The advantages of obtaining core samples that provide a visual record of the rocks and ores sampled and at the same time furnish samples for assay have been pointed out under the caption Exploration by Drilling. The diamond drill, which employs either carbons or bort as cutting media, has been the type of core drill most widely employed in metal mining, largely because it will cut the hardest rocks; it can be used to drill at any angle from horizontal to vertical, up or down.
The shot drill has been used chiefly in obtaining cores of formations at dam sites, testing building foundations and the like, and sampling flat deposits of industrial minerals, where holes drilled vertically downward will provide the desired information. The shot drill has been used extensively together with chum and diamond drills for sampling copper deposits in northern Rhodesia.
With 100-percent core recovery the core will comprise an accurate sample of the ore penetrated. This does not imply that its assay value will be that of the ore at 5 feet or other distance from the hole; and, as with other types of samples, the results are reliable for estimation of ore grades only as cores are obtained at enough properly distributed points in the ore body to give an average value that approximates the average of the entire ore body. As the percentage of core recovery diminishes, the diameter of core varies, and other irregularities occur, the assay results become less reliable.
In soft or fractured formations, pieces of core are likely to break off and be ground by the sludge; the core often will be appreciably smaller in diameter in soft layers of rock than in hard layers, and therefore the core sample will contain proportionately less soft material than hard, whereas the sludge will contain more. If, as often happens, the softer material is appreciably higher or lower in grade than the harder material, the core assay will be correspondingly too low or too high. In soft or broken ground, a double core barrel usually will obtain a higher core recovery than a single core barrel.
In soft ground, the outside diameter of the hole may be larger than in hard ground, and the volume of the sludge sample will therefore be greater (assuming that all of it is caught).
When core recovery is low it usually becomes necessary to assay both the core and the sludge from each sample depth and to combine the results in order to approximate the actual value of the material cut by the drill. Mathematical methods based upon the relative volumes or weights of core and sludge and the assay values have been worked out for combining the results of core and sludge assays. In the opinion of the authors, some of these methods are more refined mathematically than usually is warranted by the approximations of some of the factors entering into the equations. Thus, the total overcut of the drill in soft ground would be determinable from the weight and volume of the sludge, but there is no way of determining whether the overcut is uniform over the full length of the sample or is principally from one small section that may be very much higher or lower in grade than the rest. Furthermore, there is no definite way of determining whether all the sludge has been recovered or whether some of it has been lost in cracks and crevices in the wall of the hole. If some has definitely been lost, the question remains as to whether the lost material carried a higher or lower percentage of valuable material than did the part recovered in the sample.
In diamond drilling, the drilling water is circulated down the drill rods, out the bottom, and up along the outside of the rods. In some recent models this direction of circulation has been reversed to impart maximum velocity to the rising current that carries the cuttings. This causes the sludge to be classified, the lighter particles rising readily to the surface with the water and the heavier particles lagging behind and perhaps coming to rest on a rough spot or in a crevice in the wall of the hole. With churn drilling, on the other hand, the sample becomes well mixed, and for this reason churn-drill sludge samples may be more reliable than core-drill sludge samples.
In view of the above uncertainties, which constitute only a few of the variables that must be considered, it is apparent that reliable interpretation of results depends on good judgment based on close observation of all abnormal occurrences in the taking of each sample, especially when only a small amount of core is recovered. A method of combining core and sludge assays that gives good results in one kind of ground might not apply to ground in which conditions are different. In ore of uniform grade and with even distribution of ore minerals throughout the gangue matrix, the sampling errors due to loss of core or caving obviously are not as serious as where mineralization is erratic and the grade varies widely from one point to another. Methods of combining core and sludge assays will be discussed briefly in a later section of this bulletin, but the foregoing has been introduced at this point to emphasize the necessity for care in collecting the sludge, especially where core recovery is poor, and in noting and logging everything unusual that may take place during the drilling and removal of each sample from the hole, so as to throw light upon the reliability of the assays later on.
Prior to drilling out of a sample the hole must be pumped clean of cuttings from previous runs and the depth at which drilling is resumed must be checked to make sure the bit is on the bottom of the hole. If the hole is caving, it may be necessary to case it to a point below the caving section. (If caving is anticipated, it is well to start with a large bit to permit casing and then to drill through the casing with a smaller bit.) It is then necessary to check the amount of water issuing from the hole with that being pumped in to determine whether an appreciable quantity of water is being lost. In fractured ground, very little of the water, or even none, may return, in which case it becomes necessary to cement the hole to close the fissures in the rock responsible for the loss of water. In Northern Rhodesia, holes were cemented if loss of water exceeded 5 percent.
When drilling is begun the return water is diverted to suitable boxes, launders, or barrels, where the cuttings are caught and the clear water is allowed to overflow. The simplest and perhaps most satisfactory device consists merely of a number of barrels, each provided with a number of bungs at different heights. As soon as the first barrel is filled, the sludge is diverted to the next, and so on. When all the cuttings have settled, the clear water is drawn off through the bungs and the sludge is collected in one sample for each run. A more elaborate arrangement, which was used with success in Northern Rhodesia, is shown in figure 5. This is suitable for surface drilling and deep holes where the cost of its installation is warranted. A simpler device and one widely used is a box 12 to 18 inches wide, 6 to 10 inches deep, and 4 to 10 feet long, divided transversely by two or more baffle plates, behind which the sludge settles and is caught. The discharge from the casinghead is diverted into the upper end of the box and the water overflows the lower end.
When the desired depth has been drilled the drill is stopped, and full pressure is put on the pump (when a pump is used instead of gravity head) to obtain as rapid circulation of water as possible in order to remove not only the light and fine material but the heaviest cuttings as well. The water is circulated until it comes up clear and free from cuttings. The depth to the bottom of the hole is checked on the drill rods and the core is broken off and pulled. The pieces of core are removed from the core barrel and arranged in a core box in the order in which they were taken from the hole. The top and bottom depths of the sample are then marked on the box opposite the corresponding ends of the cores. When core recovery is incomplete the length of the core will be less than the depth drilled. The pieces of core are measured along their centers and the sum of their
lengths will approximate the length of core recovered. In solid, hard ground, the measured lengths may be accurate.
The cores should be inspected and the mineralogic and geologic information they reveal should be logged carefully by a competent geologist or engineer.
Is Diamond Drill Sampling Accurate
The foregoing makes it apparent that the accuracy of diamond-drill samples varies greatly and depends on a number of factors. Where the ore is fairly uniform in grade, core recovery is high, and cores are uniform in diameter core assays in general are very reliable. Where core recovery is low and assay results correspondingly doubtful the core obtained may nevertheless furnish valuable structural and other geological information. Where mineralization is erratic, as is true of most gold ores, core assays are often only indicative and are not trustworthy as a basis for estimating the grade of ore reserves.
Under most conditions, diamond-drill sampling is at least as accurate as other methods of drill sampling, and under conditions favoring use of the diamond drill, it is the most nearly accurate drilling method.
In the Southeastern Missouri district, where galena is finely disseminated in limestone but favors shaly layers and is much more friable than the limestone, diamond-drill samples are accurate. One company in this district assays only the sludge and retains the cores for reference, whereas another company bases its estimates on core assays and only assays the sludge as a check. Both report reliable results, which illustrates the point previously made that when the ore mineral is finely disseminated in the gangue and the ore is of low or moderately low grade, the reliability of samples is not seriously affected by disproportionate amounts of sludge and core, core and sludge assays being virtually the same.