An ore may be treated experimentally with cyanide solution and 90 per cent, of the gold or silver dissolved. This may be quoted as an extraction of 90 per cent., but it is preferable not to do so, since it is, at best, only an experimental extraction; and by no parallel method of operation in practice could such an “extraction” be actually realized as a recovery. The percentage obviously refers to solution, and the result of the test might as well be stated as such. In practice, an assay or theoretical extraction may be a valuable indication as to the progress of operations, but it is of no value for the purpose of economic comparison unless it happens to check exactly with actual recovery. Again, an ore may be treated by a system of concentration and 70 per cent, of the gold or silver “extracted.” Such an extraction is still further removed from actual recovery of metal or its equivalent money value. When the actual weight of the concentrate and the assay of the same are available for computation then the result may be termed an extraction; but when the amount of metal associated with the mineral is determined by a difference in assay content of the ore before and after concentration then the result is obviously only an assay extraction.
In the case where the final product from the reduction works is solely in the form of bullion the statement of result is, comparatively, an easy matter. The assay extraction shows the amount of metal which has been removed during treatment, as indicated by the difference in assay results. The actual recovery must necessarily be based on the actual amount of metal recovered in the form of bullion. In other instances where a proportion of income is derived from the direct sale of bullion, and a proportion from the sale of a concentrated product, then the assay or theoretical extraction is obtained, as before, by estimation from the difference in assay result before and after concentration and cyanidation. As a check upon this result the combined valuation of the bullion produced, together with the gold and silver indicated in concentrate returns, may be calculated as an actual extraction. Under such conditions actual recovery figures are out of the question; and any statement which adds bullion valuation to smelter returns of concentrate treatment should contain a definite designation of the total as an extraction and not a recovery. Extraction and recovery are terms which are not necessarily interchangeable and the time is ripe for due recognition of the fact.
The final result of metallurgical operation should be stated in a form bearing a definite significance and capable of general application for comparative purposes. By common consent it is invariably reported as a percentage of the original metal content in the ore. At this stage of the discussion it will be necessary to introduce other considerations which will be dealt with in detail. In the statement of assay or theoretical extraction two figures alone are necessary: namely, the result of the assay of the feed and of the residue. The latter figure is generally easily obtained and, for reasons which will be dealt with later, can usually be relied upon. With regard to original metal content in the ore it may be noted that, with most installations, a reliable sample can be obtained by automatic methods after milling. Where the ore is being milled in a cyanide solution containing gold or silver the situation is complicated. It is imperative that the personal element should be eliminated in the sampling of any unground feed and, to this end, automatic sampling and splitting of a considerable proportion of the tonnage handled are imperative. With a reliable knowledge of the metal content of both feed and residue the control of operations is considerably simplified; and the figures obtained can be subsequently checked against the actual extraction or actual recovery figures obtained by direct valuation of bullion, concentrate, or other final product; or from smelter returns giving actual metal content in such product.
Theoretical extraction may be calculated daily, but in order to average the figures obtained a knowledge of the tonnage being handled is necessary. With these data a forecast can be made as to probable yield.
In dealing with actual extraction or actual recovery it has already been pointed out that the total amount of metal ultimately produced can be found by direct valuation. Two other figures are needed before the percentage of actual extraction or actual recovery can be estimated. One of these refers to tonnage, of which a daily estimation is imperative. The difficulty of calculating the actual weight of dry ore going to the mill has already been discussed elsewhere and it is generally conceded that tonnage calculations are best made after the ore has been ground and by means of direct calculation or specific gravity methods. A disadvantage in the latter method of tonnage estimation has been suggested by T. T. Read, who refers to the possible formation of colloid hydrates during fine grinding. I should like to see further discussion on this point and in the meantime I would tentatively suggest that the colloidal state may not be induced by simple comminution. Recent research would seem to allow us to assume that all substances occur in the colloidal, if not in the crystalloidal state; and from Cornu we learn that “the gels of the mineral kingdom are the typical products of every normal weathering process.” I do not think that any amount of fine grinding could transform kaolin into colloidal clay.
In connection with the specific gravity method of arriving at tonnage figures it may be added that, in the case where slime treatment is continuous, there are obvious advantages to be gained by the inclusion in the system of two tonnage tanks, which will incidently act as agitators, and which can alternately be filled with and emptied of pulp. From these the actual dry tonnage being handled can be obtained by specific gravity methods.
