GB2443019A - Radio imaging of underground structures: Elimination of need for synchronisation channel - Google Patents

Abstract

A method of radio-imaging of geological features (particularly the imaging of coal seams) utilises the property of dispersion, by which is meant that signals of different frequencies travel at different velocities in a medium that can be classed as a good conductor. At the receiver, multiple measurements of relative phase shift made at three or more transmitted distinct frequencies can be used to deduce the conductivity of the medium. This is achieved without needing to know the absolute phase of the transmitted signal, thereby eliminating the need for a separate synchronisation channel. The transmitted frequencies may be closely-spaced to allow a single tuned antenna to be used. In one embodiment, the receiver utilises a magnetic loop antenna 41 and the signal is self-synchronised using a code-locked loop and a calculated inverse sequence 46. The relative phase-shifts are calculated using a Fourier transform within a microprocessor 44.

Description

I

Radio Imaging of Underground Structures: Elimination of need for synchronisation channel This invention relates to methods and systems for the radio-imaging of anomalous geological features and more particularly to the imaging of coal scams.

The basis of the radio imaging method is the measurement of the attenuation of medium frequency electromagnetic waves propagating over a number of intersecting paths between transmitters and receivers located in boreholes or mine workings. Typically, several individual frequencies may be used within the range 20 kHz to 2 MHz. In a typical implementation, in long-wall coal mining, measurements may be made across a panel of coal between the intake and return roadways, where the width of the coal panel may be as high as 300m. Changes in the phase and attenuation of the transmitted signals indicates anomalies in the coal seam that can be related to geology. If these anomalies are known in advance, steps can be taken to mitigate their effect, with significant benefits to coal production. The radio-imaging technique is well-established and is described in several existing patents, for example, "Synchronous radio-imaging of underground structures", US Patent 6,744,253.

Problems arise during the practical deployment of equipment using the radio imaging method (RIM), because of the need to synchronise the transmitter and receiver for the purpose of making measurements of phase shift. One method of achieving synchronisation would be to transmit a reference signal between the transmitter and receiver using a fibre-optic cable, but this is expensive and difficult to deploy because the cable has to run the entire length of the long-wall coal face. A copper cable cannot be used because of the danger that parasitic coupling would cause it to convey unwanted signals. Indeed, the problem of spurious signals usually requires mitigation by the disconnection or removal of all metalwork in the vicinity of the equipment for the duration of the measurements. The aforementioned patent describes a technique of synchronisation that involves transmitting a reference signal on a 2.5 kHz carrier. This frequency, it is claimed, is low enough not to be affected by the geological structure of the medium. Although this method conveys some advantages, it still requires the deployment of a separate transmitter and receiver, and requires the transmission of the reference signal between the synchronisation receiver and the sounding receiver using a fibre-optic cable.

To overcome these limitations, the present invention proposes a method of eliminating the need for a separate reference signal to be transmitted. This is achieved by making use of the property of dispersion -that is, that signals of different frequencies travel at different velocities in a medium that can be classed as a good conductor. Even so, any measurement of relative phase shift cannot, by itself, be used to establish an absolute phase reference for synchronisation purposes, because the velocity of propagation through the medium is a function of the conductivity, which is unknown. However, an advantage of the present invention is that multiple measurements of relative phase shift made at three or more distinct frequencies can be used to deduce the conductivity of the medium; and this goal is thereby achieved without needing to know the absolute phase shift, thereby eliminating the need for a synchronisation channel.

A further advantage of the present invention is that the probing frequencies chosen for these measurements may be closely-spaced (for example, they might be at 99 kHz, 100 kHz and 101 kHz) thus allowing the signals to be transmitted simultaneously from the same tuned antenna, which is in contrast to the system described by US Patent 6,744,253 which used two separate transmitters and two widely-spaced signal frequencies.

A further advantage of this invention is that the relative attenuation of the signal at two or more frequencies over a single transmission path may be used to derive additional information about the transmission medium.

