration effect of 155 and 165 pet, 
 respectively, and in bit life effect of 109 and 136 pet were ob- 
 tained when drilling with two concentrations of aluminum 
 chloride. These maximums correspond to the ZSC concentra- 
 tion for the two main components of the granite: quartz at 
 
101 
 
 7xl0" 7 mol/L and feldspars at 2xl0~ 6 mol/L aluminum 
 chloride. The origin of the third maximum is not understood 
 at this time. It was concluded from these results and those for 
 Sioux Quartzite that ZSC-controlled drilling is applicable to 
 rocks containing high amounts of silicate minerals. 
 
 Minnesota Taconite 
 
 Penetration and bit life effects comparing drilling of Min- 
 nesota taconite with various concentrations of aluminum 
 chloride to drilling with DDIW are given in table 1. Drill- 
 ing with aluminum chloride at the ZSC concentration of 
 1.2x 10" 6 mol/L produced simultaneous maximum increases 
 in penetration and bit life of 143 and 104 pet (figs. 5A-6A). It 
 was concluded from these results that ZSC-controlled drilling 
 is also applicable to rocks containing moderate amounts of 
 silicate. 
 
 Tennessee Marble 
 
 Penetration and bit life effects comparing drilling of Ten- 
 nessee marble with various concentrations of aluminum 
 chloride are given in table 1 . The drag, or spade-type, bit was 
 used only for drilling into this rock type. Drilling with 
 aluminum chloride at 2.3xl0~ 6 mol/L, very near the ZSC 
 concentration of 1.93 x 10~ 6 mol/L, resulted in a peak in- 
 crease in penetration improvement of 84 pet and a bit life ex- 
 tension of 146 pet (fig. 7). 
 
 It is concluded from the drilling results for Sioux Quart- 
 zite, Westerly Granite, Minnesota taconite, and Tennessee 
 marble that ZSC-controlled drilling should be applicable to all 
 rock types, whether they have high-, medium-, or no-silicate 
 contents. 
 
 EFFECT OF BIT TYPE ON ZSC-CONTROLLED 
 DRILLING 
 
 ZSC-controlled drilling enhancement with diamond- 
 impregnated bits has been demonstrated for Sioux Quartzite, 
 Westerly Granite, and Minnesota taconite. Drilling of Ten- 
 nessee marble was done with a tungsten carbide spade-type bit 
 to determine whether ZSC-controlled drilling was applicable 
 to other bit types. 
 
 As noted by Westwood (4), the cutting mechanisms of the 
 diamond coring and tungsten carbide spade bit types are suffi- 
 ciently different. Conventionally, spade bits as described here 
 are not used in rock drilling, but rather in the machine-tooling 
 trades. However, if a rock is not too hard, it can be drilled 
 with a spade bit, which works with a scraping or dragging mo- 
 tion. Often the depth of cut for sharp spade bits is greater per 
 pass than diamond bit cutting, all other conditions being 
 equal, but the bits wear down quickly. 
 
 The demonstration of successful ZSC-controlled drilling 
 enhancement for both drag bit types, i.e., diamond drills and 
 spade bits, strongly implies the applicability of ZSC-controlled 
 drilling to these types of bits. It is furthermore suggested from 
 these results that ZSC-controlled drilling should be applicable 
 to all drag bit types and possibly to all drill bit types. 
 
 Drilling of Tennessee marble with the diamond- 
 impregnated coring bit was also tried; however, no decline in 
 drilling rate was observed, even after numerous test holes. This 
 meant that any diamond drilling tests using sharp- and dull-bit 
 states, as defined by this study, would take far too long to be 
 practical in the test program. 
 
 EFFECT OF SURFACE CHARGE MODIFIER ON 
 ZSC-CONTROLLED DRILLING 
 
 Inorganic Salts 
 
 Drilling tests on Sioux Quartzite were conducted using 
 various concentrations of solutions of the chloride salts of 
 aluminum (AlCh), calcium (CaCl 2 ), sodium (NaCl), and zir- 
 conium (ZrCLt), as well as aluminum nitrate [A1(N0 3 )3], as 
 drilling fluids. The resulting penetration and bit life effects are 
 given in table 1 and plotted in figures 2 and 3. These figures 
 show that maximum penetration and bit life improvements 
 were obtained with ZSC concentrations of each of the five in- 
 organic salts tested. Although higher ZSC concentrations were 
 required for lower valence cations, the enhanced drilling per- 
 formance was still obtained. On the basis of these drilling 
 results and those for Westerly Granite, Minnesota taconite, 
 and Tennessee marble with aluminum chloride, it is concluded 
 that drilling with ZSC concentrations of inorganic salt solu- 
 tions should result in improved penetration and bit life 
 responses. Determination of the applicability of ZSC- 
 controlled drilling to all types of surface charge modifiers re- 
 quired testing of previously tested rock types with other types 
 of surface charge modifiers such as organic cationic surfac- 
 tants, cationic polymers, and nonionic polymers with some ca- 
 tionic character. 
 