In addition to the tonnage figure, an additional estimate, dealing” with gold and silver content either before or after treatment, will be needed before actual recovery percentage can be calculated. Such a figure must be reasonably accurate and for a number of reasons the assay of residue should be used for this purpose. This particular sample can be relied upon as being more representative than the feed sample for the following reasons:
- The ore is finely ground and thoroughly mixed during treatment before being sampled.
- Coarse metal of value has been removed, thus reducing the possibility of error in assaying.
- The final treatment process allows of a correct average sample being taken from a large bulk of material by simple means.
- Where classification is efficient there is no chance of the personal element entering into the question; neither is it possible to vary the metal content at will except in the rare instance where residues are discharged with different percentages of moisture, the latter being, in the form of valuable solution.
An additional reason why reliance may be placed on the residue sample lies in the fact that the figures obtained from the assay of such material may, in the great majority of instances, be regularly and periodically checked by fresh sampling and assaying. This is an important point.
A further advantage is seen in instances where, if the assay of feed had been used, an actual extraction at times of over 100 per cent, would be recorded. There is always a possibility of the actual extraction reaching or exceeding the theoretical or assay extraction, but there is no possibility of the actual extraction reaching or exceeding 100 per cent. A logical method of arriving at final result should preclude, the possibility of absurdity of statement.
The most important phase of descriptive technology dealing with metallurgical operations undoubtedly refers to final result; and in this connection the value of clarity of statement is obvious. I recently attempted to systematize the information, published in book form, and dealing with metallurgical operations on a number of important properties. I found that no comparison of result was possible. In one instance only was a serious attempt made to give information to the reader. In this case the percentage of gold recovered as bullion was calculated from the average assay content of the original ore multiplied by the tonnage treated. In three instances the “extraction” referred to certain assay results, and was obviously only as assay extraction. In two instances there was no indication as to how the published extraction figures had been obtained. In two instances the results were either contradictory or were based on figures bearing no connection whatever with the actual metallurgical extraction. In the final instance the phrase “metallurgical recovery” had been used to include the metal in the rich ore sorted at the breaker plant, the metal shipped in the concentrate, and the metal really recovered as bullion. In this instance the word “recovery” seems to have been more sadly misused than the previous term “extraction.” The percentage figures tell us nothing as to the metallurgical efficiency of the plant or the suitability of the scheme of treatment. As the amount of ore shipped to the smelter rises so the “recovery” increases. By shipping all the ore to the smelter the “recovery” would reach 100 per cent.
A definite standardization of such results is badly needed, and accounts of metallurgical operations would have an added interest and value if accompanied by a clear and concise resume of the actual metallurgical result of such operations. Actual recovery might well be stated in all instances where the final product is in the form of bullion. Actual extraction might be stated in all cases where the final product is in a less concentrated form, and where the yield consists of bullion as well as concentrate, etc. If extraction and recovery figures are unobtainable then a plain statement of the fact is alone necessary. Assay extraction by another name is of little or no value except when it coincides with actual recovery. In the latter event it might just as well be stated as an actual recovery. Assay extraction can always be checked and it is the actual extraction or actual recovery which alone is of economic significance and metallurgical importance.
The point is often overlooked that facts as well as opinions are looked for by the majority of readers. On the subject of filtration of slime pulp fully 90 per cent, of descriptive accounts of filters and filtration omit the fundamental point as to whether or no the filter can or does work efficiently—they avoid a plain statement as to what percentage of displacement of dissolved metal the filter is able to effect.
On the question of the statement of metallurgical result I venture on some definitions:
- The assay or theoretical extraction is the amount of metal which is indicated by assay results as having been removed from the ore during treatment.
- The actual extraction indicates the amount of metal removed from the ore and isolated either as bullion or in other concentrated form. Such actual extraction is estimated by the direct valuation of the bullion and the statement of return from smelter or buyer showing actual metal content in concentrated product, upon which the sale was effected.
- The actual recovery can only be based on the direct valuation of bullion, and the term is only available for use when the bullion produced is approximately the same market value, per se, as the contained gold and silver.
In the calculation of percentage extraction or recovery the following summary will explain the method of statement:
Where a = the assay of feed,
z = the assay of residue, and
t = the tonnage represented,
then 100 (a-z)/a = percentage assay extraction
and a1t1 + a2t2 + a3t3 + / t1+ t2 + t3 = average feed assay = A
and z1t1 + z2t2 + z3t3 + / t1+ t2 + t3 = average residue assay = Z
then 100 (A-Z)/A = average assay extraction percentage.
