Implementation of Tracing

SOURCE: CETRAHE, BRGM, 2019.

Artificial Tracing in Hydrogeology: Best Practices. Information System for the Management of Groundwater in Centre-Val de Loire, published online in January 2019
https://sigescen.brgm.fr/Tracages-artificiels-en-hydrogeologie-les-bonnes-pratiques.html

Before conducting a tracing operation, several preliminary steps should be considered :

The first step is to clearly define the objectives of the tracing: reconnaissance of underground flows, simulation of pollution transfer, characterization test of the aquifer with the determination of hydrodispersive parameters (flow velocity, kinematic porosity, dispersivity), etc. This step is crucial as the strategic choices made subsequently will be a compromise between objectives and cost.

The second step involves gathering as much existing information as possible, as well as documentation on previous tracings (see the article dedicated to the regional inventory). The collected information should include all geographical, topographical, geological, hydrogeological, and anthropogenic data (water uses, intakes, etc.). As for previous tracings, even if they do not meet today's evaluation criteria, they will be rich in information and very useful for avoiding certain pitfalls.

The third step is the reconnaissance of the site where the tracing will be conducted. This involves identifying potential injection points (direct access or via an unsaturated zone, absorption capacity, possibilities of loading and overflowing, need for flushing, accessibility especially for vehicles transporting water for flushing, etc.) and potential restitution points (intakes, uncaptured springs, surface water outlets, operation, accessibility, possible flow measurement, etc.). At the end of this visit, it is important to assess the feasibility of implementing various monitoring devices (manual sampling, installation of automatic samplers, installation of fluorimeters, attachment of activated carbon detectors, influence of pumping regimes, influence of chlorination, etc.) and to anticipate the hydrological conditions that may differ (and vary) at the time of the test.

After addressing these steps, one can proceed to the sizing of the tracing.

Single or Multi-Tracing? Multi-tracing involves simultaneously injecting different tracers at multiple injection points. It allows answering several questions at once, reducing costs, and saving considerable time. However, it requires a judicious choice of tracers used, sufficiently conservative in the context, and without presenting analytical interferences between them.

In practice, the quantity is estimated by experts, considering the hydrogeological context. Between empiricism, intuition, and experience, to decide, two determining elements must be considered: the dilution that the tracer should undergo, often approximated by distance, and the analytical performance of the tracer and monitoring methods.

NOTE

Tracing cannot provide information on the entire hydrological or hydrogeological system. The results relate only to the tested part. To extrapolate to another part of the aquifer, one must be certain of the homogeneity of the environment.

Best practices involve providing prior information about the tracing operation to authorities (DDT, gendarmerie, etc.) and local residents (town hall). This helps avoid concerns and alerts related to water coloration in the case of fluorescent or coloring tracers. Before any injection, it is necessary to take water samples as control samples, and if the protocol includes the use of activated carbon detectors, it is also necessary to plan the immersion of "control" fluocaptors at an appropriate frequency. For reconnaissance tracings, conducting them during high water periods generally provides more favorable conditions due to faster flows, preferably targeting a recession period. It is recommended to perform simulation tracings under contrasting hydrological conditions (low and high water), as the results obtained can fluctuate widely.

The monitoring and analysis method depends on several factors :

• Type of water point(s) monitored: spring, intake, borehole, river, etc.;
• Possibilities for equipment installation: available space, safety, electrical supply, access, etc.;
• Available budget.

The most reliable monitoring and analysis method is water sampling with laboratory analysis. Laboratory equipment today allows the detection of substances in very low concentrations. For fluorescent tracers, laboratory spectrofluorimeters (direct fluorescence measurement) can achieve very low detection limits, on the order of 0.001 µg/L for uranine. The spectral analysis performed by a spectrofluorimeter is an essential diagnostic tool for the detection and interpretation of tracing, especially as injection quantities are increasingly reduced to remain below the visibility threshold at restitution points.

Spectrofluorimeter (source: CETRAHE)

Field instruments allowing in situ measurements also contribute to improving tracer monitoring. Increasingly high-performance instruments are available: field fluorimeters, specific electrodes, sensitive conductimeters, etc. For fluorescent tracers, the use of field fluorimeters can be very useful. Easy to use, these devices provide results in quasi-real-time, even in the case of multi-tracings. However, it is advisable not to use them as the sole monitoring device, especially for multi-tracings. Indeed, variations in the natural fluorescence of waters recorded, as well as interferences between tracers, can be mistakenly interpreted as restitutions. It is therefore advisable to couple this monitoring with automatic or manual sampling to verify, through spectral analysis in the laboratory, the presence or absence of the tracer.

Regarding activated carbon detectors (fluocaptors) sometimes used for fluorescent tracers, it is advisable to use them as a last resort when field conditions do not allow other detection methods. They can also be used as a secondary detection method to spatially expand monitoring in the context of reconnaissance tracings, monitoring "secondary" points. However, caution should be exercised in interpreting the results obtained. Among the common tracers, monitoring by fluocaptor can only be considered for tracers such as uranine or eosin, with a number of precautions (see technical note No. 1 from CETRAHE). Red tracers (such as Rhodamines) cannot be monitored by this method, as activated carbon has shown an inability to fix them in laboratory conditions at water concentrations below 30 µg/L.

The fluocaptor method is also unsuitable for fluorescent tracers that emit in the blue (sodium naphthionate, amino G acid, Tinopal). Finally, ionic tracers (salts) can be dosed with high analytical precision by various devices (ion chromatography, spectrophotometry, atomic absorption spectroscopy, etc.). However, the natural presence of these ions in waters interferes with their detection at low concentrations, despite the performance of the device used. The dosage of the injected quantity must therefore be particularly studied so that it is sufficiently high to be detected at monitoring points and sufficiently moderate not to disrupt water uses (water intakes, natural environments).

Data Exploitation and Interpretation

The results of a tracing are illustrated by the tracer restitution curve, thus showing the evolution of concentrations over time at the restitution point. Mastery of flows at the restitution point allows calculating a restitution balance (restituted mass and restitution percentage) and the Residence Time Distribution (RTD), which describes the transit of the tracer in the tracing system. The RTD corresponds to the probability density function that gives the probability that a tracer molecule will reside in the system. It is indeed the distribution curve of the tracer cloud. When the injection can be assimilated to a "Dirac" impulse (i.e., a brief injection), the RTD gives the impulse response of the tracing system for the hydrological conditions in which it finds itself at the time of tracing (Lepiller M. & Mondain P-H, 1986). From the RTD, several parameters describing the transit of the tracer can be calculated, such as the mean residence time and the apparent velocity. The interpretation of the results differs depending on the objective. For reconnaissance tracings, the main objective is to accurately confirm the belonging of an injection point to the impluvium of the karst system. For quantitative tracings (simulation), it is important to precisely describe the transit modalities of the tracer, as well as the hydrodispersive parameters for tracings in porous media. Analytical tools to assist in parameter estimation exist. The TRAC software, in "Interpretation" mode of tracings, allows interpreting a tracing using different analytical solutions by adjusting the parameters of the solution and comparing them with observation data.

NOTE

Instrument detection thresholds should not be confused with real detection limits, which are highly dependent on the background noise level in natural waters and vary depending on the tracer.

Finally, at the end of the tracing operation and the interpretation of the results, the operator is invited to enter the information into the data entry application of the Tracing Database. This is the national database dedicated to the banking of data. (see the article dedicated to declaration and banking)