Presentation & Papers
Presentation & Papers
Presentation & Papers
iris-instruments.com
Range
RangeofofMagnetic
MagneticResonance
Resonancesystems
systems
MAGNETIC RESONANCE
SOUNDING SYSTEMS
DIRECT DETECTION
OF GROUNDWATER
water content
permeability estimate
depth of water layers
Determination of water level and quantity
Lateral extension of an aquifer layer
Selection of the best place to drill
Prediction of yield, after calibration
Interpretation:
Porosity vs depth
Field curve:
Signal vs pulse
Our offer:
Sale of products
Interpretation software
Rental
On-site training
Maintenance
iris-instruments.com
Loop wire
reel
Main unit
Tuning
PC
Converter
Batt
Receiver
number of
DC/DC converters
loop
dimensions
100m
one
100x100m
150m
two
150x150m
Main
Battery
Converter
Loop
Loop
Switch
unit
Transmitter
Tuning
deep
aquifer
two
aquifers
medium
aquifer
thick
aquifer
multilayer
aquifer
E0
I
T2*
t
signal (E0)
signal
signal
signal
signal
signal
Pulse moment
Pulse moment
Pulse moment
Pulse moment
Pulse moment
Pulse moment
Main unit
iris-instruments.com
Loop wire
INFORMATION OBTAINED
The interpretation of measurements permits to estimate the
water content and the mean pore size (permeability) of each
layer at depth. These parameters are useful to determine the
prospects of a groundwater reservoir before drilling.
Tx / Rx
LOOP
MRS
Bo
Static field
Mo
Proton
magnetic
moment
H proton
ORDERS OF MAGNITUDE IN MRS
The Larmor frequency: it is determined by
the amplitude of the local Earth magnetic field
and varies between 0.8 and 3.0 kHz.
The loop size: the side of the loop is of the
order of the maximum investigation depth:
typically, 50, 100 or 150 m.
The energizing pulse: its moment controls the
depth of investigation. To reach 150 m,
currents up to 400 A and voltages up to 4000 V
are required during a few tens milliseconds
(typically 40 ms).
The voltage measured in the loop is in the
order of a few tens to a few thousands
nanovolts. Stacking is used to enhance these
small amplitude signals.
The duration of a sounding: depending on
the signal to noise ratio, a complete sounding
Converter & Tuning
(16 pulse moment values) can
take one to two
hours.
Larmor
TIME CONSTANTS
T2*: transverse time constant,
related to the component of the
proton
magnetic
moment
M0
perpendicular to the Earth magnetic
field
(one
pulse
technique
measurement)
T1: longitudinal time constant,
related to the component of the
proton
magnetic
moment
M0
parallel to the Earth magnetic field
(double
pulse
technique
measurement).
The time constants are linked to the
permeability through the mean
pore size: when the pores are large,
the protons loose slowly their energy
against the grains as shocks are
less frequent than when the pores
are small.
amplitude
vs time
for a given
pulse moment
-- E0
Maininitial
unit amplitudes E0
vs pulse moment
water content
vs depth
relaxation curves
T2*
sounding curve
interpretation section
iris-instruments.com
1000
magnetic inclination 45
SHALLOW
AQUIFER
DEEP
AQUIFER
TWO
AQUIFERS
MEDIUM
AQUIFER
THICK
AQUIFER
MULTILAYER
AQUIFER
d = 20m
750
d
10m
10m 20%
20%
500
d = 50m
d = 80m
signal
250
signal
signal
signal
signal
signal
d = 100m
d = 140m
0
100
1 000
lake, river
500
gravel
200
coarse
100
medium
50
fine
20
clay
q (A.ms)
Pulse moment
Pulse moment
Pulse moment
Pulse moment
Pulse moment
Pulse moment
free
water
1000
MRS is a direct method for groundwater detection, as it directly measures the response of the water itself (H protons).
The more traditional methods (DC, TDEM, ), are indirect ones, as they measure a physical parameter which is only
indirectly linked to the presence and to the quantity of water: the electrical resistivity of the layers is a function not only of
the porosity (volume of water) but also of the resistivity of the water ; besides, the formation resistivity is also influenced by
the conductivity of clay which makes the interpretation sometimes complex. However, the MRS cannot distinguish between
fresh and salt water, as it has no relation with dissolved salts, but only to the hydrogen content.