In the case where an appreciable tonnage is associated with the metal extracted the average assay content in residue per ton milled must be calculated and used in the formula, otherwise actual extraction will exceed assay extraction.
Where M = the amount of metal in bullion, concentrate, or precipitate,
T = the tonnage milled,
t = the tonnage of concentrated product removed, and
Z = the average actual residue assay,
then 100 M/M + Z (T-t) = average actual extraction percentage.
Where M = the amount of metal in bullion,
T = the total tonnage, and
Z = the average residue assay,
then 100 M/M + TZ = average actual recovery percentage.
It is, of course, assumed that the same unit of weight would be used throughout in the calculation of metal content in ore, bullion, concentrate, precipitate, etc.
Any metal in a concentrated product removed from the ore after or during reduction, and the value realized elsewhere by direct sale or otherwise, may be included in the statement of actual extraction. The total amount of metal so extracted would form a basis of calculation for percentage actual extraction. On the other hand, the metal in rich ore abstracted before treatment of the main bulk can form no part of, nor may it be used to influence, a synopsis of such treatment. Its cost of extraction cannot be included as an expenditure per ton milled and its content cannot be added to any statement of recovery made in the reduction plant without falsifying figures which might be obtained dealing with the result of actual metallurgical treatment.
The next consideration refers to the case of calculation of extraction from ore being treated for both gold and silver content, and in the case where both metals occur in quantities of economic significance. A numerical average is of no significance, neither is extraction based on combined metal content before and after treatment of other than metallurgical import. The extraction in current metal value is sometimes of use for comparison, but such a method of statement is not of value for general purposes. The reason for this lies in the fact that the result is influenced by market fluctuations which are beyond the control of the metallurgist. The only exception to this occurs in the rare instance when the percentage extraction of both metals is identical. The extraction and recovery percentages should be stated in such a manner that an alteration in the price of one of the metals would not necessitate an explanation as to a rise or fall in the extraction or recovery figures caused solely by such fluctuation.
With regard to the system to be used in the statement of gold and silver content in ore and bullion, it must be admitted that the metric system is preferable to any other. In troy measurement the units bear no definite relation to one another, and there is no regular factor of difference. The use of the troy grain is now practically discontinued, and the juxtaposition of the facts that 24 grains equal 1 pennyweight and 20 pennyweights equal 1 ounce has often led to serious errors in calculation. It is perhaps too much to hope that the centimeter-gram system may be extended to all calculation used in ore treatment within the lifetime of the present generation.
In most English-speaking countries there seems no objection to the use of the troy ounce in the calculation of silver content in ores, and it is in general use in connection with bullion in most places. The logical equivalent for the use of gold ores is the pennyweight (dwt.), which is perhaps preferable to the use of the decimal subdivision of the ounce. The use of the American dollar as a unit of gold content is illogical, although in the technical literature of the United States its use is practically compulsory. It is common to see a statement that a gold ore has a value of $1. This may mean that the ore assays 1 dwt. of gold, or thereabouts, and this metal if completely recovered in its original state of purity would have a value of $1. The original statement is really a jumble of contradictions, because the ore may not be worth 10c. when placed on the market. The statement of metal content in terms of current coinage is a difficult matter, and verbosity is unavoidable if logic and precision are not to be entirely disregarded.
Experimental and Metallurgical Report Work:
The remarks previously made with regard to extraction are equally applicable in connection with experimental work. Extractions of metal based on differences of assay results should be definitely stated as assay extractions. When they refer to differences in gold and silver content as a result of experimental cyanide treatment such extractions are preferably referred to as solution extractions. They should, in all cases, be distinctly differentiated from probable or possible recovery. These latter figures are only obtainable in the case where the experimental plant is capable of duplicating actual practice in every detail, and where the ore being assayed is not freed from the normal amount of dissolved gold or silver generally found associated with it after ordinary treatment. When preliminary work is carried out on a small scale and where there is no working size filtration or leaching plant available, the assays of the washed ore must be used for the purpose of computation of possible recovery figures. The following ratios are available as a rough guide for this purpose:
These figures introduce the consideration that the more complicated the treatment and the finer the ore is ground the lower becomes the actual recovery of metal. This point is invariably overlooked in reports of experimental work where solution percentage is confused with possible recovery.
