GB2322937A

GB2322937A - Multiphase fluid monitor

Info

Publication number: GB2322937A
Application number: GB9704778A
Authority: GB
Inventors: Duncan Brian Condie Brown Syme; Malcolm Roderic Wormald; John Charles Fisher; Gwilym Dorien James
Original assignee: AEA Technology PLC
Current assignee: Ricardo AEA Ltd
Priority date: 1997-03-07
Filing date: 1997-03-07
Publication date: 1998-09-09
Anticipated expiration: 2017-03-07
Also published as: GB9704778D0; GB2322937B

Abstract

The fluids in a pipeline 16 are characterised by detecting the prompt gamma rays emitted when the fluids are irradiated with fast neutrons from a pulsed neutron tube 30. The gamma rays are detected by an energy sensitive detector 36, and the number counted in energy windows. At intervals the fluid in the irradiated region is controlled to have a known composition, eg. empty, so the counts can be corrected for drifts in the baseline. This may be achieved using a bypass system. A scintillation detector 48 may be used to detect the gamma rays and rays originating from a fast neutron are distinguished from those from a thermal neutron by a time gate. Operation of the detector may be stabilised by observing a known peak in the gamma spectrum, and the counts also be corrected for any variations in neutron flux. The rate of flow of the fluid can also be determined.

Description

Multiphase Fluid Monitor This invention relates to an apparatus and to a method for monitoring a multiphase fluid.

The use of neutron interrogation and/or neutron activation, in conjunction with gamma detection, has been suggested for monitoring the contents of multiphase oil pipelines, and for flow measurement in such pipelines.

For example EP-A-0 007 759 suggests the use of neutron irradiation, and measurement of gamma rays with a NaI(Tl) scintillation detector, to assess the water content of crude oil. GB 2 276 237 A describes the use of a neutron tube to irradiate the contents of a multiphase oil pipeline, with two scintillation detectors, one to detect prompt gamma rays, and the other located downstream to detect gamma rays from activated nuclei.

According to the present invention there is provided a method for monitoring the contents of a pipeline carrying a fluid, the method comprising irradiating the fluid at an irradiation region with fast neutrons from a fast neutron generating tube, detecting prompt gamma rays emitted from the irradiation region during the said irradiation, and providing first signals characterising the energy spectrum of the prompt gamma rays, at intervals controlling the contents of the irradiation region to contain a standardised fluid, performing the irradiation with neutrons and the detection of the prompt gamma rays as aforesaid and hence providing second signals characterising the energy spectrum of the prompt gamma rays in the presence of the standardised fluid, and from the first signals and the second signals determining characteristics of the fluid in the pipeline.

Measurements of a gamma ray spectrum may be made with one or more different standardised fluids at different times. For example one standardised fluid might be a gas such as air at atmospheric pressure (which gives negligible gammas), another standardised fluid might be a chemically well-defined liquid such as water, methanol, or pure kerosene completely filling the irradiation region. Such standardised spectra might be measured once every month, more preferably every week or even once every day or every few hours.

It has been found that the prompt gamma ray spectrum so measured may vary gradually over time even if there is no change in the fluid in the irradiation region. For some situations there is in any event only a small difference between the spectra obtained with an irradiation region with liquid present or with an empty (air-filled) irradiation region (referred to as the baseline), so for example if the liquid is an oil the 12 prompt Y rays from C may be only a few per cent above the baseline signal at their energy. By obtaining spectra with a standardised fluid at intervals any drift in the baseline can be allowed for, so the fluids can be characterised more accurately.

The method is particularly suitable for where the fluid is an oil/water/gas mixture, for example in a pipeline carrying fluids from an oil well, enabling measurements to be made of the water cut, i.e. the percentage of water in the liquid phase, and also the void fraction i.e. the proportion of gas in the mixture.

It will be appreciated that the complete gamma spectrum, that is to say the numbers of gamma rays at different energies, need not be determined; it is only necessary to determine the numbers of gamma rays at certain energies selected to enable the fluid to be characterised. For example with an oil/water/gas mixture the gamma rays might be counted at energies corresponding to emissions from carbon, and from oxygen. Analysis of the fluid contents is simplest in cases where each phase of the fluid emits a different gamma ray. When a gamma-emitting element occurs in more than one phase it may be necessary to make use of known relationships between the phases, for example a known ratio of gamma emissions at a particular pressure.