 Cationic Surfactants 
 
 Four cationic surfactants were tested as surface charge 
 modifiers. Three of these surfactants were familiar quaternary 
 ammonium salts: the 12-carbon dodecyltrimethyl ammonium 
 bromide (DTAB), the 14-carbon tetradecyltrimethyl am- 
 monium bromide (TTAB), and the 16-carbon hex- 
 adecyltrimethyl ammonium bromide (HTAB). The fourth sur- 
 factant was a commercially available low-molecular-weight 
 cationic polymer, Percol 402. DTAB, TTAB, and HTAB drill- 
 ing solutions were prepared in DDIW and tested on Sioux 
 Quartzite; therefore, the drilling performances obtained with 
 these additive solutions were compared with the baseline drill- 
 ing performance for Sioux Quartzite with DDIW. Percol 402 
 drilling solutions were prepared in Erie Mining Co. mine well 
 water and were tested on Minnesota taconite. Percol 402 drill- 
 ing performance was therefore compared with the baseline 
 drilling performance for Minnesota taconite with Erie Mining 
 mine well water. Drilling tests were conducted with surfactant 
 solution concentrations below, at, and above the ZSC concen- 
 tration. 
 
 Penetration and bit life effects for drilling Sioux Quartzite 
 with DTAB, TTAB, and HTAB in DDWI are given in table 1 
 and plotted versus surfactant concentration in figures 8 and 9. 
 Drilling with DTAB at a concentration of 9.8 x 10" 4 mol/L, 
 which is very near the ZSC concentration of 9.4 x 10" 4 mol/L, 
 produced a maximum increase in penetration of 1 18 pet and a 
 near maximum in bit life of 100 pet (figs. 8A-9A). Penetration 
 effect improvements for TTAB and HTAB at their respective 
 ZSC concentrations of7.2xl0 -5 mol/L and 1.6x 10' 6 mol/L 
 were both 88 pet, while bit life extensions were 87 and 56 pet, 
 respectively (figs. 8B-9B, 8C-9C). 
 
 Penetration and bit life effects for Percol 402 are given in 
 table 1 and plotted as a function of concentration in figures 5C 
 and 6C, respectively. Because the exact equivalent molecular 
 weight of Percol 402 was not known, solutions were made up 
 in parts per million expressed as volume per volume. Drilling 
 
102 
 
 with the very dilute solution of Percol 402, which produces the 
 ZSC condition at 0.64 ppm, resulted in the best drilling per- 
 formance: a 175-pct improvement in penetration effect and a 
 141 -pet improvement in bit life. 
 
 It is concluded from the drilling results for cationic sur- 
 factants that drilling with ZSC-concentration solutions of 
 these surfactants also gives rise to simultaneous maximum 
 penetration effect and maximum bit life. 
 
 Acid-Base Solutions 
 
 Surface charge neutralization also occurs at the isoelectric 
 point (1EP) that is achieved by adjusting the pH of the water 
 with either acid or base, depending upon whether the IEP pH 
 is lower or higher than the incipient pH of the water. For 
 magnetite, the IEP pH is about 5.5. Therefore, drilling of 
 Minnesota taconite was conducted with solutions whose pH's 
 were above, at, and below the IEP pH value of 5.5. Penetra- 
 tion and bit life effects are given in table 1 and plotted as a 
 function of pH in figures 5B and 6B. Buffered solutions were 
 employed to ensure that the desired pH was maintained 
 throughout the drilling test. This was important since the ma- 
 jor chemical byproduct of rock drilling has been found to be 
 hydroxide ions (OH). Monitoring the pH of both the influent 
 and effluent streams when drilling with DDIW alone has 
 shown that drilling raises the drilling fluid pH as much as 3 to 4 
 pH units. Drilling with a pH 5.5 buffered solution (acetic acid- 
 sodium acetate) resulted in simultaneous increases of 279 pet 
 in penetration and 218 pet in bit life compared with drilling 
 using DDIW alone. One possible explanation for these much 
 higher drilling performance improvements, which are about 
 twice those for drilling Minnesota taconite with ZSC- 
 concentration solutions of aluminum chloride, is that the ZSC 
 condition is maintained throughout the test by the buffered 
 solution. With the ZSC-concentration solution of aluminum 
 chloride, this is probably not the case, as the hydroxide ions 
 produced in drilling alter or vary the surface charge condition 
 during the course of the drilling test. The results of these tests 
 further point to the universal application of ZSC-controlled 
 drilling irrespective of the type of additive employed to 
 neutralize the surface charge. 
 