The Statement of Working Costs:
In conclusion, I would refer briefly to the question of the statement of working costs, and in this connection I would urge the importance of the local segregation of working cost from general expense, in order that operators in positions of responsibility should be given every opportunity of keeping those expenditures over which they have control under constant scrutiny, so that continued economies in total working costs may be effected and clearly indicated in the returns.
On the question of amortization, or the liquidation of the original cost of the plant by means of payments out of profits of treatment, there is little to be said, because in general mining work the provision of such a sinking fund has been deemed infeasible. There exists, however, no logical reason why theoretical amortization figures should not be taken into consideration in the first place when some definite scheme of treatment has to be decided upon. In the latter event two factors alone are generally supposed to influence the decision: the probable recovery, and the cost of treatment. If amortization costs were added to the cost of treatment in each case and the matter considered from a strictly business standpoint, it is obvious that these charges, coupled with the necessarily higher operating costs, would afford a sound reason against the selection of any process which involved an unnecessarily extensive equipment, involving the necessity for comminution of the ore beyond an economic limit. Even if amortization charges are not feasible in actual practice, there is no reason why they should not be considered in preliminary estimation work.
On the question of repairs and renewals, I would suggest that in a statement of working costs these two items be distinctly differentiated. I would even go further and suggest that items coming under repairs should be added after the total working costs have been arrived at. My reason for this lies in the fact that actual operating costs are for the purpose of locally checking extravagance and unnecessary expenditure and for the encouragement of the introduction of economies. I would point out that it seldom occurs that mill superintendents who are held responsible for working costs are also responsible for either the scheme of treatment or the initial plant arrangement. Repair costs are largely an index of the class of plant erected in the first instance; and if such an item were isolated and added to the total working costs it would form a powerful argument against misplaced economy in the first instance.
In the statement, of construction costs I would urge the necessity for the exclusion of all sums paid for wearing parts from the final statement of cost. These should be segregated out, placed to a suspense account, and charged out as operating expense. The absurdity of charging everything required to complete the plant to construction has been exemplified in every instance, but the practice is still in general vogue. In one instance which came under my notice a ball-mill plant ran with an entire avoidance of liner expense, the principal running expenditure, for over five months. In another case an extensive conveyor plant ran for over two years with no belting expenditure charged out to working costs. As soon as the. original equipment deteriorates there is a rise in operating cost, due to the necessity for replacement. Such fluctuations are wholly inadmissible and convey false impressions, and I have always found that the last thing to be blamed is the system of book-keeping responsible for such anomalies.
In placing before the members of the Institute the foregoing remarks on the descriptive technology of gold and silver ore treatment I realize that the subject has been dealt with from a single viewpoint only—that of the practical metallurgist. Many of my contentions are admittedly argumentative and I trust that a critical discussion will tend to throw more light on the subject.
The technological study of the treatment of gold and silver ores has been largely responsible for the phenomenal strides which have marked the progress in this branch of metallurgy during recent years. In no other application of science to industry is system more imperative at every stage; correct formulation of result is only second in importance to efficiency of operation.
Processing method has been largely the result of individual initiative, aided by the work of technical journals and metallurgical societies.
There still remains a field for profitable discussion, and it is my intention to deal with some inconsistencies of expression common to a number of writers on the subject; and also to draw attention to the false impressions created by a statement which may, possibly be only unintentionally misleading. I also take the opportunity of tracing the source of both metallurgical and economic result; and to question the desirability of the methods usually employed in arriving at a final statement.
In the first place, I wish to deal with the question of extraction. This term may mean anything or nothing, and it is often used to mislead. In a report of a mining engineer as to the result of operations subsequent to the remodeling of a treatment plant prominence was given to the statement that the extraction had amounted to 91 per cent. On examination it was found that only a small percentage of the so-called extraction was calculated on actual bullion return, a proportion being figured from the gold left in the zinc boxes, and not recovered. In addition to this, a proportion of the extraction was said to be due to gold in concentrate. As there was no concentrating apparatus on the property and as concentration did not form a part of the scheme of treatment the matter was investigated, with the result that the concentrate, of lower gold content that the original ore, was found to be the result of an inefficient pumping system which failed to provide for the whole of the battery product delivered to it. The balance was stacked and, although itself of no marketable value, the gold content was allowed to form a basis of extraction figures.
The illustration may be extreme, but it will serve to show to what lengths an engineer may go in the mutilation of the most vital term in metallurgical technology.
The Descriptive Technology of Gold and Silver Metallurgy, by A. W. ALLEN, CAMBRIDGE, ENGLAND