In terms of depth determination, MRS is influenced, as other geophysical methods, by equivalence rules, due to the fact
that it is an integrating method. However, for MRS , the eigen parameter is the product of the water content by the
thickness of the layer, which means that the total quantity of water is always fairly well determined.
A particularity of MRS is the non linear relationship between the measured signal and the energizing pulse
intensity. This means that doubling the pulse current does not mean doubling the signal: instead it increases the depth of
investigation. On the other hand, the MRS signal is linearly related to the water content of the layers, which makes the
interpretation quite quick.
bound
water
E0
Some specificities of the MRS method have to be pointed out: in case of magnetic rocks (volcanics,), the Earth
magnetic filed is non homogeneous, which makes the method difficult to apply. Besides, due to the very low amplitudes of
the signals to measure, the MRS method is sensitive to electromagnetic noise such as power lines, and to passive
conductors such as pipes, fences etc. This makes its application more difficult in industrialised areas.
time
Main unit
current
water
iris-instruments.com
MRS signal
depth
A. Legchenko
BRGM, France
legtchenko@exchange.brgm.fr
SUMMARY
The Magnetic Resonance Sounding method (MRS) has
been used in the past years with success in various
geological and geographical contexts for groundwater
surveys. This method has indeed the ability of directly
detecting the presence of water through the excitation of
the hydrogen protons of water molecules.
The frequency to which the H protons react depends on
the magnitude of the Earth magnetic field, while the
intensity of the excitation determines the depth of
investigation. The amplitude of the magnetic field
generated in return by the water of a layer is proportional
to the porosity of this layer, and the time constant of the
relaxation curve is linked to the mean pore size of the
material, that is to say tightly related to its permeability.
A loop laid on the surface of the ground is used for both
transmitting the excitation pulse and measuring the
response of the H protons. The linear relation between
the measured signal and the layer porosity permits to
interpret the 1D sounding as soon as the readings have
been collected in the field.
The main applications of this method concern the
determination of the water level and of the total quantity
of water available down to 100 to 150 m depths.
Magnetic Resonance Soundings can also help to select
the best place for drilling, to predict a yield using a
calibration, and to determine the geometry of an aquifer
layer for hydrogeological modelling
A set of field examples acquired in various countries
(Africa, Asia, Europe) points out both the advantages and
the limitations of this method and suggests the place it
should take among other geophysical methods in the
methodology of groundwater investigations.
INTRODUCTION
Groundwater exploration carried out with traditional methods
(TDEM, DC resistivity,) usually leads to good qualitative
success, even if an evaluation of the quantity of water present
into the ground is not possible due to the indirect relation
between the physical parameter measured and the water.
However, for the past five years, a new procedure based on
the Magnetic Resonance phenomenon has been introduced in
routine applications for directly detecting the presence of
Bearth
BEarth
Extended Abstracts
T 2*
I
t
area.
Figure 4. Theoretical Magnetic Resonance Sounding
curves for a 10m thick aquifer layer having 20% water, at
20, 50, 80 and 140m depths
Extended Abstracts
Interpretation:
Porosity vs depth
Field curve:
Signal vs pulse
noise
Field curve:
Signal vs pulse
Interpretation:
Porosity vs depth
FIELD EXAMPLES
Magnetic Resonance Soundings in Mauritania
In a sandy area where the water is very fresh (with a
conductivity of about 100S/cm), the DC electrical
soundings do not permit to distinguish wet sands from dry
sands inter-bedded between clayey formations. Figure 5
shows a neat response of the Magnetic Resonance Sounding
curve over a 40m depth, at least 50m thick aquifer layer
having of the order of 20% of free water. The complete
Extended Abstracts
CONCLUSIONS
Both theoretical and experimental developments show
interesting opportunities for the Magnetic Resonance
Sounding method in groundwater investigations. The ability
of directly detecting the presence of water gives the
possibility to quantify the resource both in terms of porosity
as well as permeability before deciding if a borehole has to
be drilled or not, for optimising the survey costs.
The integration of this technique with the conventional
indirect methods has to be examined on a case to case basis
according to the capabilities of each one of the methods.
More developments are necessary to adapt the methodology
of MRS in environments such as noisy areas or 2D aquifers.
REFERENCES
Figure 7. Correlation between the transmissivity given by
MRS and that given by pumping tests (BRGM & IRD)
CONDITIONS OF APPLICABILITY
The experiments and surveys carried out up to now in various
geological backgrounds permit to identify the conditions
where the Magnetic Resonance Sounding method has the
best chance of success:
- The water layer to investigate has to be located in the first
100 to 150m. This depth can be decreased if the ground is
conductive.
