Presentation at the PDAC 99 Conference in Toronto, Canada

The use of geophysical methods in diamond exploration is approaching its fiftieth anniversary. Hundreds of kimberlite bodies in Russia, and probably thousands worldwide, have been discovered using these methods. The substantial successes and long history of application allow us to analyse the available data, recognise the achievements of geophysical technologies, specifically identify, highlight and study shortcomings and errors, and, against this background, attempt to envisage the future of diamond exploration geophysics: diamond geophysical technologies of the 21st century.

A. The Nakyn precedent

For more than 5 years, a new kimberlite field, the Nakyn (or Middle Markha) field in Yakutia, Russia, has been known in the geology of diamond deposits.

Its first pipe was discovered almost accidentally during reconnaissance drilling on a 20 km by 2 km grid.

The second kimberlite pipe was discovered near the first while checking a low-contrast magnetic anomaly. As subsequently established, this anomaly was associated not with kimberlites but with traps in the kimberlite host rocks. Thus, in this case too, the pipe was discovered almost accidentally.

This year (1999), another pipe has already been discovered near the first two. It was intersected by a borehole during drilling on a 400×400 m grid in the area around the first pipe. The drilling was undertaken to confirm the absence of ore in an area of potential industrial construction.

All these kimberlite pipes have exceptionally high commercial concentrations of high-quality diamonds, and none produces sufficiently distinct anomalies in conventional geophysical fields.

Modern high-accuracy magnetic and electromagnetic surveys have been conducted in this area over the past 2–3 years, and the identified anomalies were tested by drilling, with negative results for diamond exploration. Significantly, the anomalies from the pipes listed above were not recommended for drilling.

These data allow us to speak of the Nakyn field precedent: the accidental discovery of a new kimberlite field in which all the kimberlite bodies are highly valuable but are expressed very indistinctly in data from conventional modern geophysical methods.

It seems important to analyse the nature of this precedent and consider its significance for the practical use of geophysical methods in the exploration for commercially valuable diamond deposits.

In view of the foregoing, it is appropriate to seek answers to the following three questions:

Is the Nakyn precedent an exception to the rule that modern conventional geophysical technologies are highly effective in exploring for primary diamond deposits, and is it an exception that can be disregarded?

If not, that is, if this exception cannot be disregarded, what is the probable extent of highly diamondiferous magmatism whose products are virtually unrepresented in standard geophysical (and possibly mineralogical) fields?

And if this extent proves significant, what might be the optimal technologies for exploring for deposits of this type: highly diamondiferous deposits with very low-contrast expressions in geophysical fields?

To formulate substantiated answers to these questions, let us consider some theoretical premises and practical examples of diamond exploration in Yakutia.

First, let us assess how well we know the indicator properties of diamondiferous and highly diamondiferous kimberlites, that is, the properties we use to search for them. To this end, let us consider the theoretical problem of a non-random sample.

B. The problem of a non-random sample and implications of its solution

Suppose that, before exploration begins, a particular kimberlite field actually contains a total of N kimberlite bodies. We shall call this collection of pipes the population, because it actually exists independently of us or the detection method.

It is known that pipe magnetization in well-studied kimberlite fields follows a statistical distribution close to exponential (Fig. 1). In other words, strongly magnetic pipes are relatively rare, whereas weakly magnetic pipes are very common. This distribution has the same form in all well-studied kimberlite fields, but its mean—the principal parameter of the exponential distribution—is specific to each field.

Suppose that exploration within this field is carried out by magnetometry. Clearly, under otherwise favourable conditions, all highly magnetic and large pipes will be detected by the method, whereas some or most small and very weakly magnetic pipes will be missed.

In this case, the effectiveness of the method can be assessed using the probabilistic characteristics of the exploration filter (Fig. 2).

Here the horizontal axis shows variation in the magnetic moment of individual pipes, and the vertical axis shows their probability of detection by magnetometry for a particular measurement accuracy and exploration grid.

