Search Documents

By Donald Leroux, Dave Scott, Giovanni Di Prisco

"The Lady Loretta lead-zinc deposit in Northwest Queensland, Australia contains an average of 5.9% lead, 17.1% zinc and 97 g/t silver. Ore mineralization is mostly comprised of fairly complex intergrowths of pyrite, sphalerite and galena. Starting in 1998, Noranda Inc., the major joint venture partner of the project, conducted extensive metallurgical test work, which culminated in July - August 2000 in a series of Pilot Plant tests. A common thread to all the metallurgical work has been the deportment of a large portion of silver into non-payable metallurgical products. Average silver distribution from four Pilot Plants shows a distribution of silver in lead concentrate of 28%, whereas silver in zinc concentrate was 49%, and 23% in the final tails.The consistently poor silver recovery as well as the fairly unusual silver distribution, particularly in zinc concentrate, was considered an issue to be thoroughly investigated. A mineralogical plan was devised to determine if these particular metallurgical results were somewhat related to the nature and mode of occurrence of silver carriers. Work was performed to determine discrete silver-bearing minerals and/or silver contained in solid solution in a suite of commonly occurring sulphides. Ore microscopy, SEM methods, Electron Microprobe and Ion Microprobe analyses were all used to answer these questions, and to find the silver carriers, so a metallurgical plan could be devised to improve silver recovery.INTRODUCTIONPotential economic exploitation of the Lady Loretta deposit in Queensland, Australia is driven by revenue derived from zinc and lead flotation concentrates and co-product silver recovery in lead. Relatively low recoveries of payable silver to lead concentrate, and high silver distribution to zinc concentrate necessitated an investigation into the natu"

Jan 1, 2002

By Faan Bornman

"Reeves and Platt1 stated that dense medium separation is one of the most complex unit processes in mineral processing today particularly if cyclones are used. It is also the most efficient gravity concentrating process available to process engineers. The complexity of the process and large number of design and operating variables means that the process has largely remained more an art than a science. Dense medium separation is well established in the diamond, coal and iron ore industries but is gaining popularity as a pre-concentration step in other industries. The maintenance issues are addressed that will affect the efficiency of the operation.INTRODUCTION The purpose of the dense medium separation (DMS) process is to separate the valuable material from the waste based on gravity differences. However, the specific gravity difference between the two fractions must be large enough to achieve high efficiencies. Heavy particles sink through the dense liquid and light particles float. DMS is also used for the pre-concentration of minerals before further downstream processes like communition, in order to minimise the treatment costs. Examples are the pre-concentration of diamond ores or metalliferous ores like nickel, copper and cobalt. Dense medium separation is also used to upgrade the material to meet different market specifications, for example in coal and iron ore applications. A typical dense medium separation process for a diamond application is shown in Figure 1."

Jan 1, 2014

By Patrick Pelletier, Pierre-Alexandre Bossé, Julie Fournier, Gianni Bartolacci, Jayson Tessier, Ahmed Bouajila, Carl Duchesne

"This paper summarizes research into the study of the control of flotation processes based on froth images. The use of numerical image analysis techniques for the extraction of significant patterns (froth structure and colour) from industrial flotation froths is discussed.First, the froth was analysed by the use of the Multivariate Image Analysis (MIA) method to extract the spectral variation contained in a RBG image. The Partial Least Squares (PLS) was applied to develop the empirical model for the prediction of froth grade. The results of this application are illustrated using one industrial case study.Second, the froth was also analysed by Gray Level Co-occurrence Matrix (GLCM) and Wavelet Transform Analysis (WTA) to extract features related more to morphological aspects. Since there is useful information hidden in froth texture, these methods are used as texture descriptor to classify different types of froths.Third, a control strategy based on a froth structure indicator was developed and implemented on an industrial zinc flotation circuit. The feedback controller uses a reagent flow-rate in order to maintain the froth structure at its set point. Process improvements obtained with the implemented control strategy are highlighted.Finally, the feasibility to implement a hybrid control strategy consisting in a froth structure supervision controller based on the froth concentrate grade estimated by MIA and recovery inferred by image analysis is discussed.INTRODUCTIONThe development of control strategies to achieve satisfactory control performance has always been a great challenge for engineers and the need for improved process diagnosis, control and optimization has become more and more evident for the mineral processing industry in the last decades. That is why new process variables need to be measured particularly in highly multivariable processes such as flotation. This industrial process is subject to a wide variety of disturbances mainly generated by frequent and sometimes abrupt changes in physical and/or chemical ore characteristics and also, due to some irregularities in manual process operating procedures. A flotation circuit scheme is always a complex network of more than two dozens slurry streams. Information about those streams (% solid, metal content analysis and flow rates) is important for process diagnosis performance evaluation and control. Also, because of the grade decrease and increasing complexity of the new processed ores, more instrumentation outputs and more powerful and significant pr"