This invention will now be described by way of example, and with reference to the accompanying drawings in which: Figure 1 shows a stylised plan view (not to scale) of a long-wall coal mine using retreat mining.

Figure 2 shows a block diagram of a radio-imaging transmitter in an embodiment of this invention that uses three discrete transmitter frequencies.

Figure 3 shows a block diagram of a radio-imaging transmitter in an embodiment of this invention that uses a binary sequence to generate a wideband signal.

Figure 4 shows a block diagram of the corresponding radio-imaging receiver.

A typical mining operation is as depicted by Figure 1. An intake roadway 1 and a return roadway 3 are driven into the coal seam 4 for a distance that may be several kilometres. The spacing of the roadways may be 350m and the coal panel between them, which is to be mined, may contain geological anomalies 2. A coal-cutting device (e.g. a shearer or plough) 8 traverses the long-wall coal face as shown by the adjacent double-ended arrow, and retreats from the mined area 7 back towards the start of the roadways. In order to detemiine the location of the geological anomaly 2, a radio transmitter 5 is located in the intake roadway 1, and a radio receiver 6 is located in the return roadway 3. During the surveying operation they are independently re-located along the roadways as shown by the adjacent double-ended arrows in order to obtain a multiplicity of readings of phase and attenuation of the radio signal that is traversing the coal panel. The data is then processed using a tomographic reconstruction algorithm to obtain information about the geological anomaly.

In one embodiment of this invention, the transmitter utilises three closely-spaced transmitter frequencies as depicted in Figure 2. The frequency generators 22 produce analogue sine waves, and these may be conveniently synchronised by means of a microprocessor 21 or may be digitally synthesised in a manner that will be obvious to a practitioner of electronic circuit design. The frequencies are summed in a summing amplifier 23 and amplified in a power amplifier 24 before being fed to the antenna 25 which may be an induction loop or solenoidal device, air-cored or otherwise, or an electric field antenna of a type appropriate to the frequencies being transmitted, or a current-injection type of antenna utilising electrodes placed into the coal seam. The antenna may be tuned to resonance for efficiency, in which case the frequencies transmitted will be chosen so that they are all within the useable bandwidth of the transmitter.

In another embodiment of this invention, a multiplicity of frequencies may be used for the determination of the rock parameters, and these frequencies may be transmitted sequentially or simultaneously; and furthermore, that they may arise as the Fourier components of a mathematical sequence such as a maximal-length pseudo-random binary sequence, or other specially-calculated sequence of samples. Other methods and procedures will no doubt become apparent to those of ordinary skill in the art after reading the detailed description of this invention.

In Figure 3 an embodiment of this invention is depicted in block form whereby a microprocessor 31 generates a signal that is fed directly to the power amplifier 32 and antenna 33. This signal may be analogue (following the scheme of Figure 1) or it may be a digital sequence containing spectral energy at a multiplicity of frequencies. Being binary in nature, such a sequence will lead to a higher efficiency in the operation of the amplifier than an analogue signal. The disadvantage of requiring a broad-band untuned antenna will, in such an instance, be offset by the processing gain that is possible at the receiver.

In Figure 4 a radio receiver is depicted in block form, in an embodiment of this invention that is suitable for decoding and processing a wideband binary sequence such as that transmitted by the arrangement of Figure 3. The receiver antenna 41, which may be of any of the types suggested earlier, is connected to an amplifier 42 an thence to sampler and analogue to digital converter 43. The resultant data is processed by a discrete Fourier transform (DFT) or other means of analysis under the control of a microprocessor 44.

In order to provide synchronism to the transmitted sequence (which is essential for the correct operation of the Fourier transform) a signal correlator 45 is utilised. The oulput of this correlator is filtered in an appropriate fashion by a digital filter 48, the output of which controls a voltage-controlled oscillator (VCO) 47. The output of the VCO is used to clock a look-up table 46 or other means of waveform generation in order to syrnhesise the inverse sequence to that which was transmitted. This is cross-correlated against the received signal by means of the correlator 45. The circuit elements 45 to 48 thereby comprise a code-locked or delay-locked loop which performs a similar function to a phase-locked loop, and which causes the receiver's sampler to track any drift in frequency of the transmitted signal. These circuit elements may be discrete and operating in an analogue signal domain, or they may be part of a computer program residing within a digital signal processor or field-programmable gate array. Other implementations of a synchronisation thcility with similar features will no doubt become obvious to those of ordinary skill in the art after having read the

above description.