Magnetic surveying will detect some pipes and miss others. The detected pipes may be called a non-random sample, and the missed pipes the residual population.

Figure 3 shows the non-random sample and residual population for an exponential distribution of the indicator parameter.

Analysis of the figure leads to the following conclusions:

A) The distribution of petromagnetic characteristics in the non-random sample, i.e. for the detected objects, is described quite satisfactorily by a lognormal distribution. The relative proportion of weakly magnetic pipes here is very substantially lower than in the population.

B) The distribution of petromagnetic characteristics in the residual population, i.e. pipes not detected during the exploration undertaken, is close to exponential, with a pronounced and very substantial predominance of weakly magnetic diatremes.

The principal conclusion from this example, however, is that the characteristics of the indicator properties of known, discovered kimberlite bodies, pipe clusters and individual kimberlite fields cannot be applied to bodies, clusters and fields that remain undiscovered and are targets of renewed exploration in previously explored areas.

What, then, do the actual data show about the mean magnetic characteristics of kimberlites in clusters and fields of kimberlite pipes?

Figure No. 4 presents data showing variation in the mean magnetic susceptibility characteristics of kimberlites in 15 kimberlite fields in Yakutia (excluding the Nakyn field). The distribution is distinctly lognormal. Weakly magnetic kimberlite fields are very rare. This allows the data in the figure to be regarded as a non-random sample of kimberlite fields from the entire population of such fields in the region.

In this case, the described low magnetization of kimberlites in the Nakyn field is clearly explained by its being the first representative of the residual population, turning it from a paradox into a precedent.

We stress once again that the residual population of pipes, clusters and fields should have markedly reduced indicator values of the principal indicator parameter—in our case, weak and very weak magnetization of kimberlites.

In Figure 5, the same population of fifteen Siberian kimberlite fields is shown in different colours according to their levels of economic diamond content. As the figure shows, kimberlites in economically diamondiferous fields have the lowest magnetization values, whereas kimberlites in fields containing no known deposits are highly magnetic.

Significantly, the pipes of the Mirny field have the lowest average magnetization, and economic deposits account for approximately 50% of the total number of pipes in this field. In the other two, more magnetic fields, this proportion does not exceed 5%.

Data on variation in the concentrations of pyrope, ilmenite and other indicator minerals in pipes can be considered in the same way. The conclusions from these data are virtually identical to those obtained for the petromagnetic characteristics of pipes. Figure 6 shows data on the correlation between magnetic moment and mineralogical productivity for pipes in the Mirny field. As the figure shows, this correlation is virtually linear, i.e. weakly magnetic diatremes also have lower concentrations of indicator minerals.

The data presented appear sufficient to state the first principal conclusion, or to answer the first and second questions posed above. These answers can be formulated as follows:

The Nakyn precedent is inconsistent with the rule that conventional, albeit modern, geophysical technologies (magnetometry) are highly effective in exploring for commercially valuable primary diamond deposits. The Nakyn precedent overturns this rule. It clearly, convincingly and virtually unequivocally demonstrates the very low effectiveness of conventional magnetic exploration technology in detecting pipes and fields of the Nakyn type, i.e. exceptionally diamondiferous ones.

The example of the Yakutian kimberlite province supports the assertion that kimberlite fields of the Nakyn type belong to the residual population of exploration targets systematically missed by conventional exploration. In this population, the relative proportion of commercially diamondiferous fields, clusters and pipes is markedly greater than the proportion of valuable objects in the non-random sample selected by magnetic properties from high-contrast exploration targets.

Without discussing the supporting arguments within this paper, we present some data on the specific indicator characteristics of kimberlite bodies with high commercial diamond grades. These data are illustrated in Figure 7. Variation in the principal components of the material composition of Siberian kimberlites is shown in two-dimensional coordinates. Variation in the indicator parameters used to search for pipes is shown parallel to these axes.

The plot itself shows data for individual pipes, with pipes containing commercial concentrations of diamonds highlighted.

