Environ. Sci. Technol. 2001, 35, 374-378
Determination and Distribution of Diesel Components in Igneous Rock Surrounding Underground Diesel Storage Facilities in Sweden ANDERS LORE ´ N,† LOTTA HALLBECK,‡ KARSTEN PEDERSEN,‡ AND K A T A R I N A A B R A H A M S S O N * ,† Department of Analytical and Marine Chemistry, Chalmers University of Technology, SE-412 96 Go¨teborg, Sweden, and Department of Cell and Molecular Biology and Microbiology, Go¨teborg University, Box 462, SE-405 30 Go¨teborg, Sweden
In Sweden, a preliminary investigation of the contamination situation of igneous rock surrounding underground storage facilities of diesel showed that the situation was severe. The diesel was believed to have penetrated into the rock as far as 50 m from the walls of the vaults. Consequently, the risk for contamination of groundwater and recipients could not be neglected. To be able to assess the fate of diesel components in rock, both a suitable drilling method and a method for the determination of a wide range of diesel components were needed. The analytical method presented made it possible to quantify a number of hydrocarbons in rock samples collected with triple-tube core drilling. The samples were dissolved in hydrofluoric acid (HF) with hexane in Teflon centrifuge tubes. After digestion of the rock, extraction of the analytes with hexane was performed. Determination of the individual hydrocarbons present was done with gas chromatography-mass spectrometry (GC-MS). The method was used to study the environmental impact of the underground storage of diesel. The drilling method enabled sampling without contamination risks. Our data show that the major transport of diesel components in rock occurs through fracture systems and that diffusion of diesel through the rock is of minor importance. The results have drastically changed the view of the contamination situation of diesel in the vicinity of storage facilities in hard rock in Sweden.
Introduction Storage of diesel in large underground bedrock vaults has been used for decades in many countries, both for military and civil purposes. Due to a changed policy, 45 civil storage facilities in Sweden have recently been decommissioned. Before returning these to nature, it is necessary to ascertain that the vaults are environmentally safe and by no means will threaten groundwater or surface water quality in the future. The environmental concern is mainly focused on two questions: (i) How much of different diesel components are actually left in the surrounding rock after decommissioning? * Corresponding author phone: +46 31 772 27 80; fax: +46 31 772 27 86; e-mail:
[email protected]. † Chalmers University of Technology. ‡ Go ¨ teborg University. 374
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 2, 2001
and (ii) Will possibly remaining diesel components migrate and threaten drinking water sources? The storage facilities may consist of up to six vaults, each with a volume of approximately 10 000 m3 (90 × 10 × 12 m) and are situated some 50 m below ground. The walls of the vaults are usually unlined; therefore, the stored products have been in direct contact with a large bedrock volume. The underground vaults are always located below the groundwater table. Continuous pumping of the incoming groundwater is performed from the bottom of the vaults, and consequently, a groundwater cone of depression is created. This will function as a trap for diesel because of the surrounding hydrostatic pressure, the low solubility of diesel components in water, and the density difference between diesel and water. Additionally, the continuous inflow of groundwater to the vaults will hinder diesel migration to the surrounding rock. During the fill-up phase, the groundwater level must be lower than during storage, and diesel might then have escaped into fractures. Steady state is accomplished during long-time storage, but migration of diesel into fractures in the ceiling of the vaults can take place during this period. Several investigations have dealt with the transport of contaminants in fractures in porous media, since this is most probable (1). Although the pore space was not anticipated to harbor a major fraction of the hydrocarbons because of the low average porosity of granite and gneiss, diffusion of hydrocarbons into the rock was also considered as a possible source of contamination. Thus, it cannot be excluded that large volumes of the surrounding rock are contaminated, and this poses an environmental threat to groundwater when the pumping is stopped and the groundwater returns to its normal level and flow paths. One storage facility has been opened up via the construction tunnel into the underground vaults. Inside, the vault surfaces were washed with detergents and hot water. Initial investigations were done in this storage facility and retrieval of drill-cores through walls and floor indicated the presence of diesel all along the 50-m cores. Estimates of the amount of diesel that was left in the rock around the storage indicated values up to several tons. This initiated a thorough research and technical development program to find methods to be used in the decommissioning process. Because of the enormous risk