Jan 1, 2005

By J. R. Hart, V. C. Smith, V. P. Babarovich-Hansen, J. H. Schubert

"An introduction and case study of Mintek's latest generation of Milling Circuit controller, the MillStarTM with IES (Interpreting Expert System) is presented. This paper also describes as a case study its installation on a Southern African base metals concentrator plant which has demonstrated improved operation, greatly reduced variation in product quality, as well as increased capacity and availability of the circuit.In this case study, the MillStarTMw as installed along with several other enhancements, including hydrocyclone underflow monitoring, product particle size estimation and flotation circuit stabilisation. MillStarTM utilised the particle size measurement to control grind to within 2% of setpoint, all the while ensuring sump level control, density control as well as mill power optimisation. Mintek' s expertise in the modelling of tumbling mills allowed the development of a specialised optimising controller that ensured the mill was kept in the most efficient operating region.MillStarTM augments the automatic control with useful user interface features, such as on-line trends of any logged variables, historic data warehousing and retrieval and production summary reports.A novel feature that distinguishes the MillStarTMi s the IES (Interpreting Expert System). This is a feature of the software that works with the automatic control code to generate explanations for control actions made by the controller. Often complex control systems are mysterious in nature, reducing trust and eventually resulting in them being switched off. The IES method of exposing the heart of the controller has proved popular and is seen to have potential as a training tool. 314"

Jan 1, 2000

By Peter White

Mine Planning Shaft Collar Shaft Sinking Raises Ramp Underground Water Storage

May 1, 2009

By Michael I. Brittan, Craig M. Sams

"Economies of scale and the variable characteristics of many large ore deposits have resulted in grinding installations using very large SAG mills fitted with variable speed drives. Low-aspectratio single-stage SAG mills were introduced in South Africa to process the consistent gold bearing Witwatersrand quartzites. Elsewhere, circuits were patterned after the more conventional North American circuits with high-aspect-ratio SAG mills followed by ball mills. This paper provides information on two recent Gold Fields group installations covering both types of SAG mill in large, single-stage applications. While these operations process similar tonnages, the mills represent opposite ends of the aspect-ratio spectrum. The Tarkwa mine in Ghana employs a low-aspect 27’ x 42’ 14MW SAG mill. The St Ives mine in Australia operates a conventional 36’ x 18’ 13.4MW SAG mill.INTRODUCTIONGold Fields is one of the world’s largest gold producers with annual production of approximately 4.2 million ounces. The Gold Fields group has operations in South Africa, Ghana and Australia, along with numerous exploration sites and advanced projects in several other countries. This paper discusses new milling operations in Ghana and Australia.TarkwaThe Tarkwa Gold Mine is located in Ghana, West Africa, approximately 350 kilometers west of the capital city of Accra (Figures 1 and 2). The site conditions are tropical with an average annual rainfall of 1,930 mm and a maximum temperature of around 35°C."

Jan 1, 2006

By Robert B. Parsons

"For the fourth time in five years, residents of northern Canada have seen a change in the tax treatment of housing and travel costs. The latest changes announced in December 1990 by the Minister of Finance are to become effective in 1991. When will the tinkering stop?IntroductionFor man y years prior to 1987, certain travel and housing allowances received by employees in remote areas were generally considered by Revenue Canada to be tax-free, even though some observers suspected that such generous treatment was contrary to the provisions of the tax legislation. In the mid-eighties, Revenue Canada decided to reconsider the tax treatment of the mining industry's long standing practice of providing housing and travel assistance to these employees. The tax- free treatment was continued by virtue of annual remission orders, but only pending the enactment of legislation which would tackle the issue head-on and layout the rules in black and white to be seen by all. The sequence of events since then may be chronicled as follows:February 26, 1986-Finance Minister Wilson's budget present s the framework for the future tax treatment of travel and housing benefits provided to residents in remote regions of Canada.December 22, 1986-The Minister of Finance announces detailed rules to fill in the earlier announced framework.January 1, 1987-The new tax rules come into effect.April 29, 1988-The Minister of Finance announces the establishment of a Task Force to examine the existing criteria for determining the eligibility of residents of northern and isolated areas for tax benefits.October 27, 1989-The Task Force presents its report.December 7, 1990-Finance Minister Wilson announces a new system of tax benefits for residents of Canada's northern and isolatedregions.January 1, 1991-The new rules become effective."