- There should be no magnetic material around or within the
aquifer layer, since in such a case the Earth magnetic field is
usually non homogeneous which prevents the hydrogen
protons to have the same excitation frequency
- The electromagnetic noise should be as low as possible: the
amplitudes of the MR signals are very low (tens to hundreds
of nV in ten to twenty thousand square meters loop) and
power lines, pumps, fences, pipes and magnetic storms
sometimes create difficult situations not allowing to get good
readings. Using figure-of-eight loops permit to improve the
signal to noise ratio but decreases the depth of investigation.
- The aquifer should be close to 1D conditions such as
alluviums or porous sediments or volumic hard rock
alteration. In case of fractures or faults, the difficulty may
come from the dilution of the signal within the field
generated by the large loop, also the localization of the fault
within the loop. In such case of 2D or 3D water bearing
Extended Abstracts
Loop wire
reel
Converter & Tuning
Batt
Main unit
PC
station C (50 m)
station B (30 m)
station A (10 m)
water level
sea level
CONCLUSION
The data measured and interpreted in this survey show that both the interpretation
software and the equipment (NUMIS Lite) permit to reach good quantitative results in case of
favourable conditions (reasonable EM noise, no magnetic material, presence of enough
water). The coherence of the interpretation of the various loop configurations must incite the
operator to use the one which best fits its local constraints.
The NUMIS Lite system, by its compactness and handiness will permit in the future to
test the MRS method in more geological, geographical and application contexts when the
required penetration does not exceed 50m.
10 m
station A
(10 m elevation)
station B
(30 m elevation)
25 m
station A (10 m)
station C
(50 m elevation)
station B (30 m)
60 m
40 m
60 m square
25m
60 m
25m
30 m
2 turns 30 m square
60 m square loop
Combination of electrical resistivity and magnetic resonance sounding data for mapping an
aquifer layer in Mauritania
Jean Bernard*, IRIS Instruments, Mohamed Lemine, Phy, Bassirou Diagana, Centre National des Ressources
en eau de Mauritanie, and Marc Ricolvi, Antea
Summary
Within the framework of a groundwater survey for
camel breeding purposes, electrical resistivity soundings
have been carried out for delineating the depth and the
lateral extension of an aquifer layer in the Dhar Nema
area located in the South-East part of Mauritania. The
geology basically consists in sand and sandstones, with
clay occurrences. In the middle part of the aquifer, the
DC resistivity soundings could point out the presence of
water. However, in the edge parts, they could not clearly
make the difference between dry rocks and fresh water
aquifer layers.
On the contrary, the magnetic resonance soundings
could identify the presence of water at depths of 60 to
more than 100m, and gave estimations of the values of
the porosity and of the permeability. The low EM noise
levels and the significant quantity of water existing into
the ground explain the good quality of the data. As a
whole, more than 60 MRS soundings have been carried
out in the area of approximately 150 x 100 km, and
permitted to delineate the limits of the aquifer layer.
The first drill-hole confirms the depths determined with
the magnetic resonance soundings. A campaign of new
drill-holes with pumping tests is planed to ascertain the
groundwater resources of this area.
Introduction
The Dhar Nema area is a 35 000 km desert plateau
located in the South East of Mauritania, slightly tilting
towards the East. Its West limit close to the city of Nma
consists in a 150m height cliff. The geology includes
four differentiated layers, from bottom to top:
West
DHAR
NEMA
plateau
East
dry area
aquifer area
sand dune
150m
NEMA city
dry
sandstones
aquifer
sandstones
clay
paleo-topography
with faults
shales
West
?
aquifer
sandstones
dry sandstones
clay
Figure 2: Interpreted resistivity section from Schlumberger soundings, in the Dhar Nema area
apparent
resistivity
sand
clay
sand
100 -
sandstone
water level
clay
40 to
100m
sandstone
with much water
MRS
signal
clay
AB / 2 line length
water resistivity
(wr) 100 ohm.m
sandstone porosity
(po)
20 %
wet sandstone ( 0.8x wr/ po) 2 000 ohm.m
clay resistivity
20 ohm.m
E0
T 2*
time
700 W
600
DRY
AREA
signal
pulse
DEEP
signal
pulse
TWO
signal
pulse
MEDIUM
signal
pulse
AQUIFER
AREA
1700 N
MALI
Nma
.
