EXPLORING THE ENVIRONMENTAL ISSUES OF MOBILE, RECALCITRANT COMPOUNDS IN GASOLINE. Organized by. D.L. Drogos and A.F. Diaz

Transcription

1 EXPLORING THE ENVIRONMENTL ISSUES OF MOILE, RECLCITRNT COMPOUNDS IN GSOLINE Organized by D.L. Drogos and.f. Diaz (Pages in Preprints of Extended bstracts, Vol. No. 1) Symposia Papers Presented efore the Division of Environmental Chemistry merican Chemical Society San Francisco, C March 26-, 00 EFFECTS OF ENVIRONMENTL CONDITIONS ON DEGRDTION IN MODEL COLUMN QUIFERS: II. KINETICS Clinton D. Church, James F. Pankow and Paul G. Tratnyek Environmental Science and Engineering, Oregon Graduate Institute, eaverton, OR Introduction Convincing evidence for in situ biodegradation of methy tert-butyl ether () is difficult to obtain from field data alone because of several distinctive characteristics of. For example, the interpretation of relative plume lengths for versus other gasoline constituents is difficult due to variations in the formulation and distribution of oxygenated gasoline. Decreases in concentrations over time or space (i.e., along a flow path) may be due to degradation, but they also result from dilution and dispersion 1-3. Even the discovery of likely degradation products (i.e., tert-butyl alcohol, T) is equivocal evidence for biodegradation of because T has been used as an oxygenate in some locals, and T is present as an impurity in much of the used to oxygenate gasoline 1. These difficulties with proving in situ degradation of from field data alone are exacerbated by the fact that biodegradation of is slow under almost all circumstances. n alternative approach to investigate degradation of is to use controlled model systems that are designed to simulate typical field conditions. We have prepared six model column aquifers to investigate the pathways and kinetics of biodegradation under controlled conditions simulating aquifer conditions. In Part 1 of our work with these columns 3, we found that degraded to T after a lag period of 35 days, but only under aerobic conditions and in the absence of TEX (the more recalcitrant of gasoline constituents: benzene, toluene, ethyl benzene, and o-, m-, and p-xylene). This conversion of to T in the aerobic columns eventually declined (apparently due to oxygen use by microorganisms) along with a decrease in dissolved oxygen in the column effluent. In similar columns with anaerobic or TEXcontaining influents, there was no degradation up to 1 days. For the work reported here (Part II), we designed two sets of experiments to investigate the nature of the aerobic degradation observed in the Part I studies. The first set of

2 experiments was designed to investigate whether the rate of degradation might be enhanced by the addition of a cometabolic substrate. The second set of experiments investigated the kinetics of the degradation. Materials and Methods The columns were constructed from (three or six) 2.54-cm I.D. segments of stainless steel (0-3 cm total length) connected with 2.54-mm I.D. stainless steel tubing, which allowed samples to be taken at evenly spaced points along the column. In order to compare sites that were geographically and geologically distinct, columns were prepared with sediments from four sites: two USGS (US Geological Survey) urban NWQ (National Water-Quality ssessment) field sites (Lake Erie and New Jersey); a USGS leaking underground storage tank study site (Laurel ay); and a controlled release site at ase orden, Ontario, Canada (orden). septic techniques (alcohol flame sterilized equipment, etc.) were used to collect the sediments used and to pack the columns. utoclaved site ground water in equilibrium with atmospheric O 2 was amended with (and either isopropanol, hexane, isopentane, isopentanol, malate, or ethanol for the cometabolism experiments) and used as the column influent. The column effluent was analyzed for and its potential degradation products by direct aqueous injection with gas chromatography and detection by mass spectrometry (DI-GC/MS) according to the method described previously 4. This analytical technique allowed us to quantify and T simultaneously over a wide range of concentrations and to detect and identify other substances that might be intermediates or degradation products (e.g., tert-butyl formate, isopropanol, and acetone). Results and Discussion In the experiments designed to investigate whether the rate of degradation might be enhanced by the addition of a cometabolic substrate, the conversion of to T stopped in the presence of added isopropanol, hexane, isopentane, isopentanol, malate, and ethanol (data not shown). This finding further confirmed that conversion to T occurs only in the absence of more favorable substrates. To avoid the onset of anaerobic conditions, as was observed in previous experiments performed at flow rates of 1.5 ml/d 3, the column microcosms were conditioned at ml/d until steady-state concentrations of and T were observed in the effluent (-60 days). Some data was collected at ml/d, but most kinetic data was collected after the flow rate had been lowered to. The concentrations of and T measured by DI-GC/MS on the column effluents are shown in Figure 1 (, C, and E). These plots show steady-state concentrations of for 5-7 days after breakthrough, and a steady partial conversion of to T after days, with essentially complete mass balance. The kinetics of degradation in these columns cannot be quantified reliably from the decrease in concentration because the proportion of degraded is less than 1 percent. The appearance of T is a much more sensitive indicator of degradation, however, so we have used T concentration data for our kinetic analyses. Note that mass balance calculations suggest loss of is accompanied by stoichiometric conversion to T (Figure 1, C, and E). Figure 1, D, and F show T concentrations at each port (under steady-state conditions, confirmed by agreement of kinetic parameters for ml/d and pumping rates) plotted against the elapsed contact time. The linearity of these plots suggests that the kinetics of T appearance might be zero-order.