Jan 1, 1991

By Bill Mercer

The Triggers for PDAC Involvement in H&S ??Increase in serious accidents in mineral exploration in past few years ??Increasing junior role in exploration ?? Impression that there were few H&S protocols in place in smaller companies ??Lack of national benchmarking ?? The attitude that H&S is a provincial responsibility ??Corporate Social Responsibility initiative What is the issue in exploration? ??Mining?s reputation as ?dirty and dangerous? ??In public?s and media?s minds exploration employees are ?miners? ??In Canada, regulators often do not distinguish between ?exploration? and ?mining?. ??How does exploration compare to mining from a H&S point of view?

May 1, 2008

By K. Lundholm, M. Andersson, L. Nevatalo, Å. Holmström, L. Hedström

"Several copper deposits have the arsenic bearing minerals enargite (Cu3AsS4) and tennantite (Cu12As4S12). Since this mineral contain both arsenic and copper, it is not possible to remove the arsenic by selective flotation. Partial roasting is a method to clean the concentrate from arsenic and thereby produce a suitable raw material for the copper smelter. Also mercury is commonly found in copper bearing minerals. In the partial roasting process both arsenic and mercury are volatilized, while part of the sulphur is used as energy during the roasting. This gives a calcine suitable as feed to a smelter, much like an ordinary copper concentrate. A typical analysis of calcine would be: S content 18–22%, As content below 0.3%, Cu content – not affected, grain size – not affected. However, if the concentrate contains mercury, the mercury will be volatilized during the roasting process. This will lead to very high mercury contents in the subsequent process of sulphuric acid, unless the SO2-rich process gas is pre-treated by the Boliden/Norzink process that recovers mercury as calomel, Hg2Cl2. The volatilized arsenic in the gas phase can be recovered either by dry or wet methods. However, one problem is that arsenic also contaminates the fine grained calcine dust that is captured downstream in the gas cleaning system. The calcine dust contains valuable metals such as copper, gold and silver that needs to be separated from the arsenic. The ultimate aim is to stabilize the recovered arsenic as a compound with minimum impact on the environment and recover the valuable metals. By using wet recovery methods it is possible to selectively precipitate copper, gold and silver as sulphides while arsenic is kept dissolved in the liquid. The precipitated metal sulphides can be mixed with the calcine and sent to the copper smelter. The dissolved arsenic is sent to effluent treatment and is further treated in Outotec ferric arsenate precipitation process, where arsenic is precipitated as the stable compound basic ferric arsenate, formula FeAsO4·xFe(OH)3. All concentrates containing arsenic are unique and the process design of a roaster plant must be based on relevant process data. The process data is collected during tests in lab scale and pilot scale, where the influence of critical process parameters on the quality of calcine is studied. The ability to run tests both in lab scale and pilot scale is crucial to the development of the correct roasting process. Outotec is well equipped for this."

Jan 1, 2014

Operations and maintenance departments have had a long-standing adversarial relationship. The widespread evolution and adoption of modern management techniques, such as reliability centered maintenance (RCM) and Lean, however, has driven interest in working co-operatively to optimize overall plant performance. Mining and minerals operations have long adopted RCM in maintenance and are now adopting Lean throughout organizations. Such modern management initiatives explicitly call for co-operation between the two disciplines of maintenance and production and many advances have been made in terms of people and methodology. One area that has lagged behind in supporting these developments, however, is information systems. Enterprise asset management (EAM) systems in maintenance and newer manufacturing execution systems (MES) in operations have not been truly integrated in the needed ways. Complicated plants, different software architectures, continually changing conditions and many other variables have led to systems problems, which have contributed to the maintenance/operations divide. Master data synchronization (the plant tree) and integrated reporting of real-time data with transactional data are just two of the issues that are being solved by new approaches in information technology, including web services and service orientated architectures. New technologies are coming on-stream to enable the final bridging of the gap.

