Search Documents

By R. F. Kovar, R. N. Torbin, R. E. Gundersen

I have to express my .surprise and must send a note of warning concerning the use of polyurethane foam as described in a paper by R.N. Torbin and R.F. Kovar, entitled "Innovative Shaft Lining Method," published in your February 1989 edition. In September, 1986, 177 miners lost their lives when they were exposed to fumes downwind from a fire in a tunnel lined with polyurethane foam. Whilst polyurethane exhibits reasonable fire retardent properties in small-scale tests, mass use in a tubular configuration, exacerbated by the furnace effect of cross-radiation, can lead to a rapid, uncontrollable conflagration. I do not wish these comments to detract from the engineering achievements of the authors but rather to sound a loud warning as to the application of the system. Any shaft designed for human access- be it for transport or shaft work - and any intake airway, should not consider the use of polyurethane foam. It only needs a slip with a cutting torch or an electric fault to initiate a disaster. Many countries have banned or restricted the use of polyurethane foam in the workings of a mine and maybe alternative foams (e.g. phenolics) can be used or developed

Jan 1, 1991

By G. Sanders

Discussion by G. Sanders The studies reported on in this paper were initiated to draw attention to the severe contamination problem in the Section 30 drilling program at Chimney Creek. The lithologic-subset sampling study reached a different conclusion from that presented in your paper, and I wish to comment on your subsequent analysis of the data and your conclusions. Request for more complete data In the section on subsampling, you mention that the subordinate lithologies were separated and sampled, yet only the dominantlithology gold value is plotted in Fig. 4. In a contamination study, the reader is interested in the assay values for the individual subsets. Please include a table of the subsample assay data in your reply. Also, please indicate which analytical methods were used to arrive at the gold values in the subsampling study. Turning barren rock into low-grade ore Figure 5 is very revealing and typical of all of the cross sections in Section 30. Note the long strings of low-grade mineralization spread out for hundreds of feet below the ore zones. There were some very high gold values found in certain contaminated fractions during the subset sampling. The conclusion, here, was that the distinctive, strongly-mineralized dolomite layer was probably loose and crumbly and continued to disintegrate during drilling. This caused salting of the unmineralized rock samples below. Missing the high-grade part of the ore body In your statistical analysis, you directly compare the reverse circulation assays to the diamond drill assays in Section 30. Two points argue against a direct comparison and suggest the differences are greater than the 3 % that you report. First, any core loss in a gold zone most likely means that the true gold values are greater. The drillers lost significant amounts of the clay-rich, Section-30 gold mineralization. Also, the initiated salt-mud system, an attempt to improve the core recovery, met with little success. Second, the practice of not sampling core geologically, but instead sampling on even 5-foot intervals, adds a deliberate dilution to the core assay values by including a portion of nonmineralized rock in the first and last samples of each high- grade intercept. The result is often a pair of low-grade assay values on either side of a high-grade gold zone. In reality, a high-grade gold zone has a very sharp assay wall that is often bounded by barren rock. This sampling method may make the diamond-drill core assays more like the reverse circulation values and may help explain the statistical similarities you found. However, it does not represent the true gold values in the high-grade parts of the deposit. You cannot deny that, by careful geological sampling of the drill core, higher and sharper assay values will be obtained. The low core recovery and the diamond-drill-core sampling method used act together to lower the diamond-core assay values. The 3% difference you found between the reverse-circulation and diamond-core assay values could be much larger when you consider what the true diamond-drill core values would be with optimum core recovery and a geologic sampling method for the core. Should statistics have been applied here? The statement "... that reverse circulation holes have overestimated the values of some ore zones and underestimated the values of others" (p. 345) is not correct. The subsampling confirmed what the cross sections hinted at in Section 30. Namely, beneath the high-grade zones, the reverse circulation holes created, by contamination, large intercepts of low-grade ore in regions of barren rock. Because the low-grade material was not there to begin with, this is not a process of overestimating low-grade mineralization. The next statement that "the average result is similar to that of the diamond drill holes" may apply to the data set numerically, but it is not true when viewed spatially on cross sections. Adjacent reverse circulation and diamond drill holes are almost impossible to correlate, high-grade zone values vary widely and many low-grade intercepts make no geologic sense. The subset sampling and cross sections presented in the first part of the paper show that the reverse circulation portion of the data set has some serious problems, as highlighted above, and should not have been dealt with statistically at all. Conclusion Each ore body is different, and each drilling method presents unique sampling problems. In this case, the diamond drill is the