The inverse sequence may be calculated in such a way as to result in a favourable control function (e.g. a so-called S' curve) at the output of the correlator. Additionally, the inverse may be calculated for the combination of three discrete frequencies, such as that transmitted by the arrangement of Figure 2. It will be apparent to an expert that some sequences do not have a finite and calculable inverse. However, a suitable quasi-inverse' can be calculated that merely results in the wanted output being superimposed on a zero-frequency offset.

In Figure 4 a microprocessor 44 supervises the collection of the data and its Fourier transformation into a set of frequency and phase components. According to the current invention, there exists a means of processing the data to deduce the conductivity of the transmission path without the requirement to know the absolute phase shift, and this will now be described.

In order to gain a basic understanding of the technique it will first be described on the assumption that electromagnetic waves are present solely in a plane-wave, far-field, configuration in a uniform conducting medium. This assumption is not valid, of course, and therefore a more complete model will be addressed subsequent to this preliminary description.

A simple far-field propagation model: When a monochromatic plane wave travels in a conducting medium it propagates in accordance with a complex wave number k', = Ic, -jy e1 = (1) where x is distance, Ic, is the (real) wave number and y is the attenuation coefficient, which can also be written as 1/8, where 8 is commonly referred to as the skin depth. The angular frequency of the wave is o and its phase velocity is v. In a sufficiently poor conductor, or at a sufficiently high frequency, such that <<1 (2) (where a is conductivity and c is permittivity) it is possible to derive the expressions k'=k,.-j/8, (3) (where Z is the wave impedance, defined by J(W) and is a real, i.e. not complex, quantity) and thereby deduce that the wave velocity is simply l/'J(ps). However, in a sufficiently good conductor, or at a sufficiently low frequency, such that (4) the expression resolves to k'=.!-LL, with = 8 (5) Such derivations are commonplace in textbooks on electromagnetic theory.

If we note that (1) is a representation where a complex time variation exp(ftoi) is assumed, we can deduce that the phase of the received signal is -wt -k,x or, following common convention, -wt -(Jx where r represents an arbitrary time origin and is the phase constant, equivalent to Ic, in (1).

Considering the practical measurement of phase at the receiver, we must write the phase as 4=-cot-13x+2rnt (6) where the 2,nt term represents the fact that the measured phase is distinct only over a range of 2n rad.

By transmitting signals at two frequencies, it is clearly possible to derive t in tenns of 13. However, since the conductivity of the medium is not known, 13 is unknown. Moreover, t cannot be eliminated from the equations because ii is unknown.

According to the present invention, these difficulties may be eliminated by transmitting three frequencies. If these frequencies are notated as w and w 5w the phases of the signals at the receiver are given by 4 =-(w-8w)t-31x+2n,t +2 =-(c01.60))t-132X+2nlr +0 =-cot-f30x+2nn (7) from which we can obtain +I++2-2$O=.4I+2_2J3O)x (8) which we can expand to (9) and from which we can derive the product w. If we assume that.t = j. io we can thereby denve a. We can apply a binomial expansion to the right-hand term to give, for small 8w +1 + -2$ = 2 0) (10) To be strictly accurate, we must demonstrate the phase ambiguity in + and write the above as mod(+1 + +2 -24, 2n) = I1+./(II)2x (11).

It is also pertinent to point out that there is the implicit assumption that n is the same for all three measurements. This will be the case if 13x << 2n, which equates to (12) which is clearly the case in all practical uses of this technique.