Jan 1, 2009

By Rees J. Llewellyn

High chromium white irons can be formulated to provide required combinations of abrasion and erosion wear resistance, toughness and capability to withstand corrosion attack. As a consequence, they are used widely in casting form throughout the mining/mineral processing industry, particularly as components for slurry pumps and also as mill and crusher liners, laminated wear plates in chutes and hoppers and in valves, nozzles and cyclones. Traditional cast chromium molybdenum and high chromium white irons for wear applications are covered in ASTM A532 Classes 11 and 111 and are also available as comparable proprietary alloys. They rely on heat-treated, hypoeutectic microstructures containing a high volume fraction of very hard eutectic/secondary metal carbides and martensite to provide extremely high wear resistance. The ability to increase the carbide volume fraction and hence achieve higher wear resistance, has been limited because of severe casting difficulties encountered when carbon and chromium contents are raised and hypereutectic alloys with structures also containing coarse acicular primary metal carbides are produced. The limitation has been overcome and cast hypereutectic chromium white iron pump components have established themselves in a wide variety of mineral processing service applications. Comparisons will be made between the microstructures and wear test performance data for the various alloy types and details will be provided of the service performance improvements that have been achieved through the use of hypereutectic chromium white irons.

May 1, 2004

By Konstantinos Modis

According to recently established theoretical analysis and under certain conditions, a critical sampling density can be determined for an earth-related space-distributed natural variable. Sampling above this critical limit adds little to the mapping results. The objective of this paper is the application of the above theory to the Sesklo clay quarry in Greece, where Heracles General Cement Company had conducted a sampling campaign to estimate its reserves. The structural analysis of Al2O3 values generated a variogram model, from which the critical sampling density was identified. This critical number was by far lower than the existing sampling density. Clay reserves were then estimated using geostatistics and inverse distance squares in different data schemes with variable densities. The results of the comparisons between the various models revealed that, when the sampling density was greater than its critical value, geostatistics and deterministic interpolation gave similar estimations, which is also in agreement to the theory. The optimal sampling grid finally proposed here, will result to significant cost savings in the next drilling program which is now in plan by the company.

Oct 1, 2009

By Erwin L. Zodrow

"The neglect of including the toxic metal mercury in the list of escaping elements (Zodrow, CIM Bulletin, September 1982, pp. 138-140) was pointed out to the author in a private letter by J.E. Rehder, dated September 21, 1982. I welcome this opportunity to present my preliminary mercury research data based on coal investigations supported by N.S.E.R.C and the University College.As mercury is a volatile element and liquid at room temperature, special analytical techniques are required to determine concentrations in whole coal. In the investigation, the coal is analyzed by flameless atomic absorption with a lower detection limit of approximately 10 parts per billion, which, however, is a function of the ash contents of the coals (personal communication, J.L. Brooker, XRAL, 1981).Coal seams of the Sydney Coalfield studied so far include: the Point Aconi, a thin shaly coal, the Upper Bonar, an overrider of the Harbour seam, and the Harbour, Shoemaker and McAulay seams, all of which are shown in stratigraphic succession in Figure 1. The seams have been channel-sampled taking 15-cm coal intervals as one sample for analysis. One channel sample has been investigated from each of the seams listed (two channels were taken from the Point Aconi seam), resulting in 38 samples. Table 1 shows the results, which include mercury contents, ash and total sulphur per sample interval. Figure 2 is a frequency distribution of the 38 determinations."

Jan 1, 1982

By C. H. Hitchcock

History In approaching the subject of diamond-drilling practice, a brief history of the origin and development of the diamond drill may be of interest. The use of black diamonds for boring holes in rock was discovered in 1863 by Rudolph Leschot, a French engineer. In 1864, he employed a diamond bit to bore holes for blasting purposes in the construction of the Mont Cenis tunnel between France and Italy. The method was introduced and developed in the United States in the late sixties. The first deep hole was drilled for coal in 1870, near Pottsville, Penn., by M. C. Bullock. This hole was seven hundred and fifty feet deep. It can be inferred that the diamond drill was introduced into Canada in 1871, when a bore-hole, known as Harper's bore-hole, was drilled for coal at Springhill, Cumberland county, Nova Scotia; and at the old Silver Islet mine, near Port Arthur, the diamond drill was in use as early as 1872. Diamond Drills While M. C. Bullock was developing the diamond drill for exploratory work, the Sullivan Machine Company, of Claremont, N.H., was developing the diamond channelers and gadders for quarry work. These quarry machines were very successful until the high price of black diamonds became prohibitive. In 1878, Albert Ball, chief engineer of the Sullivan Machine Company, designed the Sullivan diamond drill so well known today. The first of these drills went to the Lake Superior iron ranges. After 1880, use of the diamond drill expanded rapidly and spread to all parts of the world where exploratory work was being done. In recent years, the type of diamond drill has not changed much, but better material is going into its construction. Steel castings have replaced cast-iron. Ball and roller bearings are now used instead of brasses and babbits. Gun lubrication has eliminated the oil-can. These improvements make a drill that can be operated for long periods without expense for repairs.

