GROUNDWATER AND HUMAN DEVELOPMENT
Bocanegra, E - Martnez, D - Massone, H (Eds.) 2002 - ISBN 987-544-063-9
1484
AN INEXPENSIVE FIELD FLUOROMETER FOR HYDROGEOLOGICAL TRACER TESTS
WITH THREE TRACERS AND TURBIDITY MEASUREMENT
Pierre-Andr Schnegg, University of Neuchtel, Switzerland
Abstract. The Geomagnetism Group of the University of Neuchtel has recently designed a flow-through
field fluorometer with added spectral capabilities for hydrological tracer tests. This instrument is equipped
with four optical axes allowing water sample illumination with four independent light sources at different
wavelengths covering the full spectrum from UV to red. As many as three conveniently selected (dye) tracers
can be simultaneously measured and separated from a cocktail. Careful turbidity measurement allows for
correction of the fluorescence signals.
Keywords: Fluorometer, tracer, hydrogeology, uranine, rhodamine, Tinopal
INTRODUCTION
Flow-through fluorometers can advantageously
replace mechanical samplers in tracer tests that use
dyes. If sample archiving for future detailed
laboratory analyses is not required, field
fluorometers are the solution of choice, since they
can work unattended for days or even weeks.
Because the tracer concentration is directly
measured in a flow stream, there is no
contamination or sample ageing. Moreover unlike
mechanical samplers, flow-through cells are
insensitive to frost. Furthermore, tracing data are
immediately available during a test as well as at the
end, thereby eliminating the cost of laboratory
analyses.
The Geomagnetism Group of the University of
Neuchtel (GGUN) has been actively involved in
the development of robust and inexpensive
fluorometers suited for field work. It extended the
optical capabilities of its recently developed
instrument from 2 to 4 light sources.
Simultaneously, it has improved the measurement of
water turbidity and applied a corrective term to the
signals from the dye fluorescence prior to
mathematical separation of the concentrations.
Earlier designs can be found in Schnegg and
Doerfliger (1997), Schnegg and Kennedy (1998).
Field work in porous media is reported by Kennedy
et al. (2001a,b). Experiments using a downhole
version of the fluorometer are described in Schnegg
and Bossy (2001).
THE FLOW-THROUGH FIELD
FLUOROMETER
An optical cell made of a simple glass tube
placed along the geometrical axis of a metal
cylindrical waterproof casing measures the tracer
concentration in the water flowing through the flow
cell. The optical components used for the
measurement of dye concentration are installed
along the orthogonal axes of two square crosses in
two separate planes (Figure 1). The measurement
system consists of
an excitation section, comprising a quasi-
monochromatic light source, a filter and a
condenser lens, and
a detection section, orientated 90 to the
excitation beam, with a lens, a filter and a
photo-detector.
The light sources and the filters are selected
according to the absorption-emission spectra of the
dyes. Such a geometry allows for installation of up
to four measuring systems. One of the sets is
dedicated to the measurement of the water turbidity
while the three others are used to measure the dye
concentrations. Light sources with spectral maxima
at 370, 470 and 525 nm are ideally suited for
excitation of dyes such as Tinopal CBS-X, uranine
(Na-fluorescein) and any molecule in the rhodamine
family (amidorhodamine G, sulforhodamine B,
rhodamine WT).
Figure 1: Optical cell (glass tube) and
four sets of light sources and photo-detectors
AN INEXPENSIVE FIELD FLUOROMETER FOR HYDROGEOLOGICAL TRACER TESTS
WITH THREE TRACERS AND TURBIDITY MEASUREMENT
1485
Instrument sensitivity and linearity
The relative sensitivity to different dyes
(smallest detectable concentration) is similar to that
of laboratory spectrophotometers. Uranine is by far
the most sensitive molecule (detected at
concentrations approximately 8 times less than other
dyes). The smallest detectable concentration
variation (in clear water) is 0.02 ppb for uranine and
0.14 to 0.2 ppb for the other tracers.
Perfect linearity of the relationship between
the concentration and the fluorescence signal cannot
be achieved due to geometrical effects on the
excitation light within the optical cell. This effect is
wavelength-dependent. For the three available
sources at 370, 470 and 525 nm we observed non-
linearities of 8%, 3% and 11%. This is of concern
when dye separation is performed, since the
calculation assumes a set of three linear equations of
the concentration. Small absolute discrepancies
appear at the low-concentration end.
The measurement of single-tracer solutions
can be carried out very accurately in a range of
concentrations extending over 5 decades, from the
detection limit to 1000 ppb (10
-6
g/ml). Over this
range, the calibration curve can be fitted with a 1
st
-
degree polynomial in log-log space (Figure 2).
Therefore, the fluorometer calibration requires only
two solutions (usually 10 and 100 ppb). If higher
concentrations are of interest, additional
concentrations are used to fit higher-order
polynomials.
-10
-9
-8
-7
-6
-5
-2 -1 0 1 2 3 4
Log Signal (mV)
L
o
g
C
o
n
c
e
n
t
r
a
t
i
o
n
(
g
/
m
l
)
Figure 2: Calibration curves for uranine (),
sulforhodamine B (*) and Tinopal CBS-X ()
Turbidity measurement and correction
Turbidity effects often occur in tracer tests.
