The Geological Survey of Finland (GTK) acquired an electromagnetic tomographic drillhole device (referred to as FARA) from Russia at the beginning of the 2000s when Finland (with the permission of the Paris Club) agreed on debt conversion and the debt was partially paid with goods and services. The problem of forming a cross-sectional image or a tomographic reconstruction from its projections arises in a variety of contexts, including medical computed tomography (CT) and the radiofrequency imaging method (RIM) between two deep drillholes. Projections refer to information that is derived from transmitted electromagnetic waves propagating through an area of interest, and RIM is an electromagnetic (EM) drillhole method.
The Austrian researcher Radon developed the principles of tomography at the beginning of 1900s, and he demonstrated that a target can be exactly reconstructed from a complete set of its projections when the full-angle geometry (2π, 360 degrees scanning possibility) of the measurement is achieved. Today, tomography is widely used in the medical world and it is also gaining ground as a geophysical exploration technique for determining changes in the physical properties of subsurface targets. CT can easily provide high-resolution images of body sections, even at the millimetre scale, due to the 2π geometry. In geotomography, the 2π geometry is never met, because the transmitters and receivers are usually restricted to the earth’s surface or within a few drillholes, and insufficient measurement geometry thus leads to sparsely sampled data series.
GTK’s tomographic system is referred to as the electromagnetic radiofrequency echoing (EMRE) system, which includes the device, the image reconstruction and the 3D presentation. The transmitter is a continuous wave (CW) device with a nominal power of 2 W. The power consumption is low, and thus the batteries do not run out during a normal working day. The drillhole transmitter radiates a carrier wave (pure wave of frequencies) consisting of four medium radiofrequencies (312.5-625-1250-2500 kHz; k = kilo = 103), and without modulation the device cannot be used to measure the distance to a target due to the lack of further information on the carrier. Modulation refers to imposing input data onto a carrier, thus changing the shape of the carrier wave in some way. The frequency is the number of oscillations (or cycles) per unit of time. The receiver consists of a drillhole and a surface receiver, which are connected with a winch cable.
High frequency electromagnetic fields propagating in a lossy environment strongly attenuate due to possible conductive, dielectric and magnetic losses, which effectively restricts the measurement distances. However, according to theoretical studies, the EMRE system can even be used at the distances of ~1000 metres when the host rock electrical resistivity is >15 000 Ωm and the two lowest measurement frequencies (312.5-625 kHz) are used. Because RIM measurement is conducted between two deep drillholes, this is a restrictive factor, and RIM is thus also referred to as a limited angle problem. The lack of 2π geometry also restricts the resolution, which is tens of metres at best. However, the resolution is high when compared to traditional drillhole and surface geophysical methods. As a high-resolution technique, RIM is a powerful tool to gain a more detailed view of an existing ore body and can efficiently guide new drilling programmes and even mine planning (Figure 1).
In Finland, RIM has been used since the pioneering work by A. Korpisalo in 2005. The measurement units of the EMRE system are the amplitude and phase of the electric field. The heart of the system is the CW transmitter, which simultaneously operates at a maximum of four frequencies. The measurement frequencies can be freely selected based on the jumper settings. This is a useful property in situations where the material between the drillholes is highly conductive and, for instance, the highest frequency of 2500 kHz can be rejected and all the power can be fed to the three lower frequencies. In addition to four measurement frequencies, the vital reference frequency of 156.25 kHz is generated in the drillhole transmitter. Low-pass filters with high impedances (resistances to the alternating current (AC)) are used to isolate the transmitter and receiver antenna from the winch cables. The technique is used in multiband trap dipoles or to cut off frequencies above the reference frequency of 156.25 kHz. The reference is guided through the winch cable to the surface amplifier and further along the surface line to the surface receiver. The reference signal has an important role in the proper detection of the amplitude and phase.
Technically, the receiver is set up in a manner that does not deviate from a normal radio. The front end of the receiver (RF amplifier and mixers) is situated in the drillhole receiver and the normal components (e.g. the detector and analog-to-digital converter) in the surface receiver. The winch cable is used as a guiding path for the amplified and heterodyned signals from the drillhole receiver to the surface receiver. Heterodyning means the mixing of the frequency band derived from the local oscillator frequency (2600 kHz) of the drillhole receiver with the received measurement band to produce an intermediate frequency band (IF). The IF band is low enough to be transferred through the drillhole filters to the surface, where the final detection is performed. The low dynamic range of ~30 dB of the present receiver is a restrictive property of the system and can easily lead to situations where strong signals are cut, especially when nearby drillholes are used and the host rock resistivity is high.
The transmitter and receiver antenna serve as the transducers between the device and environment or the antenna converts radiofrequency current into EM waves, and vice versa. In the EMRE system, the transmitter and receiver antenna (vertical half-wave dipoles) are omnidirectional. Thus, the transmitter antenna radiates equal power in all directions and the receiver antenna receives power from all directions. The strongly isolated half-wave dipole antennas of the system are 20/40 m long. They are both physically and electrically short. The antennas are centre fed and consist of control tubes 36 mm in diameter for the corresponding electronics and voltage source. The tubes are 1.5 m long in the transmitter and 2.5 m long in the receiver. Intermediate cables of 10/20 m in length are used to connect the antennas to the winch cables.