In summary, the above derivation has shown that by measuring the relative phase shifts $ -$0 and 42-40 we can deduce the conductivity of the medium from equation 10. This derivation is possible because 13 is a function of frequency. Moreover, it is only possible because f3 is a non-linear function of frequency, which means that the group velocity (i.e. d /d13) is not constant and so, according to established theory, the signal will undergo the phenomenon of dispersion. We are, in effect, utilising the dispersion of the signal to measure the phase constant. We use three frequencies, and rely on the fact that the dispersion of the first pair will be different to the dispersion of the second pair.

Operation in the near field: The preceding description assumed that the propagation was in accordance with the complex wave number, as exp(-jk,r). However, this does not take into account the fields close to the transmitter. According to standard theory, when we consider the propagation of the fields close to a magnetic dipole antenna in a uniform conducting medium where a/&o>> I, we have to include an additional propagation term representing the near field, namely 2cos 8 (i + (1 + j)T) I + sin 0(1 + (1+ j)T 2jT2)e (13) where r and 0 arc polar co-ordinates and Tis the dimensionless ratio x/&, or 13x. (This immediately suggests that the dispersion is dependent on direction, which is an additiona Ifactor that has to be taken into account). Considering the co-axial situation, where 0=0, the additional phase shift 4iiS given by tanW=Z_= 1+T 1+13x (14) depending on which notation we wish to use. For T>> 1, we have, 45 , independent of distance and frequency, so the original analysis is valid. However, within the near field, at T<< 1, the expression reduces to Tor i (3x and, since this phase shift is in the opposing direction to the existing phase constant term, the implication is that the resulting waves exhibit no phase shift with distance, or travel at infinite speed. This is, of course, true' for the quasi-static approximation of the near field, thereby indicating that an exact representation for the phase shift is required, and not a quasi-static if other approximation. In practice, we can therefore expect an accurate solution for 13 to be difficult. The requirement is to solve for a the set of equations 4 +4 -24 =(w +v2 -20)-(1 13 -2130)x (15) with t3,=w,+x (16) and tan qi, = o,Xx/(1 + (17) andwhere x=/ii. (18) It will be obvious to an expert that this set of equations can be solved iteratively if no analytical solution can be derived.

In summary, the second derivation above has shown that by measuring the phase shifts 4 -$o and -4, and by assuming an appropriate propagation model for the medium, we can deduce the conductivity of the medium even if this has to be derived iteratively rather than analytically.

In a further extension of the derivations give above, a propagation model suitable to describe a panel of coal within a long-wall mining operation can be considered. Commonly, in the specialist literature such models are described in terms of a low-conductivity coal seam bounded by high-conductivity strata, and a TEM mode propagation is assumed. For those not already skilled in the art, a description of this mode of propagation is given by 1. Enislic, A. 0. and R. L. Lagace (1976). Propagation of Low and Medium Frequency Radio Waves in a Coal Seam. Radio Science 11(4), 253-261 2. Hill, D. A. (1984). Radio Propagation in a Coat Seam and the Inverse Problem. Journal of Research of the National Bureau of Standards 89(5), 385-394 3. Wait, .J. R. (1976). Note on the Transmission of Electromagnetic Waves in a Coal Seam. Radio Science 11(4), 263-265 It will, of course, be evident that the above derivations have all assumed a uniform conductivity throughout the medium, whereas the present invention claims to derive the non-uniform conductivity of the medium from measurements of phase. However, it Will now be apparent to an expert in the art that the phase-difference method described above may be applied in a straightforward manner to a tomographic reconstruction algorithm. Moreover, it will be clear that, although this invention is most easily described by way of algebraic example, such information is not required for a tomographic reconstruction that relies on a finite-element analysis of the propagation in the medium.

Commonly, tomographic algorithms compare phase and amplitude variations at different physical locations of the receiver, whereas this invention has described a method of comparing phase variations at different frequencies at the same receiver location. A further embodiment of this invention is to utilise amplitude (attenuation) measurements at different frequencies rather than phase measurements. Adopting, for clarity of explanation, the simple plane-wave model described above, the magnitude of the field strength at the receiver may be described by (19) where a is an arbitrary constant that depends on the transmitter power and the distance. If two signals are transmitted at closely-spaced frequencies o &o, the resulting ratio of attenuation at the two frequencies is = exl-Iii! (20) V2) thus demonstrating that the path conductivity may be deduced from ratiometric measurements of signal strength at a single receiver location.

