Login| Sign Up| Help| Contact|

Patent Searching and Data


Title:
A LOW-FIELD NMR DEVICE FOR MEASURING THE WATER CONTENT OF SOLIDS AND SLURRIES
Document Type and Number:
WIPO Patent Application WO/2012/028785
Kind Code:
A1
Abstract:
The invention relates to a Nuclear Magnetic Resonance (NMR) apparatus and method for measuring the water content of samples. The apparatus comprises means for producing a main magnetic field; a sample receiving space within said main magnetic field; means for exciting a measurable RF magnetization to a sample placed into said sample receiving space at an operating frequency defined by said main magnetic field; means for measuring the RF signal produced by the excited sample; and means for determining the water content in the sample based on the RF signal. According to the invention, the sample receiving space is capable of accommodating a sample having a volume of at least 0.5 dm3, and the means for producing a main magnetic field comprise a resistive electromagnet which is adapted to produce a main magnetic field corresponding to an operating frequency of 400 - 2000 kHz. By means of the invention more cost-efficient and lightweight NMR apparatuses for humidity measurement purposes can be built.

Inventors:
VIRTANEN SAMI (FI)
VIITANEN VELI-PEKKA (FI)
Application Number:
PCT/FI2011/050753
Publication Date:
March 08, 2012
Filing Date:
August 30, 2011
Export Citation:
Click for automatic bibliography generation   Help
Assignee:
METSO AUTOMATION OY (FI)
VIRTANEN SAMI (FI)
VIITANEN VELI-PEKKA (FI)
International Classes:
G01N24/08; G01N33/44
Domestic Patent References:
WO2001014847A22001-03-01
WO2003006974A12003-01-23
Foreign References:
US20060112628A12006-06-01
US20090140736A12009-06-04
EP1006366A12000-06-07
Other References:
SCHMIDT, B. C. ET AL.: "Quantitative determination of water speciation in aluminosilicate glasses: a comparative NMR and IR spectroscopic study", CHEMICAL GEOLOGY, vol. 174, pages 195 - 208, XP027214081
BELTON, P. S. ET AL.: "Pulsed NMR studies on water in striated muscle III. The effects of water content", BIOCHIMICA ET BIOPHYSICA ACTA, vol. 354, July 1974 (1974-07-01), pages 304 - 314, XP025835990
KANTZAS A. ET AL.: "Low field NMR applications in oil sands mining and extraction", INTERNATIONAL SYMPOSIUM OF THE SOCIETY OF CORE ANALYSTS, August 2005 (2005-08-01), TORONTO, CANADA, pages 1 - 12, XP055082100, Retrieved from the Internet
See also references of EP 2612135A4
THE BRITISH JOURNAL OF RADIOLOGY, vol. 71, 1998, pages 704 - 707
BARALE, P. J. ET AL.: "The Use of a Permanent Magnet for Water Content Measurements of Wood Chips", IEEE TRANSACTIONS ON APPLIED SUPERCONDUCTIVITY, vol. 12, no. 1, March 2002 (2002-03-01)
BUTLER JAMES ET AL.: "Using low-field MRI to improve tablet dissolution testing", TABLETS & CAPSULES, January 2010 (2010-01-01)
KANTZAS A. ET AL.: "LOW FIELD NMR APPLICATIONS IN OIL SANDS MINING AND EXTRACTION", INTERNATIONAL SYMPOSIUM OF THE SOCIETY OF CORE ANALYSTS, 21 August 2005 (2005-08-21)
HUCHSON J M S ET AL.: "Journal of physics E. scientific instruments", vol. 13, 1 September 1980, IOP PUBLISHING, article "A whole-body NMR imaging machine", pages: 947 - 955
R.R. RUAN; P.L. CHEN: "Water in Foods and Biological Materials", 1998, CRC PRESS
Attorney, Agent or Firm:
KOLSTER OY AB (P.O.Box 148, Helsinki, FI)
Download PDF:
Claims:
Claims

1. A Nuclear Magnetic Resonance (NMR) apparatus for measuring the water content of samples, such as solids and slurries, comprising

- means for producing a main magnetic field, - a sample receiving space within said main magnetic field,

- means for exciting a measurable RF magnetization to a sample placed into said sample receiving space at an operating frequency defined by said main magnetic field,

- means for measuring the RF signal produced by the excited sample, and - means for determining relative water content in the sample based on the RF signal, characterized in that

- the sample receiving space is capable of accommodating a sample having a volume of at least 0.5 dm3, and

- the means for producing a main magnetic field comprise a resistive electromagnet which is adapted to produce a main magnetic field corresponding to an operating frequency of 400 - 2000 kHz.