What follows from this plot? It follows that, within the entire set of studied kimberlite bodies, economic, commercially valuable deposits have not only lower magnetization but also higher electrical resistivity. Nor is it accidental that all kimberlite bodies in Yakutia discovered by electrical prospecting methods proved to be virtually barren of diamonds. Conversely, elevated concentrations of spinel and of chemical elements such as boron, zinc, chromium and nickel are characteristic of commercially valuable diatremes. Another important feature of valuable targets, confirmed by subsequent studies of kimberlites in the Nakyn field, is the greater abundance of sulphides in the pipes.

Table 1 presents generalised data on the indicator properties of economically (commercially) diamondiferous pipes in Yakutia.

Table 1. Generalised petrophysical model of commercially diamondiferous kimberlite pipes in the Yakutian province

Analysis of the table leads to the following conclusions:

Diamondiferous kimberlite pipes have consistent but only very slightly elevated anomalous magnetic susceptibility, which may appear in observed magnetic fields as anomalies with an average intensity of no more than a few nT. (Such anomalies are commonly identified by the thousands and tens of thousands in areas of airborne geophysical diamond exploration. In ground-survey data, such anomalies are usually obscured by noise from near-surface geological heterogeneities.)

Diamondiferous kimberlite pipes have a consistently elevated induced polarization parameter.

C. Experience in applying geophysical methods

For each geophysical method other than magnetometry, the history of its application to diamond exploration in Yakutia can clearly be divided into two periods. During the first, methods were used over large areas in attempts to identify anomalies caused by pipes. At first approximation, this approach seems understandable and apparently obvious. As it turned out, however, this is true only at first approximation. After the first stage, which generally yielded no results, a method was either removed from the list of mandatory exploration methods or assigned a different role in the overall cycle of geophysical work for diamonds.

Electromagnetic methods were applied in the following variants: ground-based direct-current and alternating-current electrical prospecting, induced polarization, cross-hole radio-wave transmission, and airborne electromagnetic surveying. The accompanying Table 2 briefly presents the results of applying these variants.

Table 2. Results of applying electrical prospecting methods in diamond deposit exploration

Thus, it can be stated that, with high measurement accuracy and optimised survey parameters, the method (except the induced polarization variant) predominantly identifies pipes barren of diamonds. Conversely, all these same variants reliably delineate linear kimberlite-hosting tectonic faults. The latter was also documented during work in the Arkhangelsk Region (Fig. 9). At the trial stage, the induced polarization variant identified all the commercially diamondiferous kimberlites studied (8 pipes).

Gravity prospecting was used at various scales. Its use for direct pipe exploration was limited because of low productivity, and surveys were usually conducted on relatively sparse grids. One pipe was discovered (Slavutich in the Alakit–Markha field). The pipe is barren of diamonds. Medium- and small-scale application of the method, after appropriate processing, identified kimberlite-hosting tectonic faults and approximately equant anomalies coinciding with the boundaries of kimberlite fields.

The seismic method was used to identify kimberlite pipes in covered areas. Trial work on known bodies showed that pipes can be distinguished by specific anomalies in layered host rocks. The method was applied over virtually the entire Mirny kimberlite field.

As a result, two types of anomalous objects were identified in the surveyed area:

“Pipe-type” anomalies within areas of undisturbed host-rock bedding. All these anomalies were intersected by drilling, and none proved to be a kimberlite pipe.

Linear zones of tectonic faults in which “pipe-type” anomalies could not be identified. These zones coincided, among others, with known kimberlite-hosting structures. In fact, however, they remained unexplored by the seismic method. This reflected the lack of a technology for identifying “pipe-type” anomalies in tectonic fault zones. On the other hand, the method did allow the kimberlite-hosting tectonic faults to be studied and mapped across the kimberlite field.

It should be stressed that these very kimberlite-hosting structures had previously been identified from ground and airborne electrical prospecting surveys.