 Cationic Polymer 
 
 Because drilling particulates flocculate in a ZSC- 
 concentration solution, flocculation was thought to be partial- 
 ly responsible for the enhanced drilling performance at the 
 ZSC concentration. Therefore, two polymers (one cationic 
 and one nonionic) that flocculate particulates primarily by a 
 molecular bridging mechanism, and secondarily by a charge 
 neutralization mechanism, were tested as drilling-fluid ad- 
 ditives. 
 
 The cationic polymer used as a drilling-fluid additive was 
 the water-soluble, high-molecular-weight polyacrylamide 
 (PAA). Solutions of PAA were prepared in Minneapolis tap 
 water for test drilling in Sioux Quartzite. The drilling perform- 
 ance of PAA was therefore compared with the baseline drilling 
 performance of Sioux Quartzite in Minneapolis tap water. 
 Solution concentrations of PAA below, at, and above the ZSC 
 concentration were tested. Penetration and bit life effects for 
 PAA are given in table 1 and are plotted as a function of con- 
 centration in figures lOfi and llfi. As with other cationic ad- 
 ditives, drilling with the ZSC concentration of PAA (0.25 
 ppm) produced simultaneous maximum increases in penetra- 
 tion and bit life effects, 334 pet and 187 pet, respectively. Also, 
 
 it should be noted that the PAA drilling results clearly indicate 
 that flocculation is not responsible for enhanced ZSC- 
 controlled drilling. While flocculation with PAA occurs at all 
 concentration levels above 0.1 ppm, the enhanced drilling per- 
 formance is achieved only at the ZSC concentration of 0.25 
 ppm. The results of these tests with the cationic polymer also 
 indicate the universality of ZSC-controlled drilling irrespective 
 of the type of additive employed to neutralize the surface 
 charge. 
 
 Nonionic Polymer 
 
 The nonionic polymer used as a drilling-fluid additive was 
 the water-soluble, high-molecular-weight polymer poly- 
 ethylene oxide (PEO). Drilling solutions of PEO were 
 prepared in Minneapolis tap water for testing Sioux Quartzite 
 and in Erie Mining Co. mine pond water for testing Minnesota 
 taconite. The drilling performance of PEO was compared with 
 the respective baseline drilling performance for Sioux Quart- 
 zite in Minneapolis tap water and Minnesota taconite in Erie 
 Mining mine pond water. Although PEO is available in a 
 range of molecular weights from 100,000 to 6 million, drilling 
 tests were conducted only with the 5-million-molecular-weight 
 nonionic polymer because ZSC concentrations (parts per 
 million) were found to be the same as for these other PEO 
 molecular-weight varieties. Penetration and bit life effects for 
 PEO in drilling Sioux Quartzite and Minnesota taconite are 
 given in table 1 and are plotted as a function of concentration 
 in figures 10C and 1 1C for Sioux Quartzite and SD and 6D for 
 Minnesota taconite. 
 
 PEO is an unusual surface charge modifier. Because it is 
 nominally nonionic, it was expected not to neutralize the rock 
 surface charge arid, therefore, not to produce enhanced ZSC- 
 controlled drilling. This expectation was due to the observa- 
 tion that zeta potential determinations of Sioux Quartzite par- 
 ticles in commercial nonionic surfactants, such as Tergitol 
 NPX, Surfynol 465, and hexaethyl cellulose (a high-molecular- 
 weight polymer), showed no effect on the surface charge of the 
 Sioux Quartzite particles at any concentration regardless of 
 molecular weight. Zeta potential determinations of Sioux 
 Quartzite particles in solutions of ever-increasing concentra- 
 tion of PEO, however, showed that PEO, like the cationic ad- 
 ditives (inorganic salts, cationic surfactants, and cationic 
 polymers), alters the zeta potential from an initial negative 
 value to zero. But, unlike the cationic additives, PEO does not 
 produce a positive zeta potential at higher concentrations. In- 
 stead, at higher concentrations, the ZSC condition is main- 
 tained. PEO appears to have some cationic behavior in water, 
 which allows it to neutralize the negative surface charge on 
 rocks and produce the ZSC condition. In the absence of a 
 charged rock surface, the cationic behavior, however, is not 
 active and the rock surface charge remains neutral and is not 
 made positive. 
 
