Faculté des sciences

Development of a system for "in situ" determination of chlorinated hydrocarbons in groundwater

Boutsiadou, Xanthippe ; Hunkeler, Daniel (Dir.)

Thèse de doctorat : Université de Neuchâtel, 2012.

Volatile organic compounds (VOCs), and especially chlorinated hydrocarbons, are common groundwater contaminants. Efficient monitoring that can be conducted directly in the field is needed to detect a possible pollution by organic contaminants such as chlorinated hydrocarbons. The general aim of this project is to develop a portable instrument for the in situ measurement of chlorinated... More

Add to personal list
    Summary
    Volatile organic compounds (VOCs), and especially chlorinated hydrocarbons, are common groundwater contaminants. Efficient monitoring that can be conducted directly in the field is needed to detect a possible pollution by organic contaminants such as chlorinated hydrocarbons. The general aim of this project is to develop a portable instrument for the in situ measurement of chlorinated hydrocarbons in groundwater. The instrument relies on the transfer of volatile organic compounds to the gas phase followed by gas phase measurement. This research is based on three specific objectives: (a) testing of a module for extracting the contaminant from the water phase; (b) testing of a photoionization detector for gas phase analysis; and (c) calibration and evaluation of the complete instrument, including a field evaluation.
    The first objective was to extract the organic contaminants from the water. For this purpose, hollow fiber membrane modules in gas stripping mode were tested. Two different membrane materials, polypropylene (PP) and poly(dimethylsiloxane) (PDMS) were tested under low gas sweep flow rates, in order to reach maximal sweep gas concentrations. Sorption to membranes and mass transfer was investigated in detail for selected chlorinated hydrocarbons such as tetrachloroethene (PCE) and trichloroethene (TCE). PCE had a greater affinity for both membranes than TCE. Mathematical formulations were developed to calculate the experimental overall mass transfer coefficients for both membranes. Using the resistance in series model, the overall mass transfer coefficient was compared to the mass transfer coefficients estimated for the different layers (water boundary layer, membrane, gaseous boundary layer). For the PDMS membrane, the limiting factor was found to be the water boundary layer. For the PP module, the mass transfer was shown to be independent of the gas flow rate. The PDMS hollow fiber module mass stripping method appeared to provide a promising way for on-line analysis for the investigated chlorinated solvents. At the selected low gas flow rates, the PDMS hollow fiber modules yielded more stable gas phase concentrations than polypropylene modules and hence PDMS modules were used in the developed instrument.
    The second objective was to evaluate a miniaturized photoionization detector (PID) for the determination of chlorinated and aromatic hydrocarbon concentrations in the gas phase. Studied compounds included the aromatic hydrocarbons: benzene, toluene, and ethylbenzene and the chlorinated hydrocarbons TCE and PCE. To test the detector performance, a series of standards with known concentrations of organic compounds were measured. The study investigated response curves, response time, linearity between response and concentration as well as the influence of gas flow rate and humidity. The laboratory tests showed that the response of the PID was linear over a concentration range of 10 to 500μg l-1. The correlation coefficients R2 varied from 0.94 to 0.99 for most of the compounds. The PID’s signal was stable over time. However, the PID is showed some sensitivity to humidity, as the signal in a humidified environment was 20% lower than in a dry environment in a 50min period. However, this level of variability is acceptable for an instrument mainly developed for screening purposes.
    The third objective was the evaluation of the complete system to measure chlorinated hydrocarbons in groundwater. The system involved pumping water through a PDMS hollow fiber membrane module with a surface area of 0.001m2 at one side and passing air at the other side and monitoring the effluent air with the PID. The water and gas flow rates were 50ml min-1 and 12 ml min-1, respectively. The extraction and detection of TCE, PCE and cis-dichloroethylene were (cis-DCE) studied experimentally for various concentrations. Field measurements were compared to conventional laboratory data obtained by gas chromatography. Only the total chlorinated hydrocarbons concentration can be determined using the miniaturized PID. This device was capable of detecting minimum total aqueous concentration of 20μg l1-. The Swiss legal limits according to OSites that apply downgradient of contaminated sites are 70μg l-1 for TCE and 40μg l-1 for PCE. These features demonstrate that the combination of the PDMS hollow fiber membrane module with the PID is feasible for semi-continuous monitoring of chlorinated solvents in groundwater.
    To further test the applicability of the instrument under field conditions, a field study was carried out over a 5-month period in the contaminated military area of Lyss in Bern, Switzerland. Results showed a correlation between two methods (portable instrument and laboratory measurement through gas chromatography) with correlation coefficients (R2) of 0.62-0.75. In general, the portable detector’s concentrations are lower than the laboratory concentrations even if the regular calibration of the portable device. This is mainly because of the lower temperature of the field measurements where the Henry’s coefficient is lower and consequently a lower concentration is measured in the field. Correlation coefficients are improved (R2 equal to 0.85-0.99) when the effect of the temperature dependence of the Henry’s coefficient is taken into account for points where the VC concentration is not important. Field and laboratory concentrations show a better agreement but still there are variations between the two measurements. In its current form, the prototype enables the distinction between high-, medium- and low- or non-contaminated points. It is a simple screening method for site characterization and risk assessment. The total cost of the instrument is estimated at 10.000 Swiss francs.
    In conclusion, this study describes the development and demonstrates the applicability of a portable dissolved VOC detector as a screening tool. The instrument is flexible and a relatively large number of points can be measured in a relatively short period of time, while the data are available directly in the field. Limitations of instrument are the analysis of the organic compounds as a composite index, the long stabilization time and the memory effects of the hollow fiber. So, further experiments should be considered to overcome these limitations. The use of another detector could be a possibility to detect the organic compounds individually and not as a composite index. Moreover, a heating chamber for the hollow fiber can be used with a parallel air circulation in order to accelerate the desorption of the molecules from the membrane. Finally, a possible commercial production of this device should be considered.