In a still further embodiment of this invention, a multiplicity of frequencies may be transmitted in order to refine the calculations of path conductivity. Such frequencies may be derived, for example, from a sequence of samples that may be binary values or be multi-valued and which may comprise a maximal-length pseudo-random binary sequence or some other appropriate sequence. In such instances, it may be convenient to receive the signal by sampling and to derive the frequency amplitudes and phases by means of a Fourier transform as depicted in Figure 4.

In such an instance, it is important that the frequencies employed are exact integer multiples of the Fourier transform cycle time. in other words, the interval between successive zero-crossings of the frequency components must be an exact integer number of samples. This will clearly be the case if the transmitted sequence is locked to the receiver by means of a code-locked loop. The reason for this requirement is that a discrete Fourier transform (DFT) is an identical operation to that of a bank of digital filters, each with a centre frequency of nf,/Nand a bandwidth of approximatelyf; /Nwhere n is an integer, N is the size of the DFT andf is the sample rate. If the incoming signal is not at the exact centre of one of these filters then its amplitude will be recorded incorrectly and, more importantly, its phase will be recorded incorrectly, due to the phase-shift that occurs as the digitally-synthesised band-pass filter is traversed from its lower to upper cutoff points.

Although the present invention has been described in terms of certain specific embodiments, it is to be understood that this disclosure is not to be interpreted as limiting. Various alterations and modifications will no doubt become apparent to those skilled in the art after having read the above disclosure. Accordingly, it is intended that the subsequent claims should be interpreted as covering all alterations and modifications that fall within the spirit and scope of the invention.

Claims (9)

1. Claims 1. A method and apparatus for the radio-imaging of geological

   features wherein an electromagnetic signal comprising multiple frequency components is transmitted through the medium to be investigated (which may be a coal seam) and is detected and processed by a receiver, and wherein the frequency components of the received signal are compared in a relative fashion to allow certain properties of the medium to be deduced without knowledge of the absolute phase of the transmitted signal and which thereby eliminates the need for a separate synchronisation channel.
2. 2. The method of claim 1, wherein information about the geological properties of the medium, including its effective bulk conductivity, is obtained from a transmitted signal by measurements made at the receiver of the relative phase or relative attenuation of frequency components of the transmitted signal.
3. 3. The method of claim I, wherein information about the geological properties of the medium, including its effective bulk conductivity, is obtained from a transmitted signal by measurements made at the receiver of the degree of signal dispersion that occurs between frequency components of the transmitted signal.
4. 4. The system described by claims 1 to 3, wherein the necessary phase and amplitude information is derived from a multiplicity of frequencies that are transmitted sequentially by analogue or digital means.
5. 5. The system described by claims 1 to 3 wherein the necessary phase and amplitude information is derived from a multiplicity of frequencies that are transmitted simultaneously by analogue or digital means.
6. 6. The system described by claims 1 to 5, wherein the transmitted frequencies are sufficiently close in value that the antenna may be tuned to resonance for efficiency.
7. 7. The system of claims I to 3, wherein the necessary phase and amplitude information is derived from a multiplicity of frequencies that arise as the Fourier components of a transmitted mathematical sequence such as a maximal-length pseudo-random binary sequence, or other specially-calculated sequence of samples.
8. 8. The system described by claims I to 7, wherein the necessary phase and amplitude information is obtained by the method of cross-correlation of the received signal with the inverse of the transmitted digital sequence or multiplicity of frequency tones in such a manner as to effect a useable control signal of the type known as an S-curve.
9. 9. The system described by claims 1 to 8, wherein the transmitter is of the current-injection type, utilising electrodes placed into the coal seam. "-7