 Comparison of performance graphs of the penetration 
 and bit life effects as a function of PEO concentration for 
 Sioux Quartzite or Minnesota taconite (figs. 10C-1 1C, 5D-6D) 
 with similar plots using cationic surfactants (figs. 8-9, 5C-6C) 
 and inorganic salts (figs. 2-3, 5A-6A) shows why PEO is the 
 best drilling performance enhancer. For Sioux Quartzite, 
 penetration effects of over 350 pet and bit life effects of over 
 235 pet were attained with PEO (from 3 to 125 ppm), and for 
 Minnesota taconite, penetration effects of over 650 pet and bit 
 life effects of over 420 pet were attained with PEO (from 12.5 
 to 125 ppm) compared with the much lesser penetration and 
 bit life effects obtained with ZSC-concentration solutions of 
 the cationic surfactants or inorganic salts. In addition, there is 
 
103 
 
 a wide range of concentrations of PEO that produce a max- 
 imum penetration effect, compared with the single concentra- 
 tion of cationic additives that produces the maximum penetra- 
 tion effect. 
 
 It is concluded on the basis of the tests with PEO that this 
 polymer is the best additive for accomplishing ZSC-controlled 
 drilling. Not only does PEO produce the enhanced ZSC drill- 
 ing phenomenon over a wide range of concentrations, it also 
 produces much greater improvements in bit life and penetra- 
 tion. It is further concluded on the basis of the tests with PEO 
 and the cationic additives that establishment of the ZSC condi- 
 tion is the most important factor in enhanced drilling. 
 
 EFFECT OF DRILLING-FLUID WATER ON 
 ZSC-CONTROLLED DRILLING 
 
 DDIW was used as the baseline comparison water in ini- 
 tial drilling tests to clearly show any additive effect without ad- 
 ditional spectator ions, such as those found in tap water or 
 mine water, to mask the effect. Determining the applicability 
 of ZSC-controlled drilling with regard to water type required 
 testing of previously tested rock types with other water types, 
 such as tap water and mine water. While the pH of the DDIW 
 used in ZSC-controlled drilling research was constantly in the 
 range 5.3 to 6.0, the pH of Minneapolis tap water ranged from 
 7.3 to 7.7 and the pH of Erie Mining Co. mine pond and well 
 
 waters ranged from 7.0 to 7.5 and 7.9 to 8.0, respectively. 
 Therefore, in using aluminum chloride as the additive in tap 
 water or in either of the mine waters, a pH adjustment of these 
 waters was required before aluminum chloride was added, to 
 prevent precipitation and flocculation of the Al + 3 ions in solu- 
 tion as aluminum hydroxide [Al(OH) 3 ]. Acidifying the tap 
 water to pH 3.8 with hydrochloric acid resulted in only a 
 negligible decline in drilling performance on Sioux Quartzite, 
 compared with plain tap water without the addition of acid. 
 
 Penetration and bit life effects for drilling Sioux Quartzite 
 with aluminum chloride solutions made from acidified tap 
 water compared with drilling with acidified tap water only are 
 given in table 1 and graphically displayed in figures \0A and 
 11/4. Drilling with the ZSC concentration of 1.8 x 10 -4 
 mol/L aluminum chloride in pH-adjusted tap water resulted in 
 a maximum increase of 1 14 pet in penetration and a near max- 
 imum of 60 pet in bit life compared with pH-adjusted tap 
 water alone, thereby showing that enhanced drilling perform- 
 ance could also be obtained with aluminum chloride in 
 acidified tap water. 
 
 It is concluded on the basis of these data, and results for 
 drilling Minnesota taconite with Percol 402 and PEO solutions 
 (figs. 5C-6C, 5D-6D) in mine well water and mine pond water, 
 respectively, and Sioux Quartzite with PAA and PEO solu- 
 tions in tap water (figs. 105-1 IB, 10C-11C), that ZSC- 
 controlled drilling is applicable to a wide range of drilling-fluid 
 water characteristics. 
 