The prompt gamma rays are preferably detected using a scintillation detector, such as NaI(Tl), or bismuth germanate. They are preferably detected during a prolonged irradiation period, so as to ensure good counting statistics, and this period is typically at least 15 minutes, and might be as long as several hours; these times presume a mean neutron emission from the tube 6 8 -1 in the range 10 to 10 sec , and for a more intense neutron source, shorter time periods might be used. The spectrum of gamma energies is hence related to the average concentrations of the relevant nuclei over the said prolonged time, so it will be appreciated that this approach minimises the effect of rapid fluctuations in concentrations. Furthermore the counting statistics are considerably improved by measuring over a longer period, as the numbers of gamma rays at each energy are thereby increased. The prolonged irradiation time may be achieved by operating the neutron tube continuously for the said time, or may comprise a plurality of periods of continuous irradiation separated by periods in which no irradiation occurs. Preferably the neutron tube operates in a pulsed mode throughout the prolonged irradiation, for example generating 1000 or 5000 pulses per second, each of duration 20 s or 25 Rs, and say 10 neutrons per pulse.

When the fluid is an oil/water/gas mixture the most significant elements in the fluid are hydrogen, carbon, oxygen and, if the water is saline, chlorine. The effects of neutron irradiation on these elements are indicated in Table 1. Those reactions requiring thermal neutrons are indicated by T after the value of the crosssection; those requiring fast neutrons are indicated by an F, and in these cases the stated cross-sections are those at 14 MeV.

Table 1 Nuclear Detection Reactions

Cross Gamma Gamma Decay Element Reaction Section Energy Fraction Half (mb) (MeV) (%) Life Hydrogen H(n,γ)H 332 T 2.223 100 prompt Carbon C(n,γ) C 3.4 T 4.945 68 prompt 3.684 26 prompt C(n,n'γ)C 250 F 4.439 100 prompt Oxygen 16O(n,p)16N 35 F 6.130 68 7.2 s 7.117 5 7.2 s 16O(n,n'7)16O 250 F 6.130 68 prompt 7.117 5 prompt Chlorine Cl(n,7) Cl 43000 T 6.111 20 - prompt 6.620 8 prompt 7.414 10 prompt 7.490 8 1 prompt The prompt gamma spectrum is rendered more complex by additional lines, which are a measurement artefact due to single escape and double escape, at 0.511 MeV less and at 1.022 MeV less than the values of gamma energy listed in the Table; there are also other emission lines not listed in the Table. All the lines appear above a background generated by Compton scattering of gamma rays of higher energy. An important source of background gamma rays are neutron reactions with the pipe wall, the detectors, and any shielding or other nearby materials.

The gamma spectrum thus contains gamma rays arising from both fast and thermal neutrons. However a finite time is taken by the fast neutrons from the tube to be thermalised. Hence if the neutron generator produces pulses of duration less than about 40 Rs then the gamma emissions due to fast neutrons can be distinguished from those due to thermal neutrons by means of a time gate: gamma rays detected during the neutron pulse (or within say 5 Fs of it) are those generated by fast neutrons, whereas gamma rays detected in the intervals between the neutron pulses are those generated by thermal neutrons.

This may enable gamma lines to be distinguished which might otherwise overlap, or be too close to resolve; and enables variations in the two gamma backgrounds to be distinguished from each other.

The neutron tube preferably generates fast neutrons of energy above 5 MeV, more preferably above 10 MeV. The tube may be provided with a moderator (preferably not containing any of the elements of interest in the pipeline) e.g. beryllium, to ensure the pipeline contents are irradiated by both fast and thermal neutrons. The fluid in the pipeline also acts as an effective moderator, although it is clear that its effect on the neutron spectrum within the pipeline depends on its composition: the liquid phases are much better moderators than the gas phase, and although oil and water have similar moderating properties the thermal neutron flux is strongly depressed in saline water by the large thermal neutron capture cross section of chlorine. It will also be appreciated that the cross-sections of the reactions whereby gamma rays are emitted vary with the neutron energy; for example the cross sections of the two oxygen reactions in the Table are significantly larger at 14 MeV than at 5 MeV.

The accuracy of the measurements can be further enhanced by monitoring the neutron flux, and normalising the detected gamma ray counts in accordance with the detected neutron flux. Hence any drift in the neutron source does not affect the measurements. The neutron detector might be a BF3 detector or a He detector, or a fast-neutron detector; it is preferably operated continuously, and measurements only taken after several minutes exposure to the neutrons, say 15 or 30 minutes, so that operation of the neutron detector and the neutron tube will have become stable.