Experience with airborne gamma-ray spectrometry proved particularly interesting. Initial trials in favourable landscape and geological settings showed that many kimberlite bodies stood out against the host rocks through elevated thorium concentrations and reduced potassium concentrations. However, attempts to intersect several radiogeochemical anomalies of this type by drilling yielded no positive results. It was decided to abandon the method.

However, new surveys of reference fields and their geological interpretation using a landscape-geochemical algorithm showed that the method mapped the secondary geochemical dispersion halos of the pipes rather than the pipes themselves (Fig. 10). It then became clear why drilling at the centres of the halos had not yielded positive results.

Trials using the new technology within the Mirny kimberlite field produced maps of radiogeochemical dispersion halos virtually identical to maps of the distribution of kimberlite-derived minerals (Fig. 11). The former, however, were obtained in a single airborne survey season, whereas the earlier maps had required several decades of drilling and excavation of mine workings.

This brief review of experience with geophysical methods in diamond deposit exploration, based on the examples presented and others, supports the following principal conclusions.

In principle, particularly in methodological trials in favourable landscape and geological settings, some anomalous effects in the vicinity of kimberlite pipes can be identified by most geophysical methods. However, none of these methods guarantees, with the required probability (reliability), that no kimberlite bodies will be missed, still less bodies with high commercial diamond grades. Moreover, some methods and variants, under the approach currently practised, select from the population only pipes whose average diamond content is markedly lower.

With virtually every method used, it proved possible to identify features of the regional structure of the measured fields that in one form or another indicate regularities in pipe distribution (exploration criteria), or signs of pipes near the anomalous effects detected by the methods (exploration indicators).

D. Optimal technologies for detecting highly valuable, low-contrast primary diamond deposits.

The data presented above allow us to formulate and illustrate an answer to the third question posed in describing the Nakyn precedent: what might be the optimal technologies for exploring for highly diamondiferous deposits with very low contrast?

The low-contrast expression of diamondiferous pipes in geophysical fields undoubtedly demonstrates the need for high-accuracy surveys. This first requirement is obvious and is currently met by most methods.

However, a low anomaly threshold automatically leads to the identification of hundreds and thousands of anomalies that formally satisfy the detection criterion. Investigating so many such signals requires enormous financial resources and time. It is precisely financial and time constraints that require the number of identified anomalies to be in the tens or, ideally, single figures. And it is financial constraints, rather than inadequate accuracy of measuring instruments, that raise the minimum intensity threshold above which anomalies are selected for further investigation.

The second requirement for the technology under consideration is therefore the ability to separate prospective anomalies from all measured anomalous effects without having to raise the minimum anomaly threshold.

We shall not discuss the familiar procedures for investigating and classifying anomalies based on formal characteristics of kimberlite-body-type objects (Euler deconvolution and others). These approaches generally work well for sufficiently intense anomalies and have their advantages and disadvantages. Nor shall we consider obtaining anomalous effects from pipes by using several methods. The data above show that, at least under Russian conditions, this is ineffective and leads to substantial financial losses.

There is only one way to meet the second requirement: a sequential, justified reduction in the areas being explored. Both sequence and justification are essential. In Russia, such technologies are commonly called staged exploration technologies.

A special study of the staged technology scheme for primary diamond deposit exploration identified an approach consisting of successively delineating the following intermediate exploration targets—or, more precisely, areas and sites prospective for their discovery. First comes the kimberlite field; within its probable boundaries, a cluster of kimberlite pipes; within the area of a probable cluster, local prospective sites and geophysical anomalies; and only at the end, kimberlite pipes (Fig. 12).

The anomalous effects in geophysical fields established and described above, which indicate regularities in the distribution of fields, clusters, sites and pipes and reflect the distinctive internal organisation of exploration targets, are thus the principal factors in the algorithm that makes this localisation possible.

It must be stressed that many prospective areas previously studied by geophysical and other methods do not require new surveys. In these areas, only appropriate reprocessing and reinterpretation of the data are necessary.