50 km
1600
MRS n N19
MRS nN31
(Leetig site)
THICK
signal
pulse
signal
pulse
MRS n19
0m
sand
35m
75m
MRS n31
95m
116m
water level
hard
sandstones
coarser
sandstones
finer
sandstones
version
loop size
depth
NUMIS Lite
standard
60 x 60m
50m
Fig 1: NUMIS Plus (left) and NUMIS Lite (right) MRS systems
NUMIS Plus (NUMIS Lite) system uses DC/DC converters to increase the voltage of the
batteries up to 400V (100V), and to generate pulses up to 4000V (1000V) at the Larmor
excitation frequency (1 to 3 kHz). Capacitor units are used to tune the inductance of the loop at
the Larmor frequency, which optimizes the intensity of the current generated. Both systems are
controlled by a PC computer from which the operator manages the acquisition process (Fig 2).
t
PC
Receiver
I
P
Battery
Converter
Main
unit
Transmitter
Switch
Tuning
Loop
Loop
t
I. t :
excitation pulse moment (A.ms)
Fig 2: Schematic diagram of NUMIS Plus system and parameters measured during a Magnetic Resonance Sounding
a configuration window where the operator introduces the parameters necessary to carry
out the sounding: type and dimension of the loop, value of the Larmor frequency (or that
of the Earth magnetic field), number of pulse moments to measure, stacking number,
recording time and number of pulses for the T1/T2* determination [3]. The set up of
tuning capacitors is also proposed in this window.
a system window for checking the shape of the current waveform during the injection,
also the values of the battery and converter voltages, output current, output voltage, gain
factor, signal phase.
a signal window where the operator can follow up the sounding during the acquisition
itself, with the display of the decaying stacked relaxation curves (amplitudes after first
and second pulse, noise), the estimation of the initial amplitude of the relaxation curve, its
time constant, its frequency. The sounding curve, the initial amplitude for the pulse
moment values already measured, is also displayed during the sounding. Finally, a
frequency spectrum is available for analyzing the frequency content of the relaxation
curves and of the EM noise.
the Earth magnetic field must not vary laterally more than +/- 20 nT on the surface of the
loop and in its vicinity. This corresponds to a variation of +/- 1 Hz on the Larmor
frequency.
the magnetic susceptibility of the rocks should be low enough not to perturb the relaxation
of the H protons: this means that the MRS soundings are presently quite difficult in
volcanic rocks for instance.
the MRS station must be far enough from power lines, pipes, fences, pumps, which
create EM noises sometimes too large compared to the small amplitude signals to
measure, of the order of a few tens nanoVolts in a 10 thousand square meters loop
surface.
STACKING PROCESS AND MRS LOOP SELECTION FOR REDUCING THE NOISE
To improve the signal-to-noise ratio, a stacking process is used for reducing the relative
influence of the random part of the noise. Two specific notch filtering (large, narrow) are
proposed in option, recommended when the local Larmor frequency is close to the harmonics of
the power line frequency (50 or 60 Hz).
When the quality of the signal is still poor after the stacking, a few techniques involving the
shape of the MRS loop can be used to improve the signal-to-noise ratio:
SIGNAL
SIGNAL
NOISE
N2 NOISE
1 NOISE
Tx, Rx
Tx, Rx
L = 100, 150 m
Max depth L
Signal = S
Noise = N
Signal / Noise = S/N
Tx, Rx
L/2 = 50, 75 m
Max depth L/2
Signal = s1 - (-s2) = 2s = S/2
Noise = n1 - n2 N/10*
Signal / Noise 5* S/N
Tx, Rx
L = 100, 150 m
Max depth L
Signal = S
Noise = N1 N2 N/10*
Signal / Noise 10* S/N
Rx only
Fig 3: Square loop (left), Eight-Shape loop (middle) and Compensation-Square loop (right) configurations
The eight shape loop (Fig 3, middle part) permits to improve the signal-to-noise ratio 2 to 10
times (depending on local conditions), compared to the standard square loop (left part), but limits
the depth of penetration to half of that of the same perimeter square loop enabling the same
current for a given output voltage.