In obtaining the spectrum of gamma rays, the signals from the detector are amplified, and then classified according to their signal height (which corresponds to the Y ray energy). The operation of the detector is desirably stabilised by observing a peak in the gamma spectrum, for example the out-of-pulse H line at 2.223 13? MeV or a gamma line from a gamma source e.g. Cs , and adjusting the gain of the amplifier so this peak is at a consistent energy.

The present invention also provides an apparatus for performing the method as aforesaid. The irradiation region may be a region of the pipeline, provided that means are available for controlling the contents of the pipeline; the method can thus be applied to a pipeline without modifying the pipeline in any way. However in some cases it is not acceptable to change the contents of a region of the pipeline for the requisite periods and at the necessary intervals. As an alternative the pipeline may be provided with a bypass system, the irradiation region being a region of either the pipeline or of the bypass system, and means being provided to control the contents of the irradiation region so it either contains the pipeline fluids or a standardised fluid.

In a preferred embodiment the pipeline is provided with two spaced apart branch junctions and diverter valves, the apparatus comprises a bypass pipe connectable to the branch junction, and the irradiation region is a region of the bypass pipe. The pipeline contents can thus be caused to flow through the bypass pipe for example for periods of one hour, alternating with periods of one hour in which the bypass pipe is empty. When the fluid has been characterised, the bypass pipe can be disconnected from the pipeline for use on another pipeline.

The apparatus may also include means to measure the flow rate of the fluids, comprising at least one gamma ray detector downstream of the irradiation region. For example there might be two such detectors at different distances downstream. The gamma rays signals they detect differ because of decay of the activated nuclei during passage of the fluid from one to the other, so enabling the velocity to be calculated. Alternatively there might be a detector at just one such downstream position, and the neutron source be operated so as to activate a discrete portion of the fluid by irradiating the fluid for a short period of time (for example between 0.5 and 2.0 s). The time at which gamma rays from activated nuclei are detected at the downstream position enables the velocity to be calculated. Where the fluid is an 16 oil/water/gas mixture the activated nuclei are N nuclei, which emit 6.13 MeV Y rays. The number of such y rays detected by a downstream detector enables the concentration of oxygen atoms in the fluid to be calculated, as a check on the concentration as determined from the prompt gamma ray spectrum.

The invention will now be further and more particularly described, by way of example only, and with reference to the accompanying drawings in which: Figure 1 shows a diagrammatic plan view of a multiphase flow monitor adjacent to a pipeline; Figure 2 shows a diagrammatic side view of part of the monitor of Figure 1; Figure 3 shows graphically prompt gamma spectra caused by thermal neutrons, obtained with a monitor as shown in Figure 2, with three different standard fluids; Figure 4 shows graphically prompt gamma spectra obtained as in Figure 3, but caused by fast neutrons; and Figure 5 shows a calibration graph for the monitor as shown in Figure 2.

Referring to Figure 1, a flow monitor 10 in a container 11 is arranged adjacent to a pipeline 12 carrying an oil/water/gas multiphase mixture. The pipeline 12 is provided with two spaced apart T-branches 13 and valves 14, and the flow monitor 10 is connected to the two T-branches 13 using demountable couplings 15.

The monitor 10 includes a bypass pipe 16 connected at each end via two isolation valves 18 to one or other of the demountable couplings 15, so the contents of the pipeline 12 can be arranged to flow through the bypass pipe 16. The bypass pipe 16 is provided with a fluid analyser 20 and a flow meter 22, each described in more detail in relation to Figure 2. Between the isolation valves 18 is a T-branch connecting via a valve 19 to a calibration rig 24 containing fluid pumps and fluid reservoirs (not shown).

Referring to Figure 2 the fluid analyser 20 includes a 14 MeV neutron generator 30 connected to a power supply unit 32. By way of example a suitable generator 30 is the Sodern GNT 02, in which the generator 30 is connected by a 15 m long cable to the power supply/control unit 32.

The generator 30 is arranged so as to irradiate the bypass pipe 16 and its contents with 14 MeV fast neutrons, via a lead collimator 34. In this example it generates 5000 neutron pulses a second, each of duration 20 Rs, and with 3 x 10 neutrons per pulse.

