Diamond Drilling and Core Logging Facts – Part One.

Most geoscientists know a lot about a few rocks, some know a lot about a lot of rocks, but only a privileged few train others how to interact meaningfully with all rocks – yet inexplicably their numbers are dwindling as ill-informed academes have closed several high-standing institutions, persuaded that ours is a ‘dirty science’ at a time when the furtherance and application of our knowledge and skills has never been more critical.

But why are the engineers so silent in our support? Could it simply be they work with numbers whilst we observe, describe and postulate? Yet without our data on ground conditions, neither the ore deposits we discover can be mined, nor can the mega-structures needed to sustain our civilization be built. Nonetheless, it has to be said that in general the accuracy of our subsurface maps and models is sub-standard and has led to the unnecessary removal or reinforcement of considerable volumes of rock, to the extent of jeopardizing project viability. Hence, perhaps it should not be surprising that the engineering profession could be largely insensitive to a global decrease in the number of graduate arm-wavers.

Therefore, with more than just a tangible threat to our future and that of the planet at stake, should we not enumerate the measurement uncertainties associated with our subsurface exploration and surveying techniques, using the theodolite of the underworld – the diamond drill – and research the protocols and improvements needed to surmount the difficulties revealed?

An outline of the uncertainties: We are good at locating and mapping features in outcrop, with satellite-assisted exactitude, but we are generally poor at specifying the measurement uncertainties with which our sub-surface data are located in 3D space – specifically that derived from core logging.

The coordinates of a feature logged in the core, such as a gold vein, are calculated using depth/azimuth/inclination measurements recorded in the driller’s depth log and the borehole survey. Importantly though, it is the precisions with which these three measurements are made that define the probability that this calculated location matches the true position of the feature in the subsurface, i.e. where the engineers will actually find it. Currently, all we can do is define the volume of a ‘displacement uncertainty ellipsoid’ within which there is only a 68% probability that the feature logged will be found, assuming all errors are randomly distributed.

At the site of the borehole collar the three orthogonal axes of the uncertainty ellipsoid – depth, inclination and azimuth – are each usually no more than a few centimeters long. Thereafter, it is our responsibility to determine how these errors increase with depth, as well as to quantify the effects of other random, systematic, or human errors found during QA/QC of the drilling and logging processes.

Uncertainties in borehole and drill-core depth measurement: During a recent online core logging workshop, a participant remarked that my discussion of the “standard” or “stacked” methods for depth registering drill core, reminded her of a candid camera TV show. In the episode, a carpenter was asked to install a bookshelf in an office that had been rigged with a moveable wall. His assistant measured and cut a board to fit, but it was too short. Saying nothing to disturb the office workers, the carpenter measured and cut the next plank himself – but now it was too long, making him swear aloud, then roar with laughter after he was shown the fake office wall and hidden cameras.

Geoscientists have long faced a similar frustration when logging the depth of features intersected in drill core because the rock cylinder cannot be broken flush with the cutting face of the drill bit at the end of each drill run. Hence, a stub of variable length is always left in-hole so that, although the end-of-run borehole depth measurement can be accurate to within a centimeter, the depth uncertainty of the associated core break, from where the core is depth registered, is greater by a variable amount.

In effect though, relative to the overall depth of most boreholes, this core break/borehole depth difference is a minor measurement uncertainty. However, the discrepancy gives rise to artificial core losses along with so-called “core gains” – which confusion can be exploited by drillers to avoid falling foul of a clause commonly incorporated in drilling contracts that penalizes core recoveries less than an agreed percentage, usually 95%.

This contract proviso can be dodged by simply practicing “borehole loss”. Thus, whenever bad ground or hasty drilling cause low core recoveries, the short lengths from individual runs are accumulated until a nominal ‘full drill run’ is achieved. In so doing, the logged borehole depth is shallower than the true borehole depth, and frequently by a considerable amount. However, the spin-off to the drill contractor is that the reported core loss is kept within contractual limits offsetting the revenue lost by not charging for the true borehole depth by the penalty that would otherwise be imposed for the true overall poor core recovery.

In effect, the driller moves the “office wall”, which practice has the added benefit that drill bits can be worn down sometimes to the point where no waterways remain, a significant saving. Furthermore, the drilling company builds a reputation for experienced drillers that consistently deliver ‘good core recoveries’. In the meantime the geotechnical quality of the core is seriously compromised since the weakest ground – that which is never seen in outcrop – remains hidden, washed away in the borehole cuttings.