 FIELD DRILLING TEST RESULTS 
 
 MINNESOTA TACONITE 
 
 OTHER SILICATE ROCK SUITES 
 
 Field drilling tests are being conducted on taconite at 
 various mining operations in the Mesabi Iron Range of north- 
 ern Minnesota. One of these tests consisted of drilling 
 numerous 50- ft blastholes, at 30- by 30-ft spacings, with 15-in- 
 diameter rotary tricone bits on two adjacent benches of cherty 
 taconite. Drilling performance using an air-water flushing mist 
 was compared with performance using an air-PEO solution 
 mist for drilling particulate removal and dust control. The first 
 bench was drilled solely with water, in 1984, consuming six 
 tricone bits, and has subsequently been blasted away. The sec- 
 ond bench was drilled with water (first 40 holes) and PEO 
 (next 100 holes) in 1987. Figure 12 shows the penetration rate 
 results for this test. The average penetration rate for the 165 
 holes drilled with water in 1984 was 0.55 ft/min, while the 
 average penetration rate for the 40 holes drilled with water in 
 1987 was 0.56 ft/min. The average penetration rate for the 100 
 holes drilled with PEO in 1987 was 0.93 ft/min. Assuming that 
 the penetration rate for the section of bench 2 drilled with 
 PEO in 1987 would be about 0.55 ft/min if the bench had been 
 drilled with water, then drilling with PEO results in a 69-pct in- 
 crease in penetration rate. Also, for water drilling, the average 
 life for two bits was 2,745 ± 12 ft, while for PEO drilling the 
 average life for two bits was 3,468 ±40 ft: a 26-pct increase in 
 bit life. An added benefit in using the air-PEO solution mist in 
 rotary tricone drilling of taconite is that it drastically reduces 
 the amount of dust generated. 
 
 It is concluded from these initial rotary tricone drilling 
 tests that drilling with a mist of ZSC-concentration solutions 
 of PEO in air results in increased penetration rates and in- 
 creased bit life. Field drilling tests are continuing to further 
 validate these initial results. 
 
 Field testing of PEO was also done under cooperative 
 research programs with HDRK Mining Research Limited of 
 Copper Cliff, Ontario, and Centre in Mining and Mineral Ex- 
 ploration Research (CIMMER) at Laurentian University in 
 Sudbury, Ontario, under an umbrella agreement between the 
 Canada Centre for Mineral and Energy Technology 
 (CANMET) and the U.S. Bureau of Mines. 
 
 In the HDRK cooperative research program, 3 weeks of 
 diamond core drilling and percussive drilling tests in quartzite 
 
 6 0.8 
 
 1 
 
 
 i 
 
 1 
 
 i i i i i 
 
 ■ 
 
 
 - 
 
 
 
 
 \ A / 
 
 - 
 
 
 
 J 
 
 1 
 
 
 _ 
 
 - 
 
 
 1 
 1 
 1 
 1 
 
 1 
 1 
 1 
 1 
 1 
 
 
 \/ n 
 
 V 
 
 KEY 
 
 - 
 
 - 
 
 
 1 
 
 1 
 
 
 o Water drilling, 1984 
 a Waler drilling, 1987 
 
 - 
 
 D JX 
 
 
 
 
 ■ Polymer drilling, 1987 
 
 
 ~ */ ^^\° 
 
 / i 
 
 
 
 
 - 
 
 / o / 
 
 6 % tj i 
 
 i 
 
 >a 
 
 
 i 
 
 l 1 1 1 1 
 
 - 
 
 60 80 100 120 
 
 NUMBER OF HOLES DRILLED 
 
 Figure 12.— ZSC control of penetration rate in field drilling 
 of Minnesota taconite with PEO. 
 
104 
 
 and quartz-mica schist were conducted at both underground 
 and surface rock exposures under near-production conditions 
 at the INCO Thompson Mine in Thompson, Manitoba. Dia- 
 mond core drilling tests both on the surface and underground, 
 under certain conditions, demonstrated improved bit life with 
 ZSC-concentration PEO solutions. But, because of the large 
 data scatter and small sampling population, the drilling results 
 did not conclusively demonstrate the enhanced drilling per- 
 formance with ZSC-concentration PEO solutions. The wide 
 scatter in diamond core drilling bit life (footage drilled) can be 
 attributed to inconsistencies in rock morphologies for holes 
 drilled on the surface and variability in bit characteristics for 
 holes drilled underground. 
 
 Drilling performance with ZSC-concentration PEO solu- 
 tions was also determined in long-hole percussive drilling on 
 the surface with 51-mm-diameter cross bits and underground 
 with 51-mm-diameter button bits. Figure 13 shows an unused 
 cross bit, a cross bit after drilling 100 ft in quartzite with 15 
 ppm PEO solution, and a cross bit after drilling 100 ft in the 
 same rock with water alone. It can be seen that the wear on the 
 PEO drilled bit is less than on the water drilled bit. To substan- 
 tiate this visual observation, an optical comparator was used to 
 sketch the 2-D wear profile trace for each of the bits. A typical 
 comparator trace is shown in figure 14 for one of the carbide 
 inserts from each of the three bits shown in figure 13. The area 
 of these traces was measured and then used to calculate the dif- 
 ference in the wear of the carbide inserts. For the same 
 distance of rock drilled, the cross bit used with polymer solu- 
 tion was worn 42.6 pet, whereas without polymer the wear was 
 greater at 48.7 pet. 
 