Adjacent to the part of the bypass pipe 16 irradiated by the neutron generator 30 are two diametrically opposed bismuth germanate scintillators 36 (only one is shown) with their longitudinal axes extending radially, in a plane at right angles to the neutron flux. The scintillators 36 are shielded from any direct gamma rays emitted by the neutron generator 30 by the collimator 34 and by a 150 mm thick lead shield 37, and are shielded from thermal neutrons by a thin boroncontaining sheet (not shown). The photomultipliers associated with the scintillators 36 are energised by a power supply 38, and provide signals to a pulse height multichannel analyser 40. This determines the pulse height for each signal (which corresponds to the energy of the detected gamma ray), and records the numbers of signals of each height. This is equivalent to a pulse height (or gamma energy) spectrum.

The pulse height multichannel analyser 40 receives timing signals from the power supply unit synchronised with the pulses of emitted neutrons. The analyser 40 incorporates a gate set to open 10 Rs before each neutron pulse and to remain open for 40 As. The analyser 40 stores the signals received while the gate is open separately from those while the gate is not open. The analyser 40 thus can provide two gamma energy spectra, one for gamma rays generated by the fast neutrons (received while the gate is open), and the other for those generated by thermal neutrons. If any hydrogenous fluid is present, the latter spectrum can be expected to include a well-defined peak at 2.22 MeV as a result of thermal neutron capture by hydrogen. The power supply 38 to the photomultiplier is controlled to ensure that that 13? peak (or a peak from a separate Cs source (not shown)) is observed at a consistent energy value.

The two spectra determined by the pulse height multichannel analyser 40 are supplied to a microcomputer 42. A boron trifluoride neutron detector 44 in a polyethylene sleeve is arranged adjacent to the neutron generator 30, and the number of detected neutrons is also supplied to the microcomputer 42. The microcomputer 42 normalises the gamma ray counts in accordance with the detected neutron count, for the counting period, which might be half an hour, in the energy windows of interest.

This is described in more detail in relation to Figure 4.

The flow meter 22 is several metres downstream of the irradiated region of the bypass pipe 16, and comprises six cylindrical 3 inch x 3 inch (75 mm x 75 mm) thallium-doped sodium iodide scintillators 48 arranged in a ring around and close to the bypass pipe 16 with their longitudinal axes parallel to the longitudinal axis of the bypass pipe 16, and each scintillator 48 having a photo-multiplier 50 at one end. The photomultipliers 50 supply electric signals via a single channel pulse amplitude selector 51, to a multichannel analyser operated in multichannel scaler (MCS) mode 52. The selector 51 allows only those signals corresponding to a gamma ray energy of 6.130 MeV to pass. The multichannel scaler 52 operates as a counter and timer recording the numbers of gamma counts received in successive intervals.

The scaler 52 may be arranged to start operation in synchronism with the neutron generator 30.

When it is desired to measure the flow rate, the neutron generator 30 is energised for one second, so as to activate nuclei in a discrete portion of the contents of the bypass pipe 16. The fast neutrons activate the water phase by the O(n,p) N reaction. As the fluid flows along the pipe 16 the activated nuclei decay with a half-life of 7.2 s; the gamma rays emitted during decay are detected by the scintillators 48 as the activated portion of fluid passes them. The data recorded by the multichannel scaler 52 are indicative of both the flow rate of the water, and also of the oxygen concentration.

The data is transmitted to the microcomputer 42.

Operation of the fluid analyser 20 involves irradiating the bypass pipe 16 with neutrons for a prolonged period, which might be 30 minutes or one hour, while detecting the prompt gamma rays; however to improve stability no measurements are made initially for example during the first half hour after the neutron generator 30 is switched on. As described above, the pulse spectrum analyser 40 determines two different prompt gamma ray spectra. Referring now to Figure 3, this shows three such spectra in the form of a graph of counts recorded in each channel in one hour, plotted against channel number, for thermal neutrons. The graphs were obtained with an empty pipe 16 (marked A), with the pipe 16 filled with saline water (marked B) and with the pipe 16 filled with kerosene (marked C). The higher channel numbers correspond to gamma rays of higher energy.

It is clear that at all energies the empty pipe A gives lower count rates than the fluid-filled pipes.

Considering the graph B with water present, there is a sharp peak at channel 63 (corresponding to the 2.22 MeV Y from hydrogen) and a broader peak at around channel 170 (about 6 MeV, corresponding at least partly to the 6.11 MeV Y from chlorine). Considering the graph C for kerosene, there is again a sharp peak at channel 63 (due to hydrogen), and there is also a broad peak around channel 170 - this is due to y rays from other elements in the pipe 16 (such as iron), whose emission is enhanced by the increased thermal neutron flux resulting from moderation by the kerosene.