Jan 1, 1933

By J. J. Brummer

The possibility of diamonds occurring in Canada was first raised by W. H. Hobbs in 1899, who concluded: "that the apex of the fan of diamond distribution probably lies somewhere in the strip of the territory bordering James Bay on the east." This was based on the occurrence of diamonds in surficial sediments at various locations south of the Great Lakes. There are two recorded discoveries of drift diamonds in Ontario: (a) the Peterborough Diamond, 33 carats in size, poorly documented and found sometime before 1920;and (b) the Jarvi Diamond, 0.255 carat, found during 1971 in Sheraton Township near Timmins . Both discoveries were in surficial sediments and their source(s) have not been located. During 1960, Selco Exploration Company Ltd. started a heavy-mineral sample program of river gravels in the James Bay Lowlands of Ontario. This was continued by Canadian Rock Company Ltd. (1962-1963) and the Ontario Division of Mines (1973). The surveys discovered the presence of kimberlite minerals in gravel samples from streams draining the Moose River Basin. A total of 53 pyrope grains, four magnesian ilmenites and one grain of chrome diopside were identified from 32 sample locations. A study of the distribution of these minerals suggested the kimberlite source rocks to be located along the north margin or within the Kapuskasing gravity "high" in the area immediately south of James Bay. The Jarvi Diamond, found on an esker east of Timmins, could well have originated from this area. During 1965, an esker was sampled by the Geological Survey of Canada, and the results attracted attention to the area south of Lake Abitibi. At 9 out of 34 sites sampled along the Munro esker, Lee recorded the presence of pyrope garnets and very minor chrome diopside.

Jan 1, 1978

By D. Runnels, A. Allaire, J. Torrealba-Vargas

Data mining and modelling of operational data involves the discovery of patterns in large datasets for future use in benchmarking and plant optimization. The modern pattern of discovery in large data sets involves statistical techniques, machine learning and new visualization tools. Former examples of data mining are revisited and presented in this document. INTRODUCTION A network of enthusiastic and active professionals is currently developing and maintaining several open source statistical software tools (packages) for data analysis (example Rstudio, Orange, Spyder, etc.). The advantages of the open source software philosophy are the fast development and testing of the solutions presented to the community. Books and support are available for each software application. Since each modern work activity generates massive amounts of data, it is attractive to learn and test those available tools considering that we are at the door of the new work revolution coined as “Mine 4.0”. Data mining involves collecting large datasets, the preparation and cleaning of a dataset (term named data wrangling), the uploading of the data to a database program capable of handling big datasets, the development of prediction models and the presentation of the results. Machine learning techniques help to develop mathematical models and algorithms that could be used to improve the performance on a specific task. Mine operations are producing an incredible amount of data each day and the need to analyze those results seems to be impossible due to the lack of time, as well as the lack of skills to investigate the data collected. If we add to this the traditional habit of working in silos, an additional challenge is to consolidate databases from different departments in search of trends that could help to optimize production.

Jan 1, 2019

By J. D. Nye, I. A. Farr, W. J. D. Stone

"A unique deposit of barite exists near Brookfield, Nova Scotia. The crystals are extremely fine and the whole deposit virtually free of heavy metals. The deposit size was not large enough for conventional uses, such as the production of barium carbonate and drilling mud. In an effort to obtain maximum utilization, a program was begun to determine if U.S.P.-grade pharmaceutical barium sulphate for X-ray diagnosis could be produced. Early work was sufficiently encouraging to institute a development program under the federal Program for Advancement of Industrial Technology. This has been very successfully completed. Canadian and U.S. process patents have been issued, clinical trials have been successful, a production plant has been designed and the plant’s production sold. The processing plant will include magnetic separation for the removal of iron minerals, fine grinding by paddle milling to minus one micron and acid leaching. The process is unique in that at no time is the barium taken into solution.Information is given on the deposit, process, markets, use, plant capital and operating costs, and projected profits. The initial production plant will have an annual capacity of 2,000 tons per year of end product. Capital cost is $850,000 and annual turnover more than $1.5 million. As this capacity only represents approximately 15% of the North American market, rapid plant expansion is foreseen."