 Typical comparator traces for one of the carbide button 
 inserts from an unused button bit, a button bit used to drill 200 
 ft of quartz-biotite schist with 15 ppm PEO solution, and a 
 button bit used to drill 200 ft of quartz-biotite schist with 
 water alone are shown in figure 15. The area of these traces 
 was measured and then used to calculate the difference in the 
 wear of the carbide button inserts. For the same distance of 
 
 rock drilled, the button bit used with polymer solution was 
 worn only 5.3 pet, whereas without polymer the wear was 12.1 
 pet, more than twice as much, showing a significant improve- 
 ment using polymer solution. 
 
 In terms of penetration rate, tests using both the cross bit 
 and button bit percussive drills showed no appreciable dif- 
 ference in average penetration rates when the use of polymer 
 solution was compared with use of water alone as the drilling 
 fluid. Cross bits advanced at about 1.9 to 2.2 ft/min, button 
 bits at 1.8 ft/min in the respective rock faces. 
 
 Therefore, it is concluded from these data that drilling 
 with ZSC concentrations of PEO solutions extends the life of 
 both cross and button percussive bits, in spite of the lack of an 
 improvement in penetration advance. Should these percussive 
 bits be drilled to exhaustion, about three to five times longer 
 than tested, it is expected that the polymer solution would fur- 
 ther retard the carbide wear, permitting longer usage time per 
 bit and longer intervals between carbide regrinds. 
 
 In the CIMMER cooperative research program, Longyear 
 Canada, under the guidance of staff from Laurentian Univer- 
 sity and the Bureau, tested PEO as a drilling-fluid additive in 
 exploration core drilling in granite from 3,000 to 5,000 ft 
 below the surface, using BQ-size core bits. There was no 
 measurable change in instantaneous penetration rate; 
 however, bit life (total footage drilled) was improved 2.3- to 
 3.1 -fold when drilling with PEO solutions, compared with 
 drilling using water alone. Also, drill string vibration was 
 reduced when PEO was used, compared with Dromus B solu- 
 ble oil for this lubrication purpose. In addition, below 4,000 
 ft, the diamond-impregnated bits remained continually sharp; 
 they did not have to be periodically sharpened by lowering the 
 drill rotational speed momentarily (increasing torque) to 
 abrade away the dull surface diamonds and expose a new layer 
 of sharp diamonds. It is concluded from these data that bit life 
 is extended and drill string vibration reduced in diamond core 
 drilling using ZSC concentrations of PEO solution. 
 
 Figure 13.— Three cross bits. From left to right: unused bit, bit after drilling 100 ft quartzite with PEO, and bit after drilling 100 ft 
 quartzite with water. 
 
105 
 
 KEY 
 Area, pet Wear, pet 
 
 A 100.0 
 
 B 57.4 42.6 
 
 C 51.3 48.7 
 
 
 KEY 
 
 
 Area, pet 
 
 
 Wear, pet 
 
 A 100.0 
 
 
 
 
 B 84.6 
 
 
 15.4 
 
 C 81.7 
 
 
 18.3 
 
 i-A 
 / /~B 
 
 i / r c 
 
 Reference 
 
 line 
 
 SIDE VIEW 
 
 Figure 14.— Cross-sectional view showing optical comparator wear profile of typical 18-mm tungsten carbide wedge inserts 
 from 51-mm cross bit. 
 
 Toward 
 center 
 of bit 
 
 Edge 
 of bit 
 
 KEY 
 
 Area, pet Wear, pet 
 
 A 1 00.0 0. 
 
 B 94.7 5.3 
 
 C 87.9 I2.I 
 
 Figure 15.— Cross-sectional view showing optical comparator wear profile of typical 12mm tungsten carbide buttons from 
 51-mm button bit. 
 
106 
 
 POTENTIAL APPLICATIONS FOR IN SITU MINING TECHNOLOGY 
 
 Development of an in situ mining production field re- 
 quires diamond core drilling to characterize an ore body and 
 rotary drilling to create boreholes for injection and produc- 
 tion. Thus, rock drilling is absolutely essential to in situ mining 
 methodology. 
 
 It has been shown that PEO as a drilling fluid additive has 
 improved drilling performance in a variety of diamond coring 
 and rotary drilling applications. Since all of the rocks tested 
 thus far (silicates, carbonates, oxides, and sulfide ores) can be 
 treated by dilute PEO solution to achieve a ZSC condition, it 
 would be a practical undertaking to incorporate this polymer 
 into an in situ mining development plan as a drilling-fluid ad- 
 ditive for all drills involved. 
 