2. The apparatus according to claim 1, characterized in that the electromagnet is passively cooled.

3. The apparatus according to claim 1, characterized by further comprising means for actively cooling the electromagnet, preferably by forced air circulation.

4. The apparatus according to any of claims 1 - 3, characterized in that the operating frequency of the device is 400 - 950 kHz.

5. The apparatus according to any of claims 1 - 3, characterized in that the operating frequency of the device is 950 - 2000 kHz.

6. The apparatus according to any of the preceding claims, characterized by further comprising means for measuring the mass of the sample while the sample is in the sample receiving space, said means for determining the water content of the sample being adapted to utilize the mass of the sample in determining the relative water content of the sample. 7. The apparatus according to any of the preceding claims, characterized in that

- said means for measuring the RF signal are adapted to start measuring of the RF signal after a predetermined dead time after the excitation, and

- said means for determining the water content of the sample are adapted to

extrapolate the RF signal value to the time at the excitation pulse based on the measured RF signal and adapted to utilize said extrapolated signal value in determining the water content of the sample.

8. The apparatus according to any of the preceding claims, characterized in that the volume of the sample receiving space is 0.5 - 5 dm .

9. An NMR-based method for measuring the water content of a sample, comprising

- producing a main magnetic field,

- subjecting the sample to said main magnetic field in order to produce a net

magnetization to the sample,

- exciting a measurable RF magnetization to a sample at an operating frequency defined by said main magnetic field, - measuring the RF signal produced by the excited sample, and

- determining the water content in the sample based on the RF signal, characterized by

- using a sample having a volume of at least 0.5 dm3, and - producing said main magnetic field using a resistive electromagnet adapted to produce a main magnetic field corresponding to an operating frequency of 400 - 2000 kHz.

10. The method according to claim 1, characterized in that the sample is a biomass sample.

1 1 . The method according to claim 1, characterized in that the sample is in a solid or slurry form.

12. The method apparatus according to any of claims 9-11 , characterized by using a passively cooled electromagnet and an operating frequency of 400 - 950 kHz. 13. The method according to any of claims 9-11, characterized by using an actively cooled electromagnet and an operating frequency of 950 - 2000 kHz.

14. The method according to any of claims 9- 13. characterized by measuring the mass of the sample while the sample is in the sample receiving space and determining the water content of the sample based on the mass of the sample. 15. The method according to any of claims 9-14, characterized by

- measuring of the Rl signal only after a predetermined dead time after said

excitation,

- estimating the RF signal value at the time of the excitation pulse based on the measured RF signal, and - utilizing said estimated RP signal value in determining the water content of the sample.

Description:
A low-field NMR device for measuring the water content of solids and slurries Field of the Invention

The invention relates to Nuclear Magnetic Resonance (NMR) measurement devices and methods, In particular, the invention relates to a NMR device and method for measuring water content of solids and slurries.

Background of the Invention

NMR measurements are based on providing a net magnetization to a large group of atomic nuclei using a static magnetic field (main field) and deflecting the net magnetization from the direction of the static magnetic field using a radio frequency pulse (RF pulse) of a radio frequency magnetic field (RF field), the operating frequency (Larmor frequency) being defined by the nuclei concerned and the magnitude of the main field. The relaxation of the deflected precessing net magnetization can be detected by measuring the NMR signal, i.e. the EMF induced in the receiving RF coils caused by the precessing net magnetization, which is gradually relaxing back to parallel with the main field (known also as Free Induction Decay, FID)). The relaxation speed is determined by the homogeneity of the main field and the properties of the matter under measurement.

Water content of various material samples can be measured accurately and rapidly using NMR spectroscopy or relaxometry. Wide usage of NMR based moisture content measurement devices has been hindered mainly by the high cost of measurement devices. In particular, in many applications, e.g. biomass water content measurements, the desired sample volume is of the order of several deciliters or larger, which sets practical limits for device dimensions and other specifications.