Jan 1, 1994

M.S. Azun I have many objections to the content of the author's paper. Before discussing it, however, I would like to repeat the property of semivariogram function. Second order stationary properties of regionalized variables (ReV's) such as semivariogram function y(h) are perfectly known in geostatistics. Also, the kriging equations in the language of mathematical statistics using second order stationary properties are well understood. However, the way to use the sample (estimated) semivariogram function in any one of the kriging procedures is vague. The sample semivariogram function is given as follows: [ ] where N is the total number of samples, Xi is the sample value at the i - th location, Xi+h is the sample value at the i + h - th location, and h is the distance among the samples. An estimation variance of sample semivariogram function of first lag is smaller than that of higher order lag. The theoretical semivariogram function reaches the variance of samples asymptotically. But this is not easily observable because of the larger variation involved in the estimate of semivariogram function. In general, an estimation procedure is done for h = 0, 1, 2, up to the greatest integer less than N/2, even though sample semivariogram function can be computable through N-1. After estimating semivariogram function, the critical question of how to model sample semivariogram function arises. As seen in the above equation, sample semivariogram function is discrete and can be smoothed by the model being selected. Therefore, modeling of sample semivariogram function is the most important step in geostatistics. It not only smoothes a discrete function but also affects the results of the kriging procedure. When the only aim is to model the semivariogram function, which is the basic point of the author's paper, one can employ any fitting techniques, such as curve fitting, or any arbitrary functions, which are called submodels in the paper. The term "arbitrary function" is used rather than "submodel" because there is no basic understanding of developing them. The author suggests that the sum of those submodels can also be used for the modeling of sample semivariogram function. The combination of any arbitrary functions brings many problems instead of giving an insight of the domain structure considered. The author used two arbitrary functions and the nugget effect in response to sample semivariogram function (Fig. 10). For the same example, he stated that the parameters involved in the mixed arbitrary function model can be accepted when the discrepancy between sample semivariogram function and the model is small visually. For verifying the fitting behavior of any selected model, one should not be contented with the visual satisfactory. Some statistical measure such as goodness of fit has to be used. The author's practice is no more than an exercise in curve fitting without any fundamental understanding or conceptualization of the underlying physical mechanism. Furthermore, the selection of any model is not an easy task if the purpose is the search for the "best" response to the observed second order properties of ReV's. I suggest that the Markovian model (Azun, 1983), on the basis of a theoretical understanding of underlying mechanism, which gives more information about the occurrence of regionalized variables, is used to respond all properties of ReV's. There are a lot of problems for modeling of one-dimensional sample semivariogram function. Thus, it is not appropriate to go to higher order dimensional sample semivariogram function modeling. In the meantime, I would recommend that one can connect the values of standardized sample semivariogram function rather than simple values of semivariogram function in the two-dimensional estimation. The standardized values can be computed in dividing the semivariogram function value by the number of sample pairs involved in each lag regardless of the directions. In conclusion, geostatistics is an interdisciplinary area in mining that uses the principles of mathematical statistics. Thus, it should not violate any probabilistic and statistical rules. When Matheron was developing the theory of geostatistical study in the early years of geostatistics, many mining people had a reservation accepting the geostatistical tools. However, this does not mean that we, the geostatisticians, might try to convince those people using some "strange" tools or rules as some authors implied (Baafi and Kim, 1984). Instead, we have to develop and explain the geostatistical tools staying only in the framework of statistical concepts and properties. References Azun, M.S., 1983, "Stochastic Process Modeling of Spatially Distributed Geostatistical Data," Columbia University, Ph.D. Thesis. Baafi, E.Y., and Kim, Y.C., 1984, "Discussion - Comparison of Different Ore Reserve Estimation Methods Using Conditional Simulation," Mining Engineering, Vol. 36, No. 3, p. 280. Reply by J.M. Rendu The interactive method proposed by Rendu allows practitioners to develop semivariogram models that take into account not only the numerical information obtained by sampling, but also highly significant additional information that often cannot be quantified. The geology of the deposit - including hypotheses concerning its genesis, sampling methods, assaying methods, and mathematical methods used to calculate the semivariograms - all have an influence on the numerical results obtained and on how these results should be interpreted. If all the information concerning the spatial distribution of values in a mineral deposit was contained in the sample values, it could be argued that statistical techniques alone would produce optimum models. However, this is rarely, if ever, the case. Methods that allow the user to take into account his experience and his geologic understanding of the deposit should not be rejected for the sake of theoretical statistical purity.

Jan 11, 1985

By B. Beatty

Discussion by B. Beatty This is a response to M.J. Hrebar's paper, which tracks in detail the highly productive stripping capabilities of modern large bulldozers at two separate installations. It proves the old adage "economies of scale" and quantifies the advantage of being able to move large loads of dirt over comparatively short haul distances by sliding rather than carrying the load. This discussion examines whether a bulldozer is the most efficient way of using this geotechnical principle. Case 1 - Study eastern US: Contour coal stripping application The energy consumption for the Case 1 mining method (eastern US contour stripping operation) shows the D11 averaged 117 L (31 gal) of diesel fuel per scheduled hour, and produced 333.3 bank m3 (436 bank cu yd) per scheduled hour (0.351 L/m3). Assuming 1 L (0.26 gal) of diesel fuel contains 10.4 kW-hr of energy, the unit expenditure of fuel energy for Case 1 Study was 3.66 kWh/m3. Another way of shifting this material is by using an overburden slusher (OS), as shown in Fig. 1. The OS as shown in Fig. I. The OS is electrically powered and consists of a manned low wall winch (LW ), a remote controlled (radio telemetry) highwall winch (HW) and a hoe. The winches operate to pull the hoe back and forth across the pit. The hoe carries load in the forward direction ahead of a rear mounted blade and returns empty. A spreadsheet (Lotus 123) simulation package, BOSSIM, models the OS operation on a load by load basis over a wide range of pit configurations. The geometry shown in Fig. 2 of the paper was modeled using BOSSI M and the results are summarized in Table1. Table 1 shows a LW with 150 kW (200 hp) and a HW of 120 kW (160 hp) installed power, towing a 15-m3 (19-cu yd) hoe, would accomplish the Case 1 stripping task at about the same scheduled stripping rate (D11=333 m3/hr or 433 cu yd/ hour, OS=342 m3/hr or 447 cu yd/ hour). It was assumed the winches in this equipment have an electromechanical efficiency of 90%. This gives an average electrical energy expenditure rate of 151 kW-hr, and a unit energy expenditure level of 0.44 kWhr/m-3(0.57 cu yd/kWh)-an improvement of 8291%- over the D11 alternative. The energy comparison illustrates the importance of using more efficient electric power if possible, ahead of the internal combustion engine, and delivering power from a fixed ground position rather than through a moving track system. Another important energy observation is the comparative action of the dozer blade and the OS hoe. The former implement continuously digs and spills as it works the load forward. The OS hoe digs the load until full, the blade then lifts due to the contained load pressure and a shear plane forms under the mass, enabling it to slide forward in a force reduced - transporting mode. Economic comparison Table 2 operating cost per scheduled hour and unit cost comparison follows a format used in Tables 2 and 3 of Hrebar's paper and compares Case 1 Study with the OS alternative. Drill and blast costs are kept constant in both cases, although when larger OS units are considered, it is expected these costs will reduce because bigger hoe sizes will permit larger rocks to be transported efficiently.