 An example of an applicable in situ mining development 
 is in the Santa Cruz ore deposit in Casa Grande, AZ, where the 
 Bureau is conducting an in situ mining field demonstration 
 project jointly with ASARCO Santa Cruz Inc. and Freeport 
 Copper Co. The ore deposit being drilled is composed of a 
 variety of copper oxide minerals hosted in the Oracle Granite 
 and an adjacent quartz monzonite porphyry. This deposit is 
 probably typical, as far as drilling for copper mineralization is 
 concerned. The overall grade for the deposit is about 0.7 pet 
 
 copper; the deposit has never been worked by underground or 
 surface mining, so it will be developed by surface drilling. 
 Plans call for wire-line diamond drilling to obtain 2'/2-in core 
 down to 2,000 ft; later these 4-in-diameter drill holes will be 
 used as the injection wells. Recovery wells will be drilled by 
 rotary drilling to the same depth, starting out as 12-in holes 
 and tapering to 6-in holes at depth. PEO polymer solutions 
 can be used in both instances, as the drills being used are nor- 
 mally equipped to handle fluids, and ZSC concentrations can 
 be determined for each of the rock types encountered. 6 
 Polymer concentrate can be added to the drilling water mixing 
 system at the driller's discretion. 
 
 Other deposits that the Bureau may attempt to initiate for 
 in situ mining research include cherty iron formations for 
 manganese oxides and carbonates in the Cuyuna Range of 
 Minnesota, the Duluth Gabbro formation for copper, nickel, 
 and platinum-group-element sulfides, and a variety of 
 volcanics, elastics, sediments, argillites, carbonates, and fer- 
 rous and siliceous rocks for gold and silver deposits. 
 Laboratory results with PEO suggest that the polymer can be 
 used confidently in drilling and should perform well with all 
 rocks encountered. 
 
 SUMMARY AND CONCLUSIONS 
 
 Laboratory and field ZSC-controlled drilling tests have 
 been conducted to determine the boundary conditions for the 
 enhanced drilling phenomenon. Laboratory drilling tests were 
 performed on Sioux Quartzite, Westerly Granite, Minnesota 
 taconite, and Tennessee marble, using diamond-impregnated 
 coring bits or tungsten carbide spade bits. Drilling fluids were 
 prepared from chemical additives such as inorganic salts, ca- 
 tionic surfactants, acids, cationic polymers, or nonionic 
 polymer in either DDIW, mine water, or tap water at concen- 
 trations below, at, and above the ZSC concentration. 
 
 Enhanced ZSC-controlled drilling performance was 
 observed in laboratory drilling of high-silicate rocks, Sioux 
 Quartzite and Westerly Granite; a medium-silicate rock, Min- 
 nesota taconite; and a nonsilicate rock, Tennessee marble; and 
 in the field drilling of granite, taconite, and quartz-biotite 
 schist, thereby establishing the applicability of ZSC-controlled 
 drilling to a broad range of rocks. 
 
 The applicability of ZSC-controlled drilling using drag bit 
 types was shown by the diamond-impregnated core drilling of 
 Sioux Quartzite, Westerly Granite (5-6), and Minnesota 
 taconite, and the tungsten carbide spade bit drilling of Ten- 
 nessee marble. Field testing with percussive and rotary tricone 
 and diamond-impregnated coring bits showed the applicability 
 of ZSC-controlled drilling to these bit types. 
 
 Enhanced drilling performance has been obtained in the 
 laboratory with solution concentrations of inorganic salts, ca- 
 tionic surfactants, cationic polymers, and nonionic polymer 
 with some cationic character and in the field with PEO with 
 some cationic character. All of these additives neutralized the 
 rock surface charge, thereby establishing that the ZSC condi- 
 tion is the most important factor in ZSC-controlled drilling. 
 
 The applicability of ZSC-controlled drilling with respect 
 to drilling-fluid water source or composition was established 
 by achieving enhanced drilling performance with DDIW, mine 
 water, and both plain and acidified tap water in the laboratory 
 
 and with mine waters in the field. Because of the increased 
 amounts of anions in the mine and tap waters, more additive 
 was required for surface charge neutralization in laboratory 
 tests when using these waters in place of DDIW. 
 
 Maximum penetration improvements obtained for ad- 
 ditives compared with baseline water penetration results in the 
 laboratory ranged from 84 pet for Tennessee marble drilled 
 with a near-ZSC concentration of aluminum chloride in 
 DDIW to over 650 pet for Minnesota taconite drilled with a 
 wide range of ZSC concentrations of PEO in mine water. 
 Maximum bit life improvements obtained for additives com- 
 pared with baseline water bit life results in the laboratory 
 ranged from 56 pet for Sioux Quartzite drilled with a ZSC con- 
 centration of HTAB in DDIW to over 400 pet for Minnesota 
 taconite drilled with a wide range of ZSC concentrations of 
 PEO in mine water. 
 