Prior art [e.g. Remote Automatic On-Line Sensor; Final Report. Quantum magnetics Inc] suggests that a minimum operating frequency of 5 to 6 MHz should be used in NMR-based water content measurements in order to maintain a reasonable recovery time, i.e. receive circuit deadtime following the transmitted RF pulse. This necessitates the use of a relatively high main magnetic fields (> 125 m l ). High field also increases the measurable signal amplitude. Other prior art [The British Journal of Radiology, 71(1998), 704-707] also teaches that the electromotive force induced in the receiving coils by the NMR signal is essentially proportional to the square of the main magnetic field strength. However, the use of field strengths of this magnitude increases device costs, power consumption and magnet mass, for example. In addition, the temperature variation and main field inhomogeneity effects are significant in high fields, further complicating device design. Also safety aspects, as concerns e.g. a stray magnetic field outside the device, become more relevant in high magnetic fields. The above aspects are emphasized if the desired sample volume is large. Barale, P. J. et al, "The Use of a Permanent Magnet for Water Content Measurements of Wood Chips", IEEE TRANSACTIONS ON APPLIED

SUPERCONDUCTIVITY, VOL. 12, NO. 1, MARCH 2002, present that the water content of wood chips can be measured at a 0.47 T main field, yielding for the operating frequency 20 MHz. The permanent magnet used weighed 68 kg although the possible sample volume ("good field" volume) was only less than 10 ml. Other examples of prior art include water content analysis in dissolution testing, for example as presented by Butler James et al, " Using low-field MRI to improve tablet dissolution testing", Tablets & Capsules, January 2010 at 0.5 T main field and small sample volume.

There have also been attempts to use lower frequencies and field strengths. For example, Kantzas A. et al, "LOW FIELD NMR APPLICATIONS IN OIL SANDS MINING AND EXTRACTION", International Symposium of the Society of Core Analysts, Toronto, Canada, 21-25 August 2005, have used a 1 MHz Corespec 1000™ (main field 24 liiT produced by permanent magnet) relaxometer apparatus to determine the oil and water content in 20 ml ore and froth samples. It is relatively easy to produce the required field strength with a good homogeneity over small sample volumes with permanent magnets, as in the prior art referred to above. Large sample volumes are, however, required in many applications, because of sampling standards, as well as the generally large particle dimensions e.g. of biomass samples. The device designs referred to above are, however, not suitable for measuring large sample volumes since the device designs would turn out to be expensive and difficult to manufacture if scaled up to the required size. Summary of the Invention

It is an object of the invention to provide a novel NMR-based water content measurement apparatus and method suitable for the measurement of large sample volumes (0.5 1 or more). In particular, it is an aim to provide an apparatus. A further aim is to provide an apparatus which can be designed to be lightweight.

The invention is based on the idea of utilizing a measurement volume of at least 0.5 litres and a low operating frequency of 400 - 2000 kHz, which have, according to the invention, found to be achievable using a resistive electromagnet. More specifically, the invention is defined in the independent claims. According to one embodiment, the apparatus comprises

- a sample receiving space capable of accommodating a sample having a volume of at least 0.5 dm 3 ,

- a resistive electromagnet for producing a main magnetic field over the entire

sample receiving space at a field strength of 10 - 40 m l " . - an RF coil for exciting a measurable precessing transverse magnetization to a sample placed into said sample receiving space at a L armor frequency defined by said main magnetic field,

- means for measuring the RF signal produced by the excited sample, preferably using the same RF coil which is used for excitation, and - a computing unit for determining the water content in the sample based on the RF signal.

The method comprises

- resistively producing a main magnetic field over a sample having a volume of at least 0.5 dm 3 , - subjecting the sample to said main magnetic field in order to produce a net

magnetization to the sample, - exciting a measurable RF magnetization to a sample at an operating frequency of 400 - 2000 kHz,

- measuring the RF signal produced by the excited sample, and

- determining the water content in the sample based on the RF signal. The sample can be a biomass sample typically in a solid or slurry form.

According to one embodiment, a passively cooled electromagnet is used for producing the main magnetic field, the operating frequency preferably being 400 - 950 kHz in order for the amount of heat produced being sufficiently low.

According to one embodiment, an actively cooled electromagnet is used, whereby the operating frequency can be as high as 950 - 2000 kHz.

According to one embodiment the mass of the sample is measured while the sample is in the sample receiving space and the relative water content of the sample is determined partly based on the mass of the sample.