In general, staged exploration technology is implemented according to the following scheme.

The first stage involves analysing available small-scale survey data over territories of tens of thousands of square kilometres to delineate areas prospective for kimberlite fields, ranging from hundreds to a few thousand square kilometres. Methods: aeromagnetometry and airborne gamma-ray spectrometry. Delineation criteria: a regional anomaly in a special magnetic-field transform. Delineation indicators: results of landscape-geological reduction of small-scale airborne gamma-ray spectrometric data (Fig. 13).

At the second stage, medium-scale survey data are analysed within areas prospective for fields of pipes. The objective is to identify sites prospective for clusters of kimberlite bodies (a few hundred square kilometres). Methods: airborne electromagnetic surveying, airborne gamma-ray spectrometry and aeromagnetic prospecting. Delineation criteria: low-contrast zones of active dynamic influence of kimberlite-controlling faults (magnetic and electrical prospecting—special field transforms).

Delineation indicators: distinctive radiogeochemical dispersion halos of pipe clusters in modern and ancient sedimentary collectors; individual local magnetic-field and electrical-conductivity anomalies from the highest-contrast pipes, which serve as indicators of clusters (Fig. 14).

At the third stage, existing data are analysed or high-accuracy, precision airborne surveys are conducted using a suite of methods at scales of 1:10 000–1:5 000, with a ground positioning accuracy of 2–3 m. Delineation criteria: the internal structures of tectonic faults. Indicators: local magnetic anomalies, electrical-conductivity anomalies and radiogeochemical dispersion halos. Special transforms are used to suppress noise.

Ultimately, at this stage the exploration targets comprise not only anomalies but also local prospective sites delineated using a combination of criteria and indicators.

The next stage comprises verification and appraisal work at local sites and anomalies, aimed at further studying or identifying anomalies from pipes and subsequently classifying them according to their probable association with a commercially valuable deposit.

The technology described can be characterised as rapid, sufficiently reliable, highly knowledge-intensive and minimised in cost. It may be called targeted geological and airborne geophysical technology for diamond exploration.

By targeted geological and airborne geophysical technology for diamond exploration, we mean:

a) purpose-specific selection of the required work stages, airborne geophysical methods and variants, measurement accuracy, survey scales and altitudes for the various stages (where such data already exist, special pre-interpretation preparation is required),

b) purpose-specific methods for processing and presenting information, and a set of additional necessary data,

c) a purpose-specific workflow for general and predictive interpretation of the information obtained.

This technology is currently being introduced extensively into diamond exploration practice in Russia and has led to the discovery of a commercial deposit detected in the air as a 3 nT anomaly. Appropriate areas of application include both previously studied diamondiferous fields and problem areas where alluvial diamond deposits have not been linked to primary sources (many parts of Africa, Brazil, Ukraine, etc.).

Conclusions

The Nakyn kimberlite field, other clusters of kimberlite bodies and individual pipes within known kimberlite fields are typical representatives of a distinctive, particularly valuable population of primary diamond deposits characterised, on the one hand, by high commercial value and, on the other, by markedly reduced contrast in the indicator parameters used for direct detection.
It may be supposed that detecting this type of exploration target will determine the tactics and technology of diamond deposit exploration in the 21st century.

Based on knowledge of the distribution patterns of diamond deposits and the factors governing their expression, primarily in physical fields, this technology should aim to delineate, on a justified basis, a successively smaller nested series of intermediate exploration targets. Models of intermediate exploration targets for such a series have been developed using airborne geophysical methods.

Current trends in exploration methods, based on the most advanced knowledge-intensive technologies, suggest that targeted geological and airborne geophysical technologies for diamond exploration will form the basis of diamond exploration in the near future. Their speed, together with the objectivity and high information content of the data obtained, will enable maximum reliability and efficiency both within already known diamondiferous territories and in new regions prospective for diamonds.