Referring now to Figure 4, this shows three spectra plotted as in Figure 3, and for the same pipe contents, but for fast neutron irradiation. At all energies the empty pipe A gives the lowest count rates. Between channel numbers about 110 to 130 (around the 4.44 MeV emission from carbon) graph C with kerosene present shows the highest count rate compared to the other graphs, while between channel numbers about 140 to 180 (around the 6.13 MeV Y emission from oxygen) graph B with water present shows the highest count rates.

The microcomputer 42 can hence characterise the contents of the bypass pipe 16 by considering the gamma counts (corrected for the neutron flux) in two energy windows P and Q: window P in this example is channel numbers 115 to 126, and window Q is channel numbers 145 to 176.

Operation of the monitor 10 first requires calibration, which may be carried out before the monitor 10 is installed. Measurements are made as described above of the counts in the carbon window P and in the oxygen window Q, all the counts being recorded for an appropriate time (for example half an hour) and being normalised to a standard neutron count from the detector 44. Measurements are made with an empty pipe 16, and with a range of values of water content and oil content (and of both water and the oil) up to 100% water, and up to 100% oil, using an oil such as kerosene. Measurements with an empty pipe 16 (referred to as baseline) are repeated between the other measurements, and if there is any change in the baseline counts, then the other measurements are changed by the same amount. Thus account is taken of any drift in the baseline.

The monitor 10 is then installed as described in relation to Figure 1, and the valves 14 may be operated to divert the fluid flow from the pipeline 12 through the bypass pipe 16. The counts in the carbon window P and the oxygen window Q can hence be determined in the standard time, and normalised to the standard neutron count. However, at intervals during operation the fluid flow is returned to the pipeline 12 and the bypass pipe 16 is emptied, and the baseline counts obtained in the carbon window P and the oxygen window Q. If the baseline counts are different from those obtained originally during calibration, the counts obtained with the fluid flow present are altered by the same amounts, to produce counts which are corrected for baseline drift. The corrected counts in the windows P and Q, by comparison with the corrected counts measured during calibration, enable the quantity of water and of oil (which corresponds to kerosene) in the pipeline 12 to be determined.

Referring to Figure 5, this illustrates one way in which this analysis may be performed, the graph showing the corrected counts in the carbon window P plotted against those in the oxygen window Q (in each case measured in 20 minutes) during calibration with a range of different values of volume fraction of water (w) and of kerosene (k) up to 0.6. The measured counts are plotted using symbols and the lines show how the counts vary with k for constant w, and with w for constant k.

Using such a graph the counts in the windows P and Q obtained with unknown pipe contents can be related to the corresponding volume fractions k and w of oil and water.

The volume fraction of gas must be such that the total of all the volume fractions is one. If the pressure in the pipeline 12 is so high that the gamma emission from the gas (say methane) is not negligible, then the water fraction w and the parameter k are found in the way described; the parameter k is the volume fraction of oil which would be equivalent to the oil and gas actually present, and at a particular pressure and temperature the ratio of the densities, and so the ratio of the gamma emissions, of oil and methane is a known constant, so the volume fractions of oil and gas can therefore be found.

The fluid pressure (and temperature) in the pipeline 12 may be known, or may have to be measured within the monitor 10.

It should be appreciated that drift in the baseline can be allowed for even if it is not possible to empty the bypass pipe 16, as long as it can be filled with a fluid of known composition, for which the expected counts in the windows P and Q are already known from, or can be derived from, the calibration measurements. For example the bypass pipe 16 might at intervals be filled with kerosene, and the counts in the windows P and Q measured.

If these counts differ from the corresponding counts obtained during calibration, then this difference can be assumed to be equal to the baseline drift, which is the correction to be applied to counts obtained with a fluid of unknown composition.

It will be appreciated that in characterising the fluid in the pipeline 12 it is also possible to measure the total count in the hydrogen window, that is to say in the thermal neutron spectrum, say channels 60 to 65. If the monitor 10 is also used to measure flow velocity, the number of gamma rays recorded by the counter 52 may be related to the oxygen concentration (although it is also affected by the flow rate).