According to one embodiment, the RF signal is measured only after a predetermined dead time after the excitation pulse in order for the measurement electronics to recover from the excitation pulse. The assumed RF signal value at the time of the excitation pulse is estimated based on the measured RF signal by extrapolation and said estimated RF signal value is used in determining the water content of the sample. By using such measurement sequence and signal processing algorithm the utilization of such large measurement volume and low magnetic field (low I . armor frequency) can be used, contrary to the expectations of prior art. It is an advantage of the invention that the ability to measure also tight bound water in dry biomass samples is not compromised. As concerns this problem, we refer to "Water in Foods and Biological Materials", R.R. Ruan, P.L. Chen, CRC Press 1998. Typically, the dead time referred to above is 30 - 200 LIS.

The preferred operation frequency of the apparatus is 400 - 1700 kHz. The invention provides significant advantages. First, the device can be designed to be small and cost-effective compared with NMR devices based on permanent magnets or superconducting magnets. Second, it extends the measurement volume of small-scale devices to the deciliter range while maintaining the ability to determine the water content in the sample. Thus, for example the water content of biomasses and biofuels can be measured conveniently.

Is should be noted that in the case of small sample volumes, the temperature of the main coil is not a limiting factor as the resistance of the coil is typically low and thus also the power dissipation of the coil is at low level. However, it has been found that with volumes over 0.5 dm 3 , the present operating range is desired as the temperature of the coil remains at an acceptable level.

Additional advantages obtained by means of the ability to use low main magnetic field and thus low frequency are

- low power consumption,

- low magnet temperature,

- low magnet mass.

- no or low temperature dependence of magnetic field,

- ability to use an electromagnet for producing a homogeneous magnetic field

without any shimming coils,

- safe level of stray field,

- more slack relative uniformity requirement for the magnetic field, lower cost of amplifiers, AD-converters, power supplies etc.

In summary, a more cost-effective, lightweight and safe-to-use device can be

manufactured.

According to a preferred embodiment, the measurement system comprises integral sample mass measuring means. Preferably, the weight of the sample can be measured while the sample is in the sample receiving space of the NMR device, i.e. in the NMR signal measurement position. The sample weight measured is utilized together with the extrapolated NMR signal in determination of the water content of the sample.

The term "sample receiving space" means, in particular, a zone within the main magnetic field having a field strength of 10 - 40 mT and field homogeneity better than 1000 ppm, preferably better than 250 ppm. Typically there are provided holding means in the main magnet for holding a sample vessel such that the sample is located within the sample receiving space. Depending on the shape of the main coil, the sample receiving space can have various shapes. The term "resistive electromagnet" refers to a coil typically wound from a metal conductor which is in resistive state, in contrast to superconducting state, at the operating temperature of the apparatus (typicall room temperature).

Further advantages and embodiments of the invention are discussed in the following detailed description with reference to the attached drawings. Brief Description of the Drawings

Fig. 1 shows a measurement system according to one embodiment of the invention.

Fig. 2a presents graphically typical NMR signals with their relaxation times.

Fig. 2b illustrates as a graph extrapolation of NMR signal according to one embodiment of the invention. Fig. 3 presents a typical pulse sequence in accordance with the invention.

Fig. 4 shows a temperature/power vs. resonance frequency curve of aluminum coil electromagnet.

Detailed Description of Embodiments

A measurement system according to one embodiment is shown in Fig. 1. The NMR unit is denoted with the reference numeral 1 19 and comprises a main DC electromagnet 18 and an RF coil 120 placed inside the main magnet 1 18. The main magnet 120 is powered by a DC power suppl 126. The system also comprises a control and data acquisition computer 102. The computer 102 is connected via a control signal and data transmission channel 134 to an ADC and DAC converter 106. The RF coil 120 is connected to a directional switch 1 12 which is used for transmitting both the excitation signal from the ADC/DAC 106 to the RF coil 120 and the NMR signal from the RF coil 120 to the ADC DAC 106. The excitation signal 132 is transmitted through an RF power amplifier 110 powered by a suitable power supply 108. and the NMR signal is received via input amplifiers and low pass filters 1 14. 1 16 The weight of the sample placed inside the RF coil is measured using a load cell 122 connected via a load cell amplifier 124 to the ADC/DAC 106. Power for the separate units of the system is provided using a general power supply 102 or, as discussed above, using separate power supplies such as usually required by the main magnet and the RF amplifier 1 10. Power lines are drawn an dashed lines in Fig. 1, whereas the control/data lines are drawn as dash/dot lines. The RF transmit/receive signal lines and the mass weight signal lines are drawn as solid lines.