Jan 1, 1995

L.R. Bower In regard to the paper by J.B. Gwiazda, it makes a highly technical approach to show that the µ factor used by designers of lemniscate-guided roof supports has never really been confirmed as a maximum and assumes that convergence is vertical. Also, the paper does not appear to take into account deflection of structures, which occurs when the lemniscate and base members are fully loaded to their maximum stress level, nor the front to back line of the support in relation to differential roof to floor movements caused by strata movements under pressure. It is not unusual for differential movements to be slightly diagonal to the line of the support, particularly in faulted areas and on gradient faces. The paper also does not take into account consolidation of fines immediately above and below the support. Generally speaking, any differential movement is from face to waste and under these conditions the µ of 0.3, which appears to be an international standard, has worked in practice. However, if the face end of the support is lower than the waste end, then the µ of 0.3 can be considerably increased, giving rise to the damage mentioned in the paper. The ideal design should aim for a slightly forward bias in the lemniscate guide so that the last increment of setting is toward the face, tending to close any fissures that may have developed during the support advance cycle. The support should also be fitted with positive set valves to ensure that a high setting load density is attained to minimize bed separation. As far as powered supports are concerned, convergence is irresistible and all powered supports converge at their rated yield load. A similar principle can be applied to the differential roof to floor movements to drastically reduce the very high forces that would otherwise be applied to the lemniscate structures and pins and that, in turn, are transferred to the base arrangement and floor loading. Any differential movements are usually catered for by the 0.3 µ factor or deflection of structures in the lemniscate guide arrangement and consolidation of the floor. The floor loading, due to differential movement, is in addition to the support convergence load and requires additional bearing area to avoid possible floor penetration. Some seven years ago, Fletcher Sutcliffe Wild Ltd. (FSW) introduced a lemniscate-guided shield support where the lemniscate linkage is connected to the roof bar through two horizontally converging rams to allow differential movement to take place above a given rated figure. This is a known force and can be guarded against, whereas with rigid connections the forces, as yet, are unconfirmed. By careful design, a horizontal force in excess of 6 MN (60 tons) opposes differential movements for a total ram loading of only 2.5 MN (25 tons), or 1.25 MN (12% tons) each. This principle can considerably reduce the length and weight of the support in comparison with a rigid pin-type structure ; also, the yield load rating can be increased without affecting the lemniscate forces. The graph shows the tensile and compressive forces in a lemniscate linkage of a support with and without hydrostore. These forces react into both the roof beam and base members and, as can be seen from the support height to linkage load graph, a considerable reduction in these reactions is gained by the use of the FSW patented hydrostore system. Floor loading is considerably reduced under maximum µ conditions, and by allowing the roof bar to move with the strata, some degree of improvement to strata control is achieved in line with the assumptions in the paper. In practice, these movements have only been in the region of a few millimeters, which, in turn, reflects on the improvements to strata control by the addition of positive set valves. Supports to this design of both 450- and 280-t (496-and 309-st) rating have been successfully used in the United Kingdom for several years, negotiating many faulted areas without one single reported need for repair or maintenance. This includes supports left unattended during the year-long strike, proving the reliability of the system.

Jan 8, 1986

By J. B. Gwiazda

S. Budirsky J.B. Gwiazda's article deals with an interesting problem that has not been studied thoroughly up to now. Gwiazda has proposed a technical solution that eliminates horizontal load imposed by the supports on the roof. That load is due to a discrepancy between the trajectories of the canopy and the overlying strata. It is true that the improved lemniscate guidance system proposed by Gwiazda could eliminate the horizontal load but, on the other hand, it would represent a higher cost of the supports, and it would complicate the already complicated mechanism of powered supports. That is why we must be convinced that the proposed improvement is actually needed. From that point of view, the following questions should be elucidated. As shown in Fig. 1, the point 0 travels along the curve a, which produces horizontal load on the roof in the direction of the coal face. As stated by Gwiazda, "such a high load is capable of destroying the roof above the support, causing rock debris to be scattered around the face." The results of our measurements carried out in coal mines of the Czechoslovak part of the Upper Silesian Coal Basin show that the convergence in the modern type of powered supports does not exceed several millimeters (not more than 10) per hour. Similar conclusions were drawn from field measurements performed in the Polish part of the Upper Silesian Coal Basin (see S. Romanowicz and H. Szopka in Proceedings, Scientific and Technical Symposium Simmex'85, Katowice, Poland, pp. 73-83). The small vertical closure of the supports also induces a small horizontal displacement of the canopy. In addition to this, the transmission of forces between the roof and the canopy is influenced by the layer of debris on it, i.e., the canopy slides without imposing a greater horizontal load on the roof. This was deduced from field measurements of Voest-Alpine F 4/4500 powered supports equipped with rams instead of the common front lemniscate links (denoted as 1 in Fig. 5). In these supports, the pressurizing of the space of piston rod (1a in Fig. 5) causes the rams to close, which induces horizontal load on the roof in a direction toward the coal face. After the setting of a support unit, the pressure in the piston rod space was found to drop, frequently to zero, because the small horizontal sliding between the canopy and the debris or inside the debris caused the horizontal reaction from the roof to fall to zero, although the debris were pressurized by vertical setting load. As a consequence, there was no horizontal load imposed by the canopy on the roof. A rise of the horizontal load was recorded only after the passage of the shearer and during lowering of the adjacent units. Thorough information on the results of the aforementioned field measurement is given in the paper, "Analysis of the performance of shield powered supports installed in a thick seam," published in Mechanizacja i automatyzacja gornictwa, 1982, No. 12, pp. 38-46. An English translation of the paper is available from the author. We have deduced from our observations that horizontal load imposed by the canopy on the roof in the direction toward the coal face is useful for roof control because it limits the displacement of the immediate roof toward the gob particularly during the rise working. We came to the conclusion that horizontal load on the roof toward the coal face should be induced on purpose. Could Mr. Gwiazda prove with the results of field measurements his contrary opinion? Has the small horizontal displacement of the canopy toward the face actually had an adverse effect on the roof? As shown in Fig. 2, during the setting of a support unit, the canopy imposes a horizontal load on the roof in the direction of the gob. In this case, I share Gwiazda's opinion that the horizontal load could have an adverse effect on the roof. Nevertheless, two important circumstances were neglected - the horizontal compliance of the supports and the interaction of adjacent units. The question of the compliance was analyzed precisely by I. Krumnacker in Gluckauf-Forschungshefte, 1984, No. 5, pp. 219-223. He found that due to the clearance between the hinge pins and the eyes of the sheild and due to the elasticity of the steel structure,