Accordingly, the joke is squarely on the client. Without strong and informed QA/QC protocols, the driller is in full control of the measures needed to circumvent most stipulations aimed at improving core quality.

Uncertainties in surveying the borehole path:??????A good analogy for the ‘science’ of borehole surveying is the popular wire-hoop, hand-eye coordination game, where the contestant is challenged to move a metal hoop along a length of twisted wire without touching it and setting off an alarm.

In the borehole survey scenario, the twisted wire is the curvilinear borehole path (deviations greatly exaggerated) and the metal hoop equates to the edge of the 68% probability ‘measurement uncertainty envelope’ around the survey tool – with two important differences:

1.??????In the game the hoop diameter is fixed, but the survey tool’s uncertainty envelope expands with depth at a rate controlled by the precision of the instrument’s azimuth and inclination measurements.

2.??????Unlike the game, there is no alarm to alert the borehole surveyor if the borehole path touches or exits the measurement uncertainty envelope – because this event is currently impossible to detect.

Thus opportunities for human error abound, both accidental and deliberate. Errors such as poor instrument calibration, inaccurate collar coordinates, incorrect map grid used, poor winch depth calibration, in/out so-called ‘closure’ errors faked, and even copied surveys duplicated across holes, albeit slightly modified, are not infrequently found during Quality Assurance audits.

Consequently, tool performance and data quality are best evaluated by carrying out supervised repeat surveys and comparing the end-of-hole separation errors, which vital exercise presents a large field of research. Almost two decades ago a test was run to evaluate the magnitude of the problem – that had been identified even a decade earlier (P.G. Killeen, G.R. Bernius and C.J.Mwenifumbo,1995). The proposal and results of the Voorspoed exercise – a DeBeers sponsored, controlled experiment using a 389.60m pipe the path of which was precisely surveyed and the result compared with the internal surveys of different borehole survey instruments – are appended below.

It is left to the reader to decide what progress has been made since Voorspoed – aside from improvements made in the measurement capabilities of the various survey instruments and the enabling of low-cost multiple surveys of the borehole path – which advances do partially address the measurement uncertainties involved.

Uncertainties in finding the way up:

“Use with caution”. The fact that such a warning label should be applied to any geological map or model that is compiled and constructed with little, or no reference to the geological structures occurring in the ground explored, requires no explanation.

No opportunity to obtain these data from drill core should ever be missed, especially considering that outcrop mapping is biased towards exposures of hard, weather-resistant lithologies. Furthermore, the soft mineral infills of veins, joints and fractures, which can be pathfinders to mineralization, are also rarely seen on surface. Hence, by retrieving sample from all but the weakest ground, diamond drilling provides unique insights into the evolution of any terrane – essential for predicting local ground behaviour for either extractive or constructive purposes.

In recognition of the high value of such information, manufacturers are continually fine-tuning their core orientation systems to reduce, or eliminate random and systematic errors affecting the orientation of core into the “way up” of each segment recovered. This process recreates the attitude of the core relative to the geographic vertical plane, so that the dip and dip-direction of all planar structures that are intersected, as well as the plunge and trend of lineations, can be confidently measured and mapped in the sub-surface.

The gathering of such information therefore should be routine – but it is not, begging the question, “Why?”

Over fifteen years spent undertaking full Quality Assurance geotechnical audits using StereoCore? PhotoLog to photogrammetrically overlay logged data on processed images of the core, one of the most striking results is the overall high accuracies found for alpha/beta/theta angle measurements. This is pervasive over the several tens of thousand meters of core inspected from projects over much of Africa, the Philippines, Papua New Guinea, Argentina, Kazakhstan, etc., proving that in general the operation of the different goniometers available is well understood.

Hence, given the improved electronic instrumentation used to orient core, and taking note of the fact that the goniometry of the core is generally accurate, the conclusion has to be that human error is responsible for the poor performance of the technique – which deduction is encouraging as it means the solution to the problem is more-or-less immediately to hand, requiring just a few minor changes in the approach to mounting a drilling campaign.

?

To view the appendix please use this link:

https://www.dhirubhai.net/posts/john-orpen-47a27714_diamond-drilling-and-core-logging-activity-7018525583038951424-CalN?utm_source=share&utm_medium=member_desktop

Part Two will examine possible solutions, in the meantime please see also:

?

(PDF) Researching the StereoCore? Toolkit: Towards delivering maximum benefit from diamond drilling (researchgate.net)

(J.L. Orpen, D.L. Orpen and A..J.A. Bals, January 2023)