As discussed above, the main magnet 118 is adapted to produce a magnetic field corresponding to a I .armor frequency of 400 - 2000 kHz of protons in the sample, in practice, the field strength should be about 9 - 44 mT at the sample receiving zone within the RF coil 120. The main magnet is preferably an electromagnet with a winding scheme adapted to produce as homogeneous field at the sample receiving zone as possible.

According to one embodiment, the main magnet 1 18 is wound from aluminum conductor.

The main magnet is preferably a resistive electromagnet which is passively or actively cooled. Passive cooling in this context means that heat is dissipated from the main magnet only through natural radiation, convection and conduction from the magnet material to its surroundings. Active cooling may take place in the form of forced fluid circulation, such as forced air circulation.

The NMR signal frequency is directly proportional to the main magnetic field, the proportionality coefficient being the gyromagnetic ratio. On the other hand, the magnetic field of an electromagnet is directly proportional to the electrical current flowing in the magnet coil. Further, the electrical power required to drive the current is proportional to the coil resistance multiplied by the square of the current. In practice, the coil resistance increases almost linearly with the coil temperature, which in turn increases with the increasing current; thus the power consumption of the coil is practically a steeper function of the current than the square of the current.

The RF coil 120 is adapted to produce a magnetic field perpendicular to the main magnetic field. The RF coil 120 can be of a birdcage type. The size of the RF coil 120 is sufficient to

3 ^ accommodate a sample having a volume of at least 0.5 dm , preferably 0.5 - 5 dm . In particular, the sample receiving zone inside the RF coil can be shaped to be cylindrical, but other shapes are possible too.

The sample weighing device 122 can be placed inside or outside the NMR unit 1 19.

Preferably, it is placed below the NMR unit 1 19 and the dead load of the NMR unit is taken into account by suitable calibration or computationally.

In a humidity measurement, a homogeneous DC magnetic field is generated by the main magnet into the sample to be measured, then interaction of the magnetic field with the hydrogen in the sample causes a small magnetization to develop in the sample. Next, with reference to Figs. 2 and 3, the sample is exposed to a short intense radio frequency (RF) excitation pulse 3 by the RF coil, which excites the hydrogen nuclei. In the following step the NMR signal is recorded by the RF coil, typically for a period of milliseconds. During this time, the sample undergoes NMR relaxation and returns to the original magnetization state. The signal amplitude 10 (Fig. 2) is proportional to the total amount of hydrogen from moisture of the samples. However, for practical reasons, the recordation cannot be started immediately after the excitation pulse 3, but only after a predetermined dead time (of the order of tens of microseconds following the first RF pulse, typically 50-200 microseconds). However, it is the maximum value of the NMR signal that defines the moisture content, whereby this maximum value 10 is extrapolated from the NMR signal recorded after the dead time. The NMR signal arising from the protons of the solid matter in the sample decays in <50 microseconds, and thus advantageously does not affect the definition of moisture in the sample, as described above.

In summary, according to a preferred embodiment, the NMR signal is measured after a predefined period of dead time after the excitation pulse and an algorithm is used that utilizes the signal data measured after said dead time and extrapolates the signal to zero time (at the time of the excitation pulse). On the basis of the extrapolated signal, the water content of the sample is determined. Extrapolation methods known per se can be used.

The decay of the signal as illustrated in Fig. 2 is determined mostly by the T2, i.e. the spin- spin relaxation time, of the sample. The present NMR device is especially well suited for measuring the water content in biomass. When the sample to be measured is very dry, typically meaning water content of less than 20 m-%, the signal-to-noise -ratio is low, which can be compensated for by increasing the number of successive measurements and averaging them. This easily leads to a long measurement time. The limitation for the time between successive measurements is primarily set by time factor Tl, i.e. the spin-lattice relaxation time. This is the time required for the deflected average magnetization vector to recover its original value. The recovery is enabled by energy dissipation from the protons to the lattice. If the excitation pulse is applied before full relaxation, reduced signal amplitude is observed, and the correlation coefficient between the water content and the signal amplitude is altered, and thus calibration will not be valid.