Jan 1, 1987

By J. B. Gwiazda

S.S. Peng The highlights of this paper are : • The shield supports with lemniscate links will exert a tangential force on the roof strata, causing unnecessary and harmful roof movements, thus damaging the roof structure. The principal source of the tangential force is the horizontal movement of the point O, the pivotal joint between the canopy and the caving shield. When the support height is changed during setting, the point O will move along a curve upward and rearward toward the gob. As a result, it produces a vertical movement, causing the roof to move up a distance of ?d, and a horizontal movement toward the gob, creating vertical fractures in the roof. • The major purpose for employing the lemniscate-link system is to protect the hydraulic legs against the action of the bending moments induced by the tangential force that causes deformation or destruction of the legs. • Based on these two points, the author suggested that, in order to eliminate the harmful effects of the lemniscate-link system and strengthen its advantageous features, a shield support with a new design of that mechanism has been developed. With the new design, the conventional rigid lemniscate links are replaced by hydraulic rams. When the pressure inside the rams reaches a certain level, the ram will yield.

Jan 12, 1986

By G. W. Heunisch

G.A. Carmody and L.B. Anderson We wish to comment on the paper by G.W. Heunisch (MINING ENGINEERING, January 1987), which raised several important points regarding treatment of acidic waste waters. This paper confirms, in part, tests carried out by Allyn Corp. Allyn is the manufacturer of the proprietary product Alka-Ment, one of the materials that Heunisch tested. The first point is the extreme sensitivity of the sedimentation rate on the nature of the treatment material used. This apparently results from differences in nucleation of the metal hydroxides, and is consistently higher for Alka-Ment. This decreased sludge volume is also observed when Alka-Ment is combined with flocculants, such as organic polymers (see Table 1). Our second point is to emphasize that Heunisch's work shows how dependent the sludge volume is on pH. It is common practice to overtreat AMD by raising pH to 9 or 10. This practice, in addition to being wasteful of material, actually decreases the density of the resulting sludge, probably by creating charged particles that do not readily coalesce. When treating with Alka-Ment, we recommend that the treatment be monitored by testing both pH and dissolved iron. For both tests, it is important to allow adequate time to elapse between treatment and measurement, so that dissolution of alkali, aerobic oxidation of iron, and precipitation of the metal hydroxides are nearly complete. Our tests have indicated that from 30 minutes to one hour is commonly required. Only in this way can a minimum pH be attained, with adequate treatment, denser sludge, and lower treatment rates achieved. Thirdly, Heunisch correctly recognizes that ASTM titration of Alka-Ment does not reflect the true treatment value available in a unit of Alka-Ment. Our work at Allyn indicates this is a most important consideration when evaluating various materials for the treatment of acidic waste waters. While we were unaware of Mr. Heunisch's work, we welcome comparisons between Alka-Ment and any competitive products readily available for field use. Allyn has hundreds of hours of laboratory and field data we are ready to share with interested parties treating acidic waste water.

Jan 9, 1987

By L. B. Anderson, G. A. Carmody

We wish to comment on the paper by G.W. Heunisch (MINING ENGINEERING, January 1987), which raised several important points regarding treatment of acidic waste waters. This paper confirms, in part, tests carried out by Allyn Corp. Allyn is the manufacturer of the proprietary product Alka-Ment, one of the materials that Heunisch tested. The first point is the extreme sensitivity of the sedimentation rate on the nature of the treatment material used. This apparently results from differences in nucleation of the metal hydroxides, and is consistently higher for Alka-Ment. This decreased sludge volume is also observed when Alka-Ment is combined with flocculants, such as organic polymers (see Table 1). Our second point is to emphasize that Heunisch's work shows how dependent the sludge volume is on pH. It is common practice to overtreat AMD by raising pH to 9 or 10. This practice, in addition to being wasteful of material, actually decreases the density of the resulting sludge, probably by creating charged particles that do not readily coalesce. When treating with Alka-Ment, we recommend that the treatment be monitored by testing both pH and dissolved iron. For both tests, it is important to allow adequate time to elapse between treatment and measurement, so that dissolution of alkali, aerobic oxidation of iron, and precipitation of the metal hydroxides are nearly complete. Our tests have indicated that from 30 minutes to one hour is commonly required. Only in this way can a minimum pH be attained, with adequate treatment, denser sludge, and lower treatment rates achieved. Thirdly, Heunisch correctly recognizes that ASTM titration of Alka-Ment does not reflect the true treatment value available in a unit of Alka-Ment. Our work at Allyn indicates this is a most important consideration when evaluating various materials for the treatment of acidic waste waters. While we were unaware of Mr. Heunisch's work, we welcome comparisons between Alka-Ment and any competitive products readily available for field use. Allyn has hundreds of hours of laboratory and field data we are ready to share with interested parties treating acidic waste water.

Jan 1, 1988

By R. C. Kirkman

Your interesting and well-written paper brings up the issue of the purpose or use of a mineral property appraisal, specifically the value of these evaluations when the quality and quantity of the mineral resource to be exploited have not been thoroughly investigated. I am not a practicing professional in regards to valuation of limestone investment opportunities. I simply wish to make a few generic observations about the appraisal process in order to stimulate meaningful thinking by appraisal professionals whose reports may be reviewed by professional investors and "laypersons" alike. It is my contention that the analysis given in your paper is very thorough, considering the data that was made available to you. But I would not like to be the person responsible for investment decision making without having a feel for the risk involved. Therefore, the reviewer must classify this type of analysis as relatively "subjective" because they have not received an indication of the inherent risk involved in exploiting a mineral resource. This risk can best be quantified by using probabilistic methods in conjunction with the quality and quantity of the mineral to be produced. Most appraisals passively ignore this problem. Specifically, you have accurately identified the various costs and values entering the economic appraisal process. These can usually be quantified with a great deal more certainty than the capability of the resource to provide 25 years of quality material. Also, in Cases A (West Virginia)and B (Alabama), you are fortunate enough to have previous operations data. What if there had never been production from these properties? How does a potential mineral property investor evaluate the resource capability to meet various projected financial outcomes without spending a good deal of up-front capital to drill and delineate enough quality resources for a 20-plus year operation. These questions pose a substantial risk to any potential mineral investor. Only probabilistic methods should be used to quantify this type of problem, short of drilling and metallurgical testing. I might be considered a "layperson," not being familiar