The recently experimented square shape using a compensation receiving loop (right part) also
improves the signal-to-noise ratio but maintains the same maximum penetration depth as the
standard square shape: an electronic switch unit located in series with the main square loop
behaves as a short circuit during the transmission of the current, while it connects the opposite
voltage coming from the compensation loop during the measurement of the MRS response. In
such a way, the noise from the compensation loop cancels a large part of the noise from the main
loop, while the MRS signal from the compensation loop is proved to be low enough to be
negligible in front of that of the main loop.
Other loop configurations, such as multi-turn loops, can be used when the space available in
the field for setting up the loop is not large enough to use the standard square loop. The
comparison of the interpretation results of various soundings measured at a same location with
various loop geometries usually shows a good general agreement [4].
the stacked signal curve must be over the stacked noise curve.
the first part of the stacked signal curve must be greater than its second part, because of
the relaxation process which corresponds to a decaying trend.
The main frequency computed by the processing software within the received stacked
signal must not be farther from 1 Hz of the frequency of the transmitted excitation
current.
When after stacking the operator cannot detect any significant Magnetic Resonance signal, he
has to consider one of the following hypothesizes:
-
the excitation frequency is not close enough to the good one, and has to be modified
because of some Earth magnetic field gradient.
the EM noise is too high in comparison to the amplitude of the MRS signal, for the
number of stacks used.
the presence of magnetic rocks prevent to get a measurable Magnetic Resonance signal
The validation phase of a proper MRS signal observed at a given location is the most
important one which the operator has to ensure by himself: the remaining part of the acquisition
process can be automatically handled by the equipment under the control of the PC computer.
TRENDS IN THE EVOLUTION OF THE MRS INSTRUMENTATION
The improvement of the MRS instrumentation will be contemplated through the development
of multi-channel acquisition receivers, on the one hand for carrying out simultaneous readings in
2D imaging techniques, and on the other hand for eliminating the environmental noise more
efficiently by correlation processing techniques and decreasing the measuring time.
REFERENCES
[1] Vermeersch F., 2000, NUMIS Plus equipment operating manual, IRIS Instruments document.
[2] Bernard J., 2004, MRS: Step-by-step operation of NUMIS systems, IRIS Instruments document.
[3] Bernard J., Legchenko A., Groundwater, exploration with the Magnetic Resonance Sounding method,
Proc. of the 16th ASEG Geophysical Conference and Exhibition, Adelaide, February 2003, and Invited
Paper at the 65th EAGE Conference & Exhibition , Stavanger, June 2003.
[4] Vermeersch F., Texier B., Bernard J., Comparison of various geometries in Magnetic Resonance
soundings on the Pyla sand dune (France), Proc. of the 2nd International Workshop on the MRS Method,
Orleans, 19 21 November 2003, (118 - 120).
The Proton Nuclear Magnetic Resonance method, also called the Magnetic Resonance
Sounding method (MRS), after having been a research tool during a long maturation period, is
in the way of being more and more applied in the groundwater surveys for complementing the
traditional geophysical methods. Its capacity to give quantitative information for
characterising the water layers (depth and thickness, porosity, permeability after calibration)
give it a special place in the range of geophysical tool for hydrogeologists.
Due to the low levels of the signals which are measured in Magnetic Resonance field
surveys, to make the method efficient, one must take special care of the accuracy of the
Larmor frequency used in relation with the local Earth Magnetic field and of the filtering of
the natural and industrial electromagnetic noises. The shape of the wire loop used to energise
the ground and to receive the relaxation signals (square loop, eight-square loop, compensated
square loop, ) from which the initial amplitudes and the time constants are determined is
also a matter of importance as it shares the control of the depth of penetration together with
the transmitter power and it directly acts on the way the local noise is primarily filtered. This
study will report on the advantages and the limitations of these various shape loops with
respect to these two parameters, also the basic magnetic properties of the materials which
permit a measurable expression of the relaxation effects.
Besides, to optimise the use of a new geophysical method in groundwater surveys, it is
important to have a clear view of its real output within the various backgrounds which can be
met in such surveys. Field results coming from different types of aquifer, depths, geological
conditions, and countries in the Western World, Asia and Africa show the data quality and the
field efficiency which can be expected with the present state of the art of the technology used.
Also, the correlation with hydrogeological borehole parameters, such as the yield or the
transmissivity, when available, opens the way for characterizing the geological materials from
surface geophysical readings.
info@iris-instruments.com
info@iris-instruments.com
info@iris-instruments.com
info@iris-instruments.com
info@iris-instruments.com