 The laboratory tests demonstrated that PEO with partial 
 cationic character (polarizable structure) in aqueous solutions 
 was the best additive for accomplishing ZSC-controlled drill- 
 ing. Unlike the cationic additives, which achieved the ZSC 
 condition at only one concentration, PEO maintained the ZSC 
 condition over a wide range of concentrations because of its 
 polarizable structure. Thus, with PEO, the criticality of the 
 solution concentration during drilling is greatly diminished. In 
 addition, use of PEO resulted in much greater improvements 
 in bit life and penetration than use of the cationic additives. 
 
 Initial field drilling results with ZSC-concentration solu- 
 tions of PEO have also demonstrated enhanced drilling per- 
 formance, compared with field drilling results with water 
 
 6 The Bureau has developed a brief videotape and instruction sheet detailing 
 the proper mixing procedure to get polymer into solution and in making dilu- 
 tions. A copy of the video and instruction is available upon request from the Ad- 
 vanced Mining Division, Twin Cities Research Center, Bureau of Mines, Min- 
 neapolis, MN. 
 
107 
 
 alone. In field diamond core drilling in granite, 2.3- to 3. 1-fold 
 improvements in bit life were obtained. In field percussive 
 drilling of 100 ft of quartzite with cross bits, bit wear was 
 slightly reduced. In field percussive drilling of 200 ft of quartz- 
 biotite schist with button bits, bit wear was reduced by more 
 than half. In field rotary tricone drilling in taconite, penetra- 
 tion rate was increased 69 pet (0.93 ft/min versus 0.55 ft/min) 
 and bit life was increased 26 pet (3,400 ft versus 2,700 ft). 
 
 These initial field drilling results are encouraging but need to 
 be validated by further field drilling tests. 
 
 However, development of the PEO polymer as a drilling 
 fluid is far enough along that PEO can be used readily for ex- 
 ploration, quarry, blasthole, and production drilling, as well 
 as in the development of an in situ mining well field where 
 drilling in hard rock is a necessary aspect of mining. 
 
 REFERENCES 
 
 1. Watson, P. J., and W. H. Engelmann. Chemically Enhanced 
 Drilling. An Annotated Tabulation of Published Results. BuMines IC 
 9039, 1985, 53 pp. 
 
 2. Rehbinder, P. A., L. A. Schreiner, and K. F. Zhigach. Hardness 
 Reducers in Drilling. A Physico-Chemical Method of Facilitating the 
 Mechanical Destruction of Rocks During Drilling. Akad. Nauk SSSR, 
 Moscow, 1944; transl. by Counc. Sci. and Ind. Res. (CSIR) (now 
 CSIRO), Melbourne, Australia, 1948, 163 pp. 
 
 3. Shepherd, R. Improving the Efficiency of Rotary Drilling of 
 Shotholes. Trans. Inst. Min. Eng. (England), v. 133, pt. 11, Aug. 
 1954, pp. 1029-1048. 
 
 4. Westwood, A. R. C, N. H. Macmillan, and R. S. Kalyoncu. 
 Chemomechanical Phenomena in Hard Rock Drilling. Trans. Metall. 
 Soc. AIME, v. 256, 1974, pp. 106-111. 
 
 5. Engelmann, W. H., P. J. Watson, P. A. Tuzinski, and J. E. 
 Pahlman. Zeta Potential Control for Simultaneous Enhancement of 
 Penetration Rates and Bit Life in Rock Drilling. BuMines RI 9103, 
 1987, 18 pp. 
 
 6. Pahlman, J. E., W. H. Englemann, P. A. Tuzinski, and P. J. 
 Watson. ZSC-Controlled Drilling For Enhanced Penetration and Ex- 
 tended Bit Life. BuMines RI (in press). 
 
 7. Krech, W. W., F. A. Henderson, and K. E. Hjelmstad. A Stand- 
 ard Rock Suite for Rapid Excavation Research. BuMines RI 7865, 
 1974, 29 pp. 
 
C 2H 89 
 
 : jj 
 

 .H°A .1 
 
 „4°* 
 
 V 7 -^' *°° ^ * "" ;7i * </ 
 
 w •'»;• V .'^': *W* " 
 
 * v % ... 
 
 
 <, *«7v 
 
 
 
 
 
 
 
 P 
 
 
 W 'A V « \/ . McaK . .. ,„,- „ 
 
 * V V '•«' /\ HP' ^ X '• W /\ 'W -v** % ' 
 
 
 
 ^°^ . 
 
 ^o< 
 
 » • "*V» " 
 
 
 
 v^ 1 
 
 
 
 ^<=>' 
 
 * ^ ^P • 
 
 <» *~.r.« ^o A 
 
 
 ..-/ v^-/ \'^\^ v^-/ \^^\/