The Kola Superdeep Borehole SG-3
History
The Soviet Russians became the world leader in deep drilling into the Earth's crust. The Kola superdeep well SG-3 was started in 1970. The site in Arctic Russia was chosen because it was in an area of rich copper and nickel deposits, and the crust was believed to be not too thick. In all the rock below the Earth's surface there is a balance between heat, which tends to melt the rock, and pressure, which tends to keep it solid. Both heat and pressure increase with depth. Normally the balance tilts in favour of melting at about 100 km depth, which is where the rocky lithosphere changes to the layer without strength, the asthenosphere. It is this slippy layer which permits the tectonic plates of the lithosphere to slide around.
But if you drill a hole into the upper crust, as was done at Kola, you reduce the pressure there, because you have created a hole. So the hole tends to collapse, and the rocks in the vicinity start to flow, losing their strength (heat wins out over the locally reduced pressure, making the rock plastic). The Russians tried several times to get down past the maximum depth that they had reached, 12,262 m vertical depth, by 1989. But each time the new hole tended to collapse.
The hole yielded several new scientific results, not least that the Soviet picture of the crust, where granitic rocks give way to basaltic rocks at about 7 km, was proved to be wrong. The granitic rocks continue much deeper, as western scientists believed. There was also water in the crust, an unexpected discovery, and traces of former life in the form of microfossils at 6 km depth.
The Kola-92 geophysical survey
The international geophysical experiment which I conceived in 1988 came to fruition in 1992. Known as Kola-92, it involved more than thirty scientists from Scotland, Norway, the USA and Russia. We carried out our experiments between February and May 1992 during late winter. Luckily the collapse of the former Soviet Union at the end of 1991 did not deter the scientists. But we had succeeded in getting there just at the right time. Any earlier, and the Soviet authorities would never have allowed westerners to enter the Murmansk region, which is a military zone; any later, following the collapse of the USSR, the political situation would have become too lawless and corrupt.
The aim of Kola-92 was to make a cross-section image of the crust - a technique in which we in the west were way ahead of the Russians - at the only place where there was the ground truth of actual rock samples from the deep borehole. Elsewhere, the crustal images already obtained by our seismic reflection technique, both offshore and onshore, often looked very impressive, but we would never know what the layers comprised. But the marrying together of our Kola-92 image with the Kola Superdeep borehole, showed, for example, that flat reflective layers seen on cross-section images could be due simply to water in the crust, and not due to different kinds of rock.
Legacy
Activity stopped at the site in 1995 after unsuccessful attempts to keep it going as a unique natural laboratory by attracting foreign interest and currency. Dismantling of the drilling tower and ancillary buildings began in 2008.
In retrospect, the Kola well is like a 'moonshot'- a one-off study which will never be repeated nor surpassed. Similarly our Kola-92 experiment, perhaps the most ambitious and difficult geophysical field experiment ever carried out, was also a moonshot which could never happen again.
I am proud to have conceived and organised the Kola-92 experiment; a unique piece of solid earth science research. But it was a disaster for me professionally; I had spent up to three years working on it at Glasgow University, while publishing almost nothing, but then failed to get the necessary research grant. Lay people reading this far should know that doing some bold and interesting research counts for nothing with academic employers like my university; all that matters is numbers of publications (never mind the quality!) and how much cash you are bringing in to pay the university overheads - in effect, the salaries of the bean counters.
The UK research agency NERC turned down my research grant application to fund the UK share of the Kola-92 project, despite enthusiastic backing from everyone I talked to. In contrast, this international joint project was evaluated and fully funded by the three other participating countries, suggesting that there was no fundamental problem to justify withholding support. In my view this failure was simply an early symptom of the decline in UK science. Risky, far-reaching and innovative science was no longer to be funded, only incremental and predictable steps forward, preferably with short-term commercial spin-offs. I shall explain elsewhere why I failed to get the essential grant.
Media and publications
In the optimistic period before my grant application was rejected I wrote an article for the Glasgow University magazine Avenue, explaining in lay terms what we were planning to do in Kola.
I carried a VHS camera with me throughout the Kola-92 survey, to film both the geophysical experiment and the operations at the wellhead. Once back in Glasgow I compiled a half-hour video film of our work. It concentrates on the Kola-92 survey, with the aim of persuading future funding agencies that what we had done was a success; however it does include unique footage of the operations at the well head, during a period when the Russian drillers were trying to retrieve some drill pipe dropped down the hole. Here is a downloadable version of the video, including supplementary footage.
I wrote a summary report of the Kola-92 experiment, which appeared in Eos, published by the American Geophysical Union, in 1994.
I was interviewed for the BBC World Service programme Witness History in 2024, recounting my experiences at the superdeep well. Here is a downloaded version for when the BBC link is no longer available.
The academic publications from Kola for which I was a co-author are listed on my publications page.
Data availability
The seismic data from the Kola-92 experiment were brought to Glasgow, where we spent another year reading and transcribing the tapes, then distributing copies to all our partners abroad. Despite the lack of funding at Glasgow we had to keep our promise to share the data as agreed. It never fails to astonish me that the 400-500 kg weight of hard-won raw magnetic data field tapes, totalling some 60 GB of digital storage, can now fit onto a USB stick!
The main datasets were duly processed and published by our American and Norwegian colleagues over the next few years. But there remains a large quantity of potentially interesting and unique data which to this day has never been examined, because I had to move onto other research projects and try to bring in some funding. In 2021, long after retirement, I spent a few months revising and correcting the locations of the survey line. Up till then we had relied on the Russian surveyor's estimates of our positions along the 40 km line through the well. Inaccurate position can lead to degradation of the processed seismic data. But with the advent of high-resolution satellite imaging I was able to accurately relocate our line locations, because the tracks of the caterpillar tractors and vehicles, which are a couple of metres wide, can still be seen in the taiga (boreal forest) some thirty years later!
My final plan is to merge the improved positioning data with the seismic data themselves, then make the whole dataset publicly and freely available for anyone with the interest and the necessary tools to reprocess, improve, and extract new findings from this unique dataset.