Jan 1, 1991

By C. W. Hunt

I wish to register strong disagreement with the choices of drilling sites chosen by the Continental Scientific Drilling Committee, while endorsing the stated objectives. Of special significance is the lack of any site involving the metals most needed today in the US, while 9 to 13 km of drilling are allocated to gold and silver districts and 11 to 17 km to base metal (mainly copper) districts. Why is nothing proposed for platinum group or chromite? Is this because the committee did not know of such a site? If so, I offer such a site in California (see Fig. 1). The mineralization shown in Table 1 occurs in a large peridotite body in western Tehama County, CA. Petrography has demonstrated unequivocal evidence of migrated PGE deposited in peridotite and chromite of hypabyssal nature. Active diapirism (due probably to serpentine buoyancy) above a reactivated mantle to surface conduit marks the position of the conduit. Young explosion breccia cemented loosely by volcanic glass and tuff demonstrates the significance of the conduit as the prime source of the various emissions. Deep drilling could result in major metals discoveries as well as shedding light on the grand questions of platinum group mineralization. What role water has played in this occurrence is not yet certain; but chlorine and asphaltite have been detected in and adjacent to native PGM, suggesting a major role for water. The explosion breccia suggests that this excellent site could overlie entrapped geothermal steam.

Jan 1, 1988

By C. Meyer, J. J. Eidel

C. W. Hunt I wish to register strong disagreement with the choices of drilling sites chosen by the Continental Scientific Drilling Committee, while endorsing the stated objectives. Of special significance is the lack of any site involving the metals most needed today in the US, while 9 to 13 km of drilling are allocated to gold and silver districts and 11 to 17 km to base metal (mainly copper) districts. Why is nothing proposed for platinum group or chromite? Is this because the committee did not know of such a site? If so, I offer such a site in California (see Fig. 1). The mineralization shown in Table 1 occurs in a large peridotite body in western Tehama County, CA. Petrography has demonstrated unequivocal evidence of migrated PGE deposited in peridotite and chromite of hypabyssal nature. Active diapirism (due probably to serpentine buoyancy) above a reactivated mantle to surface conduit marks the position of the conduit. Young explosion breccia cemented loosely by volcanic glass and tuff demonstrates the significance of the conduit as the prime source of the various emissions. Deep drilling could result in major metals discoveries as well as shedding light on the grand questions of platinum group mineralization. What role water has played in this occurrence is not yet certain; but chlorine and asphaltite have been detected in and adjacent to native PGM, suggesting a major role for water. The explosion breccia suggests that this excellent site could overlie entrapped geothermal steam.

Jan 6, 1987

By M. M. Botz, F. W. DeVries

Discussion by F.W. DeVries Comments on natural cyanide attenuation. The paper by Botz and Mudder in the November 2000 issue of Minerals and Metallurgical Processing had one disappointing aspect. Table 1 (p. 229) failed to include a significant species - neutral copper cyanide (CuCN). This compound has been most trouble-some in many plant-scale cyanide cleanup processes, includ- ing acidification-volatilization- reneutralization (AVR), hy- drogen peroxide cleanup and air/SO2 cleanup. It has also plagued many pilot efforts to use ion-exchange (IX) resin technology. This author elucidated the concept that, under some conditions, copper cyanide failed to precipitate out of treated effluents and appeared to exist as a filterable colloid that reported as copper and cyanide, with the latter measured as WAD (weak-acid-dissociable) cyanide in that analytical method. The concentrations measured were very near the levels of 0.2 ppm, often the regulatory limit for WAD cyanide. The copper concentration, using the 1 : 1 ratio for this com- pound, would then be more than 0.4 ppm, which could also be unacceptable. We discovered that it was unsafe to assume that the Cu(CN),- would surrender both cyanide ions simultaneously and allow cupric copper to be removed subsequently as, say, cupric ferrocyanide (Hatchett's Brown). This compound could be ignored as safely as the "iron blues," as essentially an inert solid, stable in geologic time if not exposed to sunlight. Under poorly controlled oxidation of the cyanocuprite ions (copper with four, three or two cyanides complexed to the copper), CuCN could well be formed and could not be assumed to precipitate out as a solid. The authors of the paper, by failing to include the CuCN species in their model, may have left users of the model prone to ignore the possible occurrence of the colloidal(?) form in the discharge, thus finding an effluent that failed to meet regulatory limits. (See addendum below.) This somewhat ephemeral material may well have been responsible for some large-scale failures of cyanide destruction technology to succeed at sites in North America and elsewhere. Two specific references alluding to this occurrence may be found: Griffiths et al., 1987, "The detoxification of gold-mill tailings with hydrogen peroxide," Journal of the South African Institute of Mining and Metallurgy, September, p. 282. DeVries, 1989, "Improved understanding in detoxification of cyanide leach liquors using peroxide" Rand01 Gold Conference, Sacramento CA, p. 15 1. It is advisable for workers in this area of cyanide removal or recycle to avoid assuming that CuCN will, with certainty, precipitate out of solution, or that the dicyanocuprite ion will part simultaneously with both cyanide ions and allow cupric copper to be dealt with by other reactions. One further piece of work may be incomplete. That is, a definitive study of the long-term behavior of CuCN that does precipitate when exposed to other compounds present in neutralized tailings. Addendum. The extremely complex reaction of cupric ion on cyanide- which, of course, results in oxidation of cyanide to cyanate, probably through the intermediate stage of forming cyanogen [(CN)2] from oxidation of HCN with its hydrolysis to cyanate and cyanide ion - involves two cyanide moieties and the reduction of the cupric ion back again to the cuprous state. This, then, can form the CuCN potential "solid" that was already the undesirable species. In cyanide cleanup, with either air/SO2 or hydrogen peroxide, copper is generally employed as the catalyst, so it is unclear what happens when the CU(CN)2-1 is ultimately exhausted. Certainly, the field evidence is quite clear that all