Two measurement techniques can generally be used. The measurements can be performed using the MOM (multi-offset measurement) or ZOM (zero-offset measurement) technique. MOM is the traditional method, in which a stable transmitter is fixed in one drillhole and a mobile receiver is moved in the other drillhole (fan-beam geometry). The main idea is to measure a large number of angles passing through the space between the drillholes. In ZOM measurement, both the transmitter and receiver are moved synchronously in the drillholes (parallel-beam geometry). This is a rapid and simple technique to locate and define anomalous attenuation zones. In a homogeneous environment, signal levels should be same at each location. Normally, the measurement is performed as a two-way measurement, where the transmitter and receiver are interchanged in the drillholes to accomplish a full tomographic survey between them.
When an electromagnetic field propagates in lossy material, the attenuated field propagates at a lower speed than light in air (visible light that the human eye can perceive is also a part of the electromagnetic spectrum, having wavelengths of 380-740 nm (n = nano = 10-9) corresponding to a band of 430-770 THz (T = tera = 1012)), and its wavelength is also shortened. The wavelength of a wave is the distance between any similar features on the waveform (e.g. successive positive or negative peaks). In air, the wavelength of 2500 kHz is 120 m (according to the formula λ [m] = 300/f [MHz], where λ is the wavelength in metres and f the frequency in MHz), but in lossy material, the wavelength (in the EMRE band) can be as short as ~40 m. Thus, the 20-m dipole antenna of the device is very close to the resonance state of the 20-m half-wave dipole and the antenna can operate effectively. Shorter wavelengths in lossy material always mean that the propagation speed of the attenuated electromagnetic field must be only a fraction of the speed of light in air. The measurements can be monitored in real time on the screen, thus allowing the operator to gain initial insights into the subsurface geology at the site and also to modify the measurement plan if necessary. In Figure 2, a typical measurement result from the Pyhäsalmi mine area is presented.
RIM is generally conducted in the transmission mode (transmitter and receiver in separate drillholes) when the radio waves are attenuated due to the electrical and magnetic properties of the medium between deep drillholes (Figure 1). Greater conductivities always mean higher attenuation rates, and both the range and resolution are frequency dependent. For most cross-drillhole survey geometries, the angular aperture is limited and the horizontal resolution is less than the vertical resolution. Usually, the central sections are best covered and contain the greatest amount of information. The ray coverage is typically low at the top and bottom of the inter-drillhole plane, resulting in ghost structures and smearing of anomalies, which reduces the resolution in these areas.
Commonly, the smallest structure that can be resolved is estimated by the width of the first Fresnel zone. The Fresnel zones are ellipsoids drawn between the transmitter and receiver. The first Fresnel zone is a region around a ray that mostly affects the propagation of the wave. A transmitted wave can alternatively reach the receiver by a direct route or, when the wave is spread out at an angle, it may not even reach the receiver at all. On the contrary, when, for instance, the wave strikes an obstruction, the reflected wave may reach the receiver as a reflected wave. When the object is an off-plane target (not in the drillhole plane) and the transmitted wave is diffracted due to the object (the wave bends slightly and passes around the edge of the object), the effect can generally only be detected with one specific measurement frequency and the others are blind to it due to the different wavelengths. The width of the Fresnel zone is spatially variable and is determined by the wavelength and the distance between the drillholes, being smallest near the transmitter and receiver drillholes and largest midway between the drillholes (Figure 3).
Due to the frequency band (long wavelengths) generally used in RIM, the method is not sensitive to thin discontinuities (e.g. conductive black schists). Conversely, a massive sulphide deposit appears as an excellent target. The optimal geometry consists of two parallel vertical drillholes. The drillholes should be roughly in a common plane, and the minimum starting depth should be greater than the horizontal separation of the drillholes, thus clearly below the air–earth interface to minimize the disturbing effects of reflected rays from this interface. In addition, the drillholes should be deeper than the distance between them. On the contrary, when the transmitter is located deeper than the receiver, measurement can be started at even shallower depths. A scanning aperture of 30-50 degrees together with dense receiver stations results in reasonably smooth tomographic reconstructions.
When contrasting electrical properties in the drillhole section are moderate, the transmitter–receiver separation is several wavelengths and displacement currents predominate, a cross-sectional image can be reconstructed using techniques that are based on the far-field solution of the electric field. In addition, the field is assumed to propagate as straight rays and changes in the electromagnetic field are generated by changes in the material properties along the ray lines (linear based tomography). However, it should be noted that electromagnetic fields and subsurface properties are non-linearly related, and reflection, scattering and refraction of the rays should be taken into account.