Jan 1, 1980

By Gilles Landry, Mohamed Ourriban, Pierre Pelletier, Ahmed Bouajila, Claude Gagnon

"This paper summarizes the work carried out to develop and implement the gold cyanidation by size class (CSC) process, a novel concept aiming to reduce cyanide consumption.First, the authors will present the methodology and results of laboratory tests conducted to evaluate the response and the expected gains for the selected ore when submitted to the CSC processing. In the second part, focus will be on the methodology behind the CSC scheme development prior to its implementation at Sleeping Giant Mine concentrator. Finally, the evaluation of the modified circuit performance (operating since July 2007) will be addressed as a reduction of cyanide consumption by more than 30% and of free NaCN in the tailings by more than 50% is confirmed.INTRODUCTIONCyanidation is the most widely used technique to extract gold from hard rock ore. Its use of oxygen as an oxidant still makes it the cheapest process available after more than a hundred years. Unfortunately, gold cyanidation is known to be a slow process that requires high consumption of cyanide. In fact, cyanidation plants are designed with very long residence times (Habashi, F., 1967). Drawbacks of long cyanidation times are, among others, loss of gold in tailings, high reagent consumption and high capital investment and maintenance costs.Responding to their members expectations for the improvement of cyanidation process, COREM undertook a challenging R&D program covering the development of alternative concepts (processes) and the understanding and solving of mass transfer, gold passivation and process water chemistry problems. One of the interesting findings of the R&D program is a patent pending process technology based on the cyanidation by size class (CSC) concept (Côté and Ourriban, 2004; Gagnon and Ourriban, 2007; Gagnon and Ourriban, 2004).The new technology was tested at the Sleeping Giant concentrator, a plant member of COREM. The existing flowsheet was modified in order to implement the CSC process in the full-scale plant. The plant has been operated using the modified CSC based process since August 2007. Laboratory testwork was conducted prior to the industrial application in order to properly design the new process scheme. The modifications included:1) Reclassification of the grinding circuit product through an hydrocyclone2) Leaching of the coarse size fraction3) Addition of the fine fraction to the coarse leaching circuit tails."

Jan 1, 2009

By Lewis Martin

Whereas many researchers and mine engineers have conducted tests using different types of grout, grout ratios, and physical modifications of the cable to determine cable bolt load characteristics, few studies have been conducted on cables fitted with internal instruments. Those studies that have been done have concentrated on cable behavior averaged over significant cable lengths. Researchers at the Spokane Research Laboratory (SRL) of the National Institute for Occupational Safety and Health in Spokane, WA, are investigating the physical properties of cable bolts by replacing the conventional king wire with a modified king wire on which strain gauges have been installed . This paper documents test results on these modified cable bolts. Three tests were conducted at SRL on 1.83-m-long cables grouted into two 0.91-m-long pull-tube assemblies. Load along the cable was monitored with 20 strain gauges installed along the length of the cable. The instrumented cable bolt provided reproducible point measurements of cable load as opposed to load measurements averaged over significant cable lengths. Such point measurements can assist in interpreting the influence of cable confinement, grout quality, rock mass stiffness, and other factors. The instrumented cable bolt is a practical field and research tool because it can predict point loading along the cable and thereby allow interpolation of maximum possible loads on a cable.

May 1, 2002

"The measurement of error in assays collected as part of a mineral exploration program or mining operation historically has been undertaken in a variety of ways. Different parameters have been used to describe the magnitude of relative error, and each of these parameters is related to the standard measure of relative error, the coefficient of variation. Calculation of the coefficient of variation can be undertaken in a variety of ways; however, only one produces unbiased estimates of measurement error: the root mean square coefficient of variation calculated from the individual coefficients of variation.Thompson and Howarth’s error analysis approach has also been used to describe measurement error. However, because this approach utilizes a regression line to describe error, it provides a substantially different measure of error than the root mean square coefficient of variation. Furthermore, because regression is used, Thompson and Howarth’s results should only be used for estimating error in individual samples, and not for describing the average error in a data set. As a result, Thompson and Howarth’s results should not be used to determine the magnitudes of component errors introduced during geochemical sampling, preparation, and analysis.Finally, the standard error on the coefficient of variation is derived, and it is shown that very poor estimates of relative error are obtained from duplicate data. As a result, geoscientists seeking to determine the average relative error in a data set should use a very large number of duplicate samples to make this estimate, particularly if the average relative error is large."

Jan 1, 2007