Jan 1, 2002

By E. M. Spokes

A.B. Rushton The work and resulting publications of the Committee on Underground Coal Mine Safety and The President's Commission on Coal were appropriate at the time of publication. The statistics cited in the publications appeared to be valid, and the actions proposed by the committee had merit. Although the actions proposed to achieve further improvements in safety may have been worthy of repeating, to associate statistics (Table 1) that are seven to eight years old with company names is bad. Statistics are readily available from the Mine Safety and Health Administration (MSHA) and are usually within three months of being up to date. A review of the most recent Coal Controlling Co. Report will reveal significant changes in the rankings. The more recent statistics should have been cited, or at least compared with those used in the article. Westmoreland Coal Co. is listed as having a 21.1 disabling injury rate for the period (1/78-12/80). The disabling injury rate for the Company for the year ending 1985 was 7.4 for the underground mines and 6.25 for the total facilities. Through March 1986, the disabling rate for the underground mines was 2.86. The management and employees of Westmoreland Coal Co. are committed to safety. The improvements achieved over the past five years attest to that commitment. We are justifiably proud of the achievement and dislike being misrepresented as in the April 1986 article. Reply by E.M. Spokes I regret that the lapse of time since the original work gives a picture of safety practice that is no longer true today. That Westmoreland has reduced its disabling injury rate by 65% to 7.4 in 1985 is most commendable, and I congratulate you for this accomplishment. It is typical of the progress sought by the National Academy of Science and National Research Council when they funded the original study and published "Toward Safer Underground Coal Mines" in 1982. If a significant number of companies do as well as Westmoreland, the report will have served its purpose. Since the paper was primarily a condensation of the original report, it would not have been appropriate to seek access to later statistics. The purpose in seeking its publication at the time of presentation was to reach those who might have missed the original report. The fact that publication was delayed has not voided that purpose and may have reminded some who were aware of the original that its recommendations still apply to improvement of underground coal mine safety.

Jan 6, 1986

By S. D. Thompson, L. Adler

J.R. Sturgul Thompson and Adler (1988) present a new simulator for use in designing conveyor belt systems for underground mines. Their model uses a large Fortran program to do a Monte Carlo simulation. Their approach is similar to that done by other researchers, such as Stanford (1965) and Newhart (1977). All Fortran based simulation models use an imaginary clock that is advanced a time increment. At each advance of time increment the program has to determine what is to take place in the system. Since many things can take place at different places in the system, this approach requires that a large amount of logic is needed at each step of the program. This, then, necessitates a great amount of jumping between different portions of the program. In fact, the Fortran program, CMBCS, takes up to 15 to 20 minutes to run on the mainframe and several hours for the PC. Such execution times are typical of Fortran based simulation studies. There have been other examples of Fortran based simulation studies in mining. Cross and Williamson (1969) presented a model of a working open-pit copper mine. Fosinato (1982) presented a model of the bucketwheel excavators for the large open-pit mines in Southern Victoria in Australia. These excavators discharge into a complex conveyor system, which can be thought of as being similar, but more complicated than the one modeled by CMBCS. All previous simulation studies using Fortran as the source code have several things in common: they all are very lengthy, often being of the order of 10,000 lines of code and take a great deal of CPU time to execute. The program of Fosinato takes more than 4.5 hours to run on a large mainframe. Also not mentioned by Thompson and Adler is the actual time it took to write the Fortran code. This, too, can be considerable, as Fosinato's program took several man years to write and de-bug. In order to change it, as the parameters change or as the mine increases, further large amount of time is needed. An alternate to such Fortran based programs is provided by using a special computer language known as General Purpose Simulation System (GPSS). This language is designed to solve a wide variety of discrete event simulation problems and appears to be particularly suited for mining situations as shown by Sturgul and Yi (1987). Although GPSS was introduced as far back as 1961, it has not been used in mining to any great extent. However, the original form was quite crude and inefficient compared to modern versions. Perhaps with the increasing popularity and availability of these recent versions, this will change. Two such versions are GPSS/H for the mainframe and GPSS/PC for the personal computer. GPSS/H allows for interactive usage, such as allowing input data to be entered via a menu. GPSS/PC has excellent graphical capabilities and one can actually view the simulation in progress. In addition, one can halt the simulation and change values of various parameters. If a particular portion of the simulation is in progress via a concept known as "microwindows." These new versions have greatly increased the potential for using the language for mining situations. Mining applications of GPSS have been given by Sturgul and Yi (1987), Sturgul and Harrison (1987), Sturgul (1987), and Sturgul and Singhal (1988). Schriber (1987) presents a summary of the recent advances in GPSS.