Mathematical methods (e.g. the simultaneous iterative reconstruction technique, SIRT) that are based on the straight ray assumption are algebraic methods. They can easily been implemented and give rapid solutions. The methods can, however, be semi-convergent: at first they can converge towards a reasonable solution, but later the solution can even diverge. Even today, suitable forward codes for simulation studies are only just becoming available, and are still very computer time intensive. With increased computer power, a true 3D numerical modelling technique could also be applied for geotomography, resulting in more precise reconstructions of drillhole sections when realistic antennas and electric and magnetic losses in the earth’s subsurface can be taken into account. The use of a true 3D technique would replace linear inversion with full-waveform nonlinear inversion, but rapid linear results can still be useful in serving as seeds for these modern methods.
We have utilized the RIM method in two quite different areas: the Olkiluoto site and the Pyhäsalmi mine. In Finland, the final disposal of spent nuclear fuel is planned to take place in crystalline bedrock at the Olkiluoto nuclear power plant site in western Finland. Olkiluoto was a radical, interesting and pioneering area for RIM measurements, because conductive mineralizations in the drillhole section were not the focus but possibly existing fractures. Fractures filled with salt water can be detectable with RIM if the wavelengths used are short enough or the frequencies are sufficiently high. The Pyhäsalmi massive volcanogenic massive sulphide (VMS) copper–zinc deposit is an ideal research area for RIM. In both of the cases, the results were very promising, despite the device malfunctioning at the Olkiluoto site. In a recent article published in Geophysics in 2019, the Pyhäsalmi case is thoroughly considered. The measurements consisted of three sections, but only the results from sections PYS127-MPYS113 and MPYS113-MPYS119 are discussed in the paper (Figure 4).
The RIM results were compared with time-domain electromagnetic (TEM) data and electric logging data. Electric logging reliably revealed the nearby conducive mineralizations, but when compared with RIM data, the continuation of attenuating formations could be better predicted from RIM data. The intersections interpreted from the TEM data were consistent with the RIM data. However, continuation of the attenuating domains could only be established from the RIM data (Figure 5). All matters considered, the attenuation measurements were well suited to estimating subsurface conductivity properties and continuity information between drillholes at Pyhäsalmi.
RIM is an EM exploration method with high potential that can offer much better resolution and sensitivity to conductive targets than traditional ground-level and drillhole EM methods. Electric logging reliably reveals intersections and conductive mineralizations in the close vicinity of drillholes, and TEM measurements can more profoundly delineate them. However, being less sensitive to conductive mineralizations further from the drillhole, the continuation of mineralizations between drillholes separated by hundreds of metres could not be inferred using these methods. Cross-drillhole measurement effectively senses the whole space between the drillholes, and the RIM method more powerfully establishes the continuation of the attenuating mineralizations. Some significant problems exist in geophysical tomography, the nonuniqueness of the reconstructions and the accuracy of the used methods. Nonuniqueness is generated by the limited angle data. Smearing in the direction of the ray path can never be avoided when simple iterative ray-tracing methods are used. In addition, these methods are semi-convergent. The reconstructions become smooth, the well-defined boundaries are blurred and artefacts are possible. Thus, true nonlinear numerical inversion methods will probably result in more realistic and accurate reconstructions of the subsurface sections. Suitable forward codes for simulation studies and interpretation are only just becoming available, and they are still very computer time intensive.
We have a unique RIM device at GTK, but because it was possibly built as early as in the 1970s, the device is very fragile. Some of the parts should quickly be rebuilt to keep the device in proper working condition and secure the RIM measurements. Modernizing the drillhole transmitter and receiver would raise the functionality of the system to a new level (e.g. 60 dB dynamic range and phase resolution of 1-2 degrees can easily be achieved) and help to collect data of even higher quality. Together with modern non-linear numerical inversion methods, the interpretation would also be enhanced.
Korpisalo, A., 2019, Tomographic reconstructions of borehole sections using the radio imaging method at Pyhäsalmi massive sulfide deposit in Finland: Geophysics, 84, B217-B233, doi: 10.1190/GEO2017-0332.1
Korpisalo, A., 2016, Electromagnetic geotomographic research on attenuating material using the middle radio frequency band: Ph.D. thesis, University of Helsinki.
Korpisalo, A., 2014, Geotomographic studies for ore explorations with the EMRE system: Measurement, 48, 232–247, doi: 10.1016/j.measurement .2013.11.016.
Korpisalo, A., and E. Heikkinen, 2014, Radiowave imaging (RIM) for determining the electrical conductivity of the rock in borehole section OL-KR4–OL-KR10 at Olkiluoto, Finland: Exploration Geophysics, 46, 141–152, doi: 10.1071/EG13057.
Text: Arto Korpisalo
Arto Korpisalo received his MSc (1996) in medical physics from Kuopio Korkeakoulu (Finland) and PhD (2016) in physics from Helsinki University. He has worked since 1997 in Geological Survey of Finland (GTK) as a physicist. His research interests include forward and inverse problems in drillhole thermophysics and electromagnetism.