Low magnetic field and low Larmor frequency enable the construction of a large sample volume measurement system that has much lower mass, power consumption and cost than what can be expected based on the prior art.

According to one embodiment, the RF signal is measured only after a predetermined dead time after the excitation pulse in order for the measurement electronics to recover from the excitation pulse. For obtaining the most accurate estimate for the water content, the assumed RF signal value at the time of the excitation pulse is extrapolated based on the measured RF signal. This can be done following the principles illustrated in Figs. 2a and 2b. Referring to Fig. 2b, free induction decay signal amplitude 23 is very small compared to the original 90 degree excitation pulse 21. In order to record data without serious interference, it is necessary to wait until the excitation pulse as well as the noise and ringing 22 it has induced in the measurement electronics is attenuated to a harmless level, or otherwise utilize only data recorded after t=tl . Free-induction decay signal attenuation can advantageously be described by an exponential

S = A0 - e n

or Gaussian function

S = A0 e ^

Assuming that tl-tO > 50 microseconds, which is generally true for NMR devices operating below 2 MHz Larmor frequencies, the signal arising from solids is not detectable and the water content of the sample is proportional to extrapolated amplitude AO at t=t0. Since the spin-spin relaxation time T2 is strongly dependent on the material and its water content, one cannot directly use the amplitude value at t= 11 to determine the water content.

In order to find out the value of AO, it is necessary to mathematically extrapolate the envelope 24 of the signal amplitude recorded at t> tl backwards to tO, advantageously by fitting the abovementioned functions to the data at t>tl .

It should be noted that such problem relating to long dead time does not occur if the sample volume is small, especially if combined with a high Larmor frequency as in the prior art. This is because the RF power needed for exciting a small sample is low and the excitation pulse, including unwanted transients and ringing, can be short, whereby the dead time of the receiver is also very short. Also the inductance of the RF coil can be kept low with small samples even if its gain were high, which aids in minimizing the dead time of the circuitry. However, in the case of large sample volumes, the dead time inevitably becomes a significant factor, which can be taken into account using the algorithm shortly described above. The dead time can naturally be shortened by significantly reducing the Q- value of the RF-coils, but this would degrade the signal-to-noise ratio below acceptable levels.

It has been found by the inventors that the upper l imit of the present frequency range arises from the thermal limitations of a passively cooled or air-cooled electromagnet. For a passively cooled magnet with 6 kg of aluminum conductor, a magnetic field exceeding 18 ni corresponding to Larmor frequency exceeding 950 kHz leads to magnet surface temperatures that are above the acceptable values both for the sample as well as for the user. Table 1 below lists ergonomics data to establish temperature limit values for hot surfaces according to the guideline EN 563. Forced air cooling may increase the usable magnetic field/frequency range up to about 35 mT/'l 700 - 2000 kHz.

Table 1. EN 563: 1994 Safety of machinery - Temperatures of touchable surfaces - Ergonomics data to establish temperature limit values for hot surfaces

It should also be noted that the temperature aspect is not relevant for small sample volume devices, because permanent magnets can be used or the required current densities in the main coil are significantly lower, resulting in low heat dissipation.

The lower frequency limit, on the other hand is set by two different phenomena. First, the detected signal amplitude is approximately proportional to the square of the magnetic field strength. Thus, the S/N-ratio decreases rapidly when the Larmor frequency decreases. S/N- ratio is a significant problem when measuring the water content of dry samples with a low magnetic field producing a low Larmor frequency, as is the case in the invention.

Secondly, the dead time after excitation is basically inversely proportional to the Larmor frequency. When the frequency drops to one half, the time required for the electronics and the RF-coil to recover from the excitation pulse doubles (basically same number of attenuation cycles), and exceeds manyfold the spin-lattice relaxation time constant of samples with tightly bound water, thus limiting the measurement range to wet fuels. The dead time problem is aggravated by the large sample volume needed, which leads to large inductance of the excitation/receiving coils and thus slow recovery. Below the present lower frequency limit of 400 kHz, the dead time is too long for practical purposes, and the accuracy of the signal extrapolation and thus the accuracy of the measurement is compromised.

Integrating a weighing device into the system is preferred so as to be able to express the moisture content as the ratio of water mass and total mass. Typically in the prior art. this has been achieved by measuring also the NM signal emanating from the solids in the sample. This is a less accurate method, and furthermore impossible to realize with the low frequencies used in this invention.