Jan 12, 1988

By J. R. Sturgul

Thompson and Adler (1988) present a new simulator for use in designing conveyor belt systems for underground mines. Their model uses a large Fortran program to do a Monte Carlo simulation. Their approach is similar to that done by other researchers, such as Stanford (1965) and Newhart (1977). All Fortran based simulation models use an imaginary clock that is advanced a time increment. At each advance of time increment the program has to determine what is to take place in the system. Since many things can take place at different places in the system, this approach requires that a large amount of logic is needed at each step of the program. This, then, necessitates a great amount of jumping between different portions of the program. In fact, the Fortran program, CMBCS, takes up to 15 to 20 minutes to run on the mainframe and several hours for the PC. Such execution times are typical of Fortran based simulation studies. There have been other examples of Fortran based simulation studies in mining. Cross and Williamson (1969) presented a model of a working open-pit copper mine. Fosinato (1982) presented a model of the bucketwheel excavators for the large open-pit mines in Southern Victoria in Australia. These excavators discharge into a complex conveyor system, which can be thought of as being similar, but more complicated than the one modeled by CMBCS. All previous simulation studies using Fortran as the source code have several things in common: they all are very lengthy, often being of the order of 10,000 lines of code and take a great deal of CPU time to execute. The program of Fosinato takes more than 4.5 hours to run on a large mainframe. Also not mentioned by Thompson and Adler is the actual time it took to write the Fortran code. This, too, can be considerable, as Fosinato's program took several man years to write and de-bug. In order to change it, as the parameters change or as the mine increases, further large amount of time is needed. An alternate to such Fortran based programs is provided by using a special computer language known as General Purpose Simulation System (GPSS). This language is designed to solve a wide variety of discrete event simulation problems and appears to be particularly suited for mining situations as shown by Sturgul and Yi (1 987). Although GPSS was introduced as far back as 1961, it has not been used in mining to any great extent. However, the original form was quite crude and inefficient compared to modem versions. Perhaps with the increasing popularity and availability of these recent versions, this will change. Two such versions are GPSS/H for the mainframe and GPSS/PC for the personal computer. GPSS/H allows for interactive usage, such as allowing input data to be entered via a menu. GPSS/PC has excellent graphical capabilities and one can actually view the simulation in progress. In addition, one can halt the simulation and change values of various parameters. If a particular portion of the simulation is in progress via a concept known as "microwindows." These new versions have greatly increased the potential for using the language for mining situations. Mining applications of GPSS have been given by Sturgul and Yi (1987), Sturgul and Harrison (1987), Sturgul (1987), and Sturgul and Singhal (1988). Schriber (1987) presents a summary of the recent advances in GPSS.

Jan 1, 1987

By A. G. Mayton, L. Adler

The paper under discussion provides an excellent and concise study of compartment ergonomics for thin seam equipment. However, in accomplishing this, some very important operational aspects were ignored. Since these are likely to be determinant, the authors' response is solicited. The novel shuttle car with its cross-car, end-mounted compartment is operationally very different from the current one. The latter can be considered a highly mobile extension of the continuous miner or loader, with a pass-through conveyor, dump boom, and extra storage. As such, the vehicle's kinematics are the simple "I," or shuttle pattern, bent for going around pillars, with an intermediate stop at the "change-out point" (Adler, 1986 ). The action of the cable reel adjusts well to this pattern. The proposed shuttle car's motion either superimposes a "Y," time is lost (5 seconds) for each cycle, space is needed, line of sight is compromised, and a cable anchor is probably required at the dump. A chute is almost prohibitive in construction time and cost. The lack of a dump boom further curtails flexibility in the proposal. The value of a standard hand and opposite standard hand configuration also needs consideration. Many front-end loaders have a rocker (overshot) capability that is designed to achieve an "I” pattern. And crawler loaders sometimes use a side dumping bucket variant for the same purpose. These examples indicate the pattern's importance. For underground, thin seam application, the ram car has become popular. This is an articulated truck, battery-powered, and with an ejector dump. Its operational patterns would be similar to the novel shuttle car, except without the loading flexibility provided by a conveyor. Since the cab location of the proposed shuttle car largely obviates some of the major advantages of the present shuttle car, the question arises whether this configuration could be adapted to the current ram car and scoop

Jan 1, 1991

By Y. C. Kim, F. Szidarovszky

G.M. Philip and D.F. Watson The article cited above reflects the difficulty that is often encountered in interpreting different aspects of kriging. Baafi et al. write : "Due to negative kriging weights, negative block grades may be obtained in some extreme situations, which makes no practical sense." Their solution to this particular problem is to constrain the kriged estimates to data values that have positive weights, i.e., reduce the subset size contributing to the estimation. As Journel (1986, p. 133) has already observed, "Negative weights are no evil." Indeed, they are an ineluctable result of the Lagrangian multipliers in the system of simultaneous equations. The fact that data distant from the estimation point may be given negative weights simply means that the influence of such distant values must be negated because their influence is contradictory to closer values (as illustrated by Baafi et al. in their Figs. 3 and 4). Of course negative kriging weights applied to distant grade values will not in themselves induce negative grade at some central interpolation point. Negative grade values, rather, seem to be a particular artifact of automatic methods of estimation involving accu¬mulations (Philip and Watson, 1985a, 1985b, 1987b). The elimination of negative weights, suggested by Baafi et al., is likely to exacerbate a quite different problem of kriging and piecewise spline interpolation discussed elsewhere (Philip and Watson, 1987a). In interpolating large data sets of scattered observations, changing local subsets or neighborhoods must be selected automatically. Because of the mathematical formulation of kriging and spline interpolation, discontinuities then occur between closely-spaced estimation points as the subset used in the estimation changes. If this subset is reduced in number and size, the discontinuities will be amplified and more common. ? References Journel, A.G., 1986, "Geostatistics: models and tools for Earth Sciences," Journal of Mathematical Geology, Vol. 18, No. 1, pp. 119-139. Philip, G.M., and Watson, D.F., 1985a, "Theoretical aspects of grade-tonnage calculations," International Journal of Mining and Geological Engineering, Vol. 3, pp. 149-154. Philip, G.M., and Watson, D.F., 1985b, "A deterministic approach to computing ore reserves. I, Grade Estimation," Earth Resources Foundation, Occas. Paper 2, The University of Sydney, 24 pp. Philip, G.M., and Watson, D.F., 1987a, "Neighborhood discontinuities in bivariate interpolation of scattered observations," Mathematical Geology, Vol. 18, No. 1, pp. 69-74. Philip, G.M., and Watson, D.F., 1987b, "Estimating interactions of topological functions," Mathematical Geology, Vol. 18, (at press). Reply by Y. C. Kim Before replying to the comments by Philip and Watson, it is desirable to establish the original perspective of our paper, which was intended primarily for the practitioners of mining who possess a working knowledge of ordinary kriging. The paper tried to convey the following practical points: (1) there are circumstances in which one observes negative weights, (2) illustration of circumstances under which one can expect negative weights, (3) should one obtain a negative estimated grade, one can resort to elimination of negative weights using the technique outlined in the paper, and (4) one could normally select only the first layer of surrounding assays to estimate the block grade. Under such a perspective, most of the comments by Philip and Watson are irrelevant and misdirected. For example, the practitioners do not find it difficult to "interpret" the results of ordinary kriging, as evidenced by the fact that they have been using the results for nearly a decade. Nor are the practitioners concerned with discontinuities in estimated block grades due to changing neighborhood of assays during kriging. What matters to the practitioners is the fact that kriging does provide the best estimate among all the estimation tools available to them as of today. ?