3 T 2 1 Concentration (µm) C 7 E T T T Concentration (µm) D F Time (Days) Contact Time (Days) Figure 1., C, and E show and T concentrations in the effluents of three columns containing previously uncontaminated sediments. () ase orden, (C) Lake Erie, (E) New Jersey. The columns were challenged with aerobic influent containing 1.13 µm., D, and F show T concentrations derived from sampling ports along the length of columns, C, and E, respectively, while columns were at steady-state. However, further analysis (Figure 2) shows that the extent of reaction in our column was not great enough to distinguish between zero-order and first-order conversion of to T. To distinguish between these possibilities, one column (the one containing orden sediment) was exposed to three different initial concentrations of (ranging over two orders of magnitude). The results (Figure 2, -C), give nearly identical rate constants when fit to a first-order model for appearance of T. Similar results were found for the other two columns (data not shown). The average first-order rate constant for the experiments were (7.1 ± 0.1) x -4 d -1, (1.1 ± 0.1) x -3 d -1, and (9.9 ± 0.1) x -4 d -1 for the ase orden, Lake Erie and New Jersey sediments, respectively. From the combined results of this study, the average half-lives for conversion to T are 2.7 years, 1.7 years, and 1.9 years for ase orden, Lake Erie and New Jersey sediments, respectively. The similarity among these values is remarkable, and suggests that slow biodegradation of to T (with half-lives ca. 2 years) may

4 occur in a wide variety of aerobic aquifers. In fact, this estimate is consistent with available estimates of biodegradation rates in the field, which include half-lives of and 2.8 years 6. Such slow degradation rates are very difficult to characterize from field data, however, so the comparison with our column data should not be interpreted as more than qualitative agreement. In addition, it is important to recall that we only observed degradation when our columns were operated under aerobic conditions and in the absence of more favorable substrates (like the TEX compounds). Under typical groundwater conditions in and around plumes from many gasoline spill sites, our column results indicate that biodegradation will usually be negligible (7. ±.08) x -4 d -1 ml/d 2 (7.19 ± } x -4 d -1 ml/d Concentration (µm) C (6.9 ± 0.1) x -4 d -1 (7.9 ± 0.2) x -3 µm d -1 ml/d (7.9 ± 0.3) x -5 µm d D (7.94 ± 8) x -4 µm d -1 C Contact Time (Days) Figure 2.,, and C show T appearance kinetics for three influent concentrations in ase orden sediment. Influent concentrations were () 11.3 µm, () 1.13 µm, and (C) µm. D is a compilation of the plots shown in,, and C, showing that all of the data can be placed on one first-order appearance curve for a rate constant equal to 7.1 x -4 d -1. References 1. Landmeyer, J.E.; Chapelle, F.H.; radley, P.M.; Pankow, J.F.; Church, C.D.; Tratnyek, P.G. Fate of relative to benzene in a gasoline-contaminated aquifer ( ). Ground Water Monit. Rem. 1998, 18, Odencrantz, J.E. Implications of for intrinsic remediation of underground fuel tank sites. Remediation 1998, 8, Church, C.D.; Tratnyek, P.G.; Pankow, J.F.; Landmeyer, J.E.; aehr,.l.; Thomas, M..; Schirmer, M. Effects of environmental conditions on degradation in

5 model column aquifers. In: U.S. Geological Survey, Toxic Substances Hydrology Program, Proceedings of the Technical Meeting, Charleston, SC, 7-12 March 1999; Morganwalp, D.W.; uxton, H.T., Ed.; U.S. Geological Survey: West Trenton, NJ, 1999; Water Resources Investigations Report 99-18C, Vol. 3; pp Church, C.D.; Isabelle, L.M.; Pankow, J.F.; Rose, D.L.; Tratnyek, P.G. Method for determination of methyl tert-butyl ether and its degradation products in water. Environ. Sci. Technol. 1997, 31, orden, R.C.; Daniel, R..; Lerun, L.E., IV; Davis, C.W. Intrinsic biodegradation rates of and TEX in a gasoline-contaminated aquifer. Wat. Resour. Res. 1997, 33, Schirmer, M.; arker, J.F. study of long-term attenuation in the orden aquifer, Ontario, Canada. Ground Water Monit. Rem. 1998, 18,