Suspended particles entering into the optical cell
scatter the excitation light and produce a stray
signal, simulating the presence of the tracer, since
excitation/detection filters partly overlap (a few
0
/
00
of transmitted light). A dedicated optical system
measures the amount of light scattered at 90 from
the excitation beam. With clean water, this signal is
close to zero (only Raman scattering), but increases
with the number of suspended particles. The
wavelength involved in this measurement must be
selected in the red part of the spectrum, so that the
light cannot generate fluorescence if a tracer is
present in the water. To remove turbidity effects, the
fluorometer must be calibrated with different turbid
suspensions (we use typically 1, 10 and 100 NTU
(nephelometric units, formazine standards). The
measuring set (equipped with a red 660 nm light
source) is calibrated (Figure 3) and the polynomial
coefficients of the calibration curve obtained. Note
that the relationship between the optical signal and
the turbidity is linear in log-log space.
-2
-1
0
1
2
3
-1 0 1 2 3 4
Log Signal (mV)
L
o
g
C
o
n
c
e
n
t
r
a
t
i
o
n
(
N
T
U
)
Figure 3: Calibration curves for turbidity
A second calibration is required to determine
how much stray signal is produced by turbidity on
each measuring set. Thus, a full measurement is
done by collecting two signals, the tracer signal and
the turbidity, measured both in mV. Inverting the
latter yields the amount of stray signal that can then
be removed from the tracer signal. Figure 4 shows
the effect of a 40 NTU turbidity peak on a cocktail
with constant (~ 8 ppb) concentration of uranine and
sulforhodamine B.
Pierre-Andr Schnegg
1486
0
10
20
30
40
50
60
Time (minutes)
C
o
n
c
e
n
t
r
a
t
i
o
n
(
p
p
b
)
Uranine
Sulforhodamine B
Turbidity
a b
Figure 4: Influence of turbidity on tracer
concentration measurement, without (a) and with (b)
correction.
The same data set is shown on Figures 4a and
4b, but the turbidity correction has only been
applied to Figure 4b. Obviously, the sulforhodamine
measurement would be distorted at turbidity levels
in excess of 20 NTU. The applied correction
completely removes this distortion.
Separation of three tracers
There is a considerable interest in measuring
tracer concentrations that were injected at different
locations during a tracer test. The resulting tracer
cocktail measurements can be separated to obtain
the time curve for each tracer. This can be done with
the GGUN fluorometer, provided the various tracers
are properly selected, i.e. their excitation/emission
spectra do not overlap excessively.
The separation of the three tracers is achieved
by solving a set of 3 linear equations. Each equation
gives the amount of fluorescence signal V
i
on
photodetector P
i
produced by each tracer under
excitation by lamp L
i,
with i=1,2,3. For small tracer
concentrations such as found in many
hydrogeological tests (< 1 ppm), tracer signals are
additive. As an example we take the hypothetical
case of a water containing three different tracers
with concentrations , and . Previous calibration
of the fluorometer yields the fixed coefficients C
j
i
of
three different sets i of lamps, filters and
photodetectors for a fixed concentration (100 ppb)
of each tracer j. The set of equations
i
i i i
V C C C = + +
3 2 1
, i=1,2,3 (1)
yields following solution:
3
3
3
2
3
1
2
3
2
2
2
1
1
3
1
2
1
1
3
3
3
2 3
2
3
2
2 2
1
3
1
2 1
C C C
C C C
C C C
C C V
C C V
C C V
=
3
3
3
2
3
1
2
3
2
2
2
1
1
3
1
2
1
1
3
3 3
3
1
2
3 2
2
1
1
3 1
1
1
C C C
C C C
C C C
C V C
C V C
C V C
=
3
3
3
2
3
1
2
3
2
2
2
1
1
3
1
2
1
1
3
3
2
3
1
2
2
2
2
1
1
1
2
1
1
C C C
C C C
C C C
V C C
V C C
V C C
=
Stability of this solution depends on the choice
of cut-off wavelengths for the various filters, on the
central wavelength of the light sources and also, on
the choice of tracers in the cocktail. Good tracer
compatibility is achieved with dyes such as Tinopal,
uranine and any type of rhodamine. Cocktails of
uranine and eosine (or pyranine) do not fulfil the
compatibility conditions, because optical
characteristics of these tracers are too similar to
each other in terms of wavelengths. The same
remark holds for cocktails with different types of
rhodamine (amidorhodamine G, sulforhodamine B,
rhodamine WT).
To test the validity of the separation method,
increasing quantities of tracer #1 solution (up to
107.5 ml @ 992 ppb) were added to a fixed volume
(1 litre) of a solution containing tracer #2 and tracer
#3 (100 ppb each), in which the sonde was
immersed (Figure 5 a,b,c).