Jan 2, 1987

By W. E. L. Minter

Discussion by E.S. Cheney Minter (1990) has provided an excellent summary of the sedimentology and theories of origin of the Witwatersrand paleoplacer Au/U deposits. However, recent stratigraphic, metamorphic and structural -tectonic research is radically changing the concepts about the Witwatersrand that now occur in most textbooks, review articles and North American minds. Presently, these concepts are biased by the past sedimentological research so well described by Minter. Specifically, the Witwatersrand Supergroup is Archean, does not occupy a depositional basin and is significantly metamorphosed and deformed. Thus, many of its sedimentary features, including the mineralogy of its ores, have been modified. The importance of the Witwatersrand ores cannot be overestimated. Since their discovery in 1886, they have produced 52% of the gold mined in all of human history (Pretorius, 1989). Setting Only one depositional item needs to be added to the excellent

Jan 1, 1992

By W. E. L. Minter

Discussion by R.W. Hutchinson Minter's excellent recent paper (1990) on the great paleoplacer gold deposits of the Witwatersrand and its subsequent discussion (Cheney, 1991) and reply (Minter, 1991) provide an informative, interesting and up-to-date review of the geology of these unique ores and some long enigmatic questions regarding their origin. Clearly summarized by Minter (1990) are the compelling sedimentologic and stratigraphic relationships indicating that the Central Rand Group conglomerates are ancient placers and documenting the fluvio-deltaic conditions of their transport and deposition. As Minter (1991) indicates, Cheney's discussion broadens Minter's 1990 coverage of the subject by considering higher and lower stratigraphic units and by adding detailed comments concerning recently published information about metamorphic and age relationships. Cheney's discussion and Minter's reply also readdress the old controversy of epigenetic hydrothermal vs. syndepositional placer origin for these deposits. This controversy has been resurrected by: • recent recognition of hydrothermally-altered granites (HAGS) in the older Archean basement rocks that lie unconformably beneath the Dominion Group rocks north and west of the Witwatersrand depository; • recognition of greenschist metamorphic assemblages in the Witwatersrand strata; and • the presence of interpreted epigenetic mineral assemblages and textures in some of the deposits. These relationships are briefly discussed by Cheney who suggests they may result from later, superimposed metamorphism and deformation and do not prove that the gold was introduced by epigenetic/metamorphogenic processes. In his paper, Minter (1990) cited the 1986 abstract of a subsequently published paper by Hutchinson and Viljoen (1988) who also considered many of these aspects and attempted to integrate them all in abroad genetic hypothesis. In addition, they emphasized the additional important problem of determining the source for the Witwatersrand gold. Although this issue was partially addressed by Reimer (1984), Mossman and Dyer (1985) and Reimer and Mossman (1990), the question has been generally under emphasized in earlier work and was, therefore, not stressed by Minter and Cheney. Reimer (1985) also emphasized the importance of weathering of the basin's Archean hinterland and its affects on Witwatersrand sedimentation. Since the epigenetic hydrothermal hypothesis was essentially discarded in favor of placerist origin, the gold source for these great deposits has generally been attributed to the weathering of lodes in greenstones or granites of the uplifted, older Archean hinterland to the north and west. However, Hutchinson & Viljoen (1988) present evidence that this explanation is inadequate, as well as data in support of an alternative explanation. They suggest that the gold and accompanying abundant, clearly detrital and auriferous pyrite in the conglomerate reefs were derived by erosional degradation of auriferous-pyritic exhalite previously deposited, by shallow marine discharge of hydrothermal fluids, along the fault-bound northwestern margins of the Witwatersrand depository. The auriferous-pyritic exhalites were generated by fluid-rock reactions in the subsurface, particularly in the highly-altered lavas of the Dominion Group but also in adjacent basement granitic rocks. The process envisaged is comparable to that observed in sea floor hydrothermal systems today, although, in this case, it occurred at the margin of a rapidly subsiding, shallow marine or continental basin. Important evidence for this explanation is the anomalous gold content of the ferruginous, shaley strata occurring throughout the Witwatersrand Supergroup and including the distinctive contorted bed that is an iron formation. These layers have been interpreted (Hutchinson and Viljoen, 1988 and 1990) to represent the more distal, finer-grained, mixed chemical/clastic sedimentary strata. Thus, they are the basinal equivalents of the proximal, auriferous-pyritic exhalites that were deposited along the basin's margin and subsequently reworked by sedimentary processes. This gold source explanation resolves many of the enigmatic questions regarding Witwatersrand geology and the reawakened controversy between proponents of hydrothermal and placerist genetic hypotheses. The theory invokes a hydrothermal origin for the gold in the basin-margin pyritic exhalites, thus explaining the visible hydrothermal characteristics (Figs. 1 and 2). It also provides compelling evidence for placer origin by encompassing all of the fluvio-deltaic processes of clastic sedimentary transport and deposition that are clearly summarized by Minter. These characteristics result from the degradational reworking of the pyritic exhalites along the depository's margins. Fluvial reworking, transport and deltaic redeposition of this detritus, along with additional detritus carried into the depository from the hinterland, then formed the conglomerate reefs. Other enigmatic aspects of Witwatersrand geology include: • the rarity of magnetite and ilmenite (the two most common heavy minerals in placer deposits) and the contrary abundance of leucoxene in the conglomerate reefs; • the presence of differing varieties of pyrite (authigenic, allogenic, compact and porous) all of which are auriferous; and • the presence of well-rounded pebbles, even cobbles, of pyrite that, in view of the brittleness of the mineral, could not have been transported great distances from a hinterland.

Jan 1, 1993