It will also be appreciated that a fluid flow monitor may differ from that described above while remaining within the scope of the invention. For example the prompt gamma rays may be detected by any type of gamma sensor which gives an energy-dependent signal. In particular they may be detected by a high resolution gamma spectrometer such as a high purity germanium detector, or a different scintillator, for example a thallium-doped sodium iodide scintillator. Furthermore the pulse spectrum analyser 40 might not incorporate the time gate to distinguish between gammas generated by fast and thermal neutrons; and might count only those gammas in the carbon window P and the oxygen window Q rather than counting over the complete gamma spectrum.

Additional sensors may also be provided, such as a pressure sensor, or a gamma densitometer, to enhance or extend the data or its interpretion.

There are some situations where the provision of branch ducts 13 is not acceptable. In this case the invention may be applied by arranging the fluid analyser 20, and the flow meter 22 if required, adjacent to the pipeline 12 itself; it is still necessary to calibrate the monitor by measurements with the pipeline 12 empty, and full of water, and full of a carbon-containing fluid.

During use of the monitor it is still necessary to take account of any baseline drift by taking measurements, at intervals, with a fluid of known composition for which calibration counts are available. It may for example be acceptable to pass water through the pipeline 12 at intervals, or methanol, and in the latter case it would therefore also be necessary to take calibration counts with methanol filling the pipeline 12.

Claims

1. A method for monitoring the contents of a pipeline carrying a fluid, the method comprising irradiating the fluid at an irradiation region with fast neutrons from a fast neutron generating tube, detecting prompt gamma rays emitted from the irradiation region during the said irradiation, and providing first signals characterising the energy spectrum of the prompt gamma rays, at intervals controlling the contents of the irradiation region to contain a standardised fluid, performing the irradiation with neutrons and the detection of the prompt gamma rays as aforesaid and hence providing second signals characterising the energy spectrum of the prompt gamma rays in the presence of the standardised fluid, and from the first signals and the second signals determining characteristics of the fluid in the pipeline.

2. A method as claimed in Claim 1 wherein the prompt gamma rays are detected using a scintillation detector.

3. A method as claimed in Claim 1 or Claim 2 wherein the fast neutrons are generated in brief pulses of duration no more than 40 As, and wherein prompt gamma rays generated by fast neutrons are distinguished from those generated by thermal neutrons by means of a time gate.

4. A method as claimed in any one of the preceding Claims also comprising monitoring the neutron flux in the vicinity of the irradiation region, and normalising the first and the second signals in accordance with the measured neutron flux.

5. A method as claimed in any one of the preceding Claims wherein the fluid comprises oil and/or water and/or gas, and wherein the first and the second signals each represent the numbers of gamma rays in two energy windows around 4.44 MeV and around 6.13 MeV.

6. Apparatus for monitoring the contents of a pipeline carrying a fluid, the apparatus comprising means to irradiate an irradiation region with fast neutrons from a fast neutron generating tube, means to detect prompt gamma rays emitted from the irradiation region during the said irradiation, and means to provide signals characterising the energy spectrum of the prompt gamma rays, the apparatus also comprising means to occupy the irradiation region with the pipeline fluid, and means to control the contents of the irradiation region to contain a standardised fluid, and analysis means to characterise the pipeline fluid from the spectrum-characterising signals obtained with the pipeline fluid and with the standardised fluid.

7. Apparatus as claimed in Claim 6 wherein the apparatus includes a bypass system through which the pipeline fluid may be caused to flow.

8. Apparatus as claimed in Claim 7 wherein the irradiation region is in the bypass system.

9. Apparatus as claimed in any one of Claims 6 to 8 also comprising a monitor to monitor the neutron flux in the irradiation region and to provide signals to the analysis means representing the neutron flux.

10. Apparatus as claimed in any one of Claims 6 to 9 wherein the neutron tube generates neutron pulses of duration no more than 40 ps, and wherein a time gate is provided to distinguish between prompt gamma rays generated by fast neutrons from those generated by thermal neutrons.

11. Apparatus as claimed in any one of the preceding claims wherein the gamma detector means is a scintillator with a photomultiplier, and wherein the gain of the photomultiplier is stabilised by monitoring the energy at which a peak is observed in the gamma spectrum.

12. A method for monitoring the contents of a pipeline substantially as hereinbefore described with reference to, and as shown, in the accompanying drawings.

13. Apparatus for monitoring the contents of a pipeline substantially as hereinbefore described with reference to, and as shown, in the accompanying drawings.