100 ppb uranine - 100 ppb sulforhodamine cocktail
1
10
100
1000
1 10 100
Added Tinopal solution (ml)
C
o
n
c
e
n
t
r
a
t
i
o
n
(
p
p
b
)
Figure 5a: Tinopal CBS-X added to
sulforhodamine B and uranine cocktail
AN INEXPENSIVE FIELD FLUOROMETER FOR HYDROGEOLOGICAL TRACER TESTS
WITH THREE TRACERS AND TURBIDITY MEASUREMENT
1487
100 ppb Tinopal + 100 ppb uranine cocktail
1
10
100
1000
1 10 100
Added sulforhodamine solution (ml)
C
o
n
c
e
n
t
r
a
t
i
o
n
(
p
p
b
)
Figure 5b: Sulforhodamine B added to uranine and
Tinopal CBS-X cocktail
100 ppb sulforhodamine B - 100 ppb Tinopal CBS-X cocktail
1
10
100
1000
1 10 100
Added uranine solution (ml)
C
o
n
c
e
n
t
r
a
t
i
o
n
(
p
p
b
)
Figure 5c: Uranine added to sulforhodamine B and
Tinopal CBS-X cocktail
Indices 1, 2 and 3 refer to cyclical
permutations of uranine, sulforhodamine B and
Tinopal CBS-X. Generally, good tracer separation is
achieved within 1% accuracy. Tinopal has the
smallest sensitivity and shows slight departure from
the theoretical curves (full lines) at lowest values.
Note that the bending of tracer #2 and #3 curves
results from dilution with tracer #1 solution (1 to 10
volume ratio).
A second experiment tested the separation
power for sudden concentration variations of tracer
#1 in a cocktail of tracers #2 and #3 (Figure 6). No
significant concentration variations are observed in
the cocktail. Similarly, the turbidity measurement is
not influenced by the steep variation of tracer #1.
When tracer #1 level is low and close to the
instrumental noise, accurate separation cannot be
obtained if concentrations of tracers #2 and #3 are
more that 10 times larger.
0.01
0.1
1
10
100
1000
Time (minutes)
C
o
n
c
e
n
t
r
a
t
i
o
n
(
p
p
b
)
Uranine
Sulforhodamine B
Tinopal
Turbidity
a b
c
Figure 6: a,b: Dynamic behaviour of the separation
scheme. c: Erroneous signal produced by
uncalibrated tracer (eosine).
Careful calibration is mandatory for obtaining
reliable tracer separation. Figure 6c shows what
happens if an unexpected, uncalibrated molecule is
detected (in this case, eosine). The equation set
leads to a solution in terms of the tracers for which
the instrument had been calibrated.
CONCLUSION
The GGUN field fluorometer equipped with
three optical sets and an additional channel for
turbidity can be successfully used for real-time
separation of a cocktail of up to three tracers. After
reduction of turbidity effects on each signal, the
resolution of a set of linear equations leads to
individual tracer concentrations. The success of the
operation relies on accurate calibration of the
instrument with the same chemicals as used during
the tracer test. Finally, it is important to note that the
concentration ratio should not exceed a factor of 10
for concentrations close to the detection limit.
References
Kennedy K, Mller I, Schnegg PA., Rossi P, Kozel
R. 2001. Characterization of the Kappelen
Groundwater Research Site (BE), Switzerland, and
Preliminary Bacteriophage and Solute Tracer
Component Responses. Beitrge zur Hydrogeologie,
52, 158-180.
Kennedy K, Niehren S, Rossi P, Schnegg PA,
Mller I, Kinzelbach W. 2001. Results of
Bacteriophage, Microsphere and Solute Tracer
Migration. Comparison at Wilerwald Test Field,
Switzerland. Beitrge zur Hydrogeologie, 52, 180-
210.
GROUNDWATER AND HUMAN DEVELOPMENT
Bocanegra, E - Martnez, D - Massone, H (Eds.) 2002 - ISBN 987-544-063-9
1488
Schnegg PA., and Doerfliger N. 1997. An inexpensive flow-through field fluorometer. In: Proc. 6th
Conference on Limestone Hydrology and Fissured Media, la Chaux-de-Fonds, 10-17 August 1997, Vol. 2,
47-50.
Schnegg PA., and Kennedy K. 1998. A new borehole fluorometer for double tracer tests. In: Mass
Transports in Fractured Aquifers and Aquitards, Geoscience Center, Copenhagen, 14-16 May 1998, 60-63.
Schnegg PA., and Bossy F. 2001. Sonde for downhole measurement of water turbidity and dye tracer
concentration. In: New Approaches Characterizing Groundwater Flow, Munich 10-14 September 2001, 795-
799.
Authors address:
Groupe de Gomagntisme
Universit de Neuchtel
CH-2007 Neuchtel
Switzerland
[Link]@[Link]
[Link]
Test in a karst spring at St Blaise (NE) 5.04.2002
Demonstration of three dye tracer separation. Three cocktail solutions of 0.1 litre, then 1 litre @ 10,000 ppb
have been injected 300 m upstream every two minutes. Stream flowrate: ~100 litre s
-1
0.1
1
10
100
0 5 10 15 20 25 30 35 40 45
Minutes
C
o
n
c
e
n
t
r
a
t
i
o
n
(
g
/
l
)
Uranine
Rhodamine
Tinopal