of explosions associated with opening and cleaning of the vaults, together with very high costs, it was decided that all sampling and cleaning processes for other storage facilities have to be done from the ground. To be able to assess the fate of diesel components in the surrounding bedrock, both a reliable drilling method and a method for the determination of a wide range of diesel components were needed. Several analytical methods have been applied to determine the occurrence of diesel for porous and homogeneous matrixes, such as soil, ash, and sedimentary rock, but their applicability to igneous rock is low. They include Soxleth extraction (2), laser fluorescence (3), pyrolysis (4-6), and supercritical fluid extraction (SFE) (2, 7, 8). A wide range of analytes has been quantified with these methods, from low molecular weight compounds (CH4) to polyaromatics (benzoperylene). However, to obtain good extraction yields, a large number of parameters such as solvent, pressure, temperature, additives, and time need to be optimized for each method. Also, degradation of analytes has been reported for SFE (2), and discrimination has been observed for SFE and laser pyrolysis (6). It is also of interest to minimize analyte loss through evaporation, especially for light diesel fractions. 10.1021/es991168r CCC: $20.00
2001 American Chemical Society Published on Web 12/07/2000
This can be achieved by keeping the solid samples in one piece and avoid grinding. In this paper, we present the development of an accurate drilling method and a reliable analytical procedure with few manipulation steps for the determination of diesel components in igneous rock. Triple-tube core drilling with thorough cleaning of the equipment was applied. The analytical method developed dissolves the silicate skeleton of the rock by means of hydrofluoric acid (HF). The solid-phase rock is thereby converted into a solution, and the analytes can be extracted into an organic phase prior to GC-MS analysis.
Experimental Section Analytical Procedures. Drill cores were put on CO2(s) and subsequently smashed with a sledgehammer. Subsamples (≈4 g) were collected from the center of the core and put into Teflon centrifuge tubes, and 15 mL of hydrofluoric acid (48%, Riedel-de-Haen, p.a.) and 0.5 mL of hexane (J. T. Baker, p.a.) were added. The digestion of the silicate skeleton proceeded at room temperature until the rock was dissolved completely. During the digestion (≈10 d), the samples were shaken intermittently. After dissolution, the rock samples were extracted for 30 min with hexane, where the hexane layer was removed and the internal standard 1-chlorohexadecane (Merck) was added. The separations and determinations were performed with a gas chromatograph (Varian 3400) coupled to an ion-trap mass spectrometer (Varian Saturn 2000). The separations were made with a 30-m capillary column (XTI-5, RESTEK) with an i.d. of 0.25 mm and a film thickness of 0.25 µm. Helium was used as carrier gas with a flow of 1 mL/min. The column temperature was initially 70 °C for 1 min and then increased to 250 °C at 5 °C/min and held for 15 min, which gave a total time of 58 min. The injector temperature was 275 °C, and 4 µL was injected splitless with an automatic injector (Varian 8100). The needle remained in the injector for 10 s after injection. Electron impact (EI) mass spectra of GC effluents were scanned from m/z 40 to 399 at a rate of 2 scan/s at an ion source temperature of 200 °C. The electron beam energy was 70 eV. The substances determined were normal alkanes from dodecane (C10H22) up to pentacosane (C25H52). All analytical standards were prepared in hexane, either from solutions (AccuStandard) or from neat compounds (AccuStandard, Merck, Supelco). The determinations were made with onepoint internal standard calibrations within the linear range of the instrument. Calibration graphs for the individual compounds were made. Specific mass fragments were used for quantification according to Table 1. The extraction efficiency was determined by extracting diesel-saturated water several times. Stability Tests. To show the resistance of organic compounds to degradation in HF, neat n-tetradecane was added to HF and a mixture of minerals (red granite, leptite, and gray granite). The samples (n ) 3) were stored at room temperature for 14 d. Controls (n ) 3) with C14H30 without additions of HF were prepared and treated in the same way as the samples. The extraction of the n-alkane was made with hexane (5 mL) for 30 min. The hexane was removed, and the internal standard was added. The extracts from each tube were diluted to three different concentrations (40, 22, and 4 mM) and analyzed with GC-MS. This yielded normalized area counts for the compound in each tube corresponding to the amount added to the tubes. To compare the HF-treated samples with the nontreated controls, the normalized area counts were scatter-plotted against each other with the independent variable (untreated) on the x-axis. A linear regression was performed, and the standard deviations for the intercept and slope were calculated on the 95%
TABLE 1. Parameters Describing the Standard Curves for Individual Alkanesa compound
r2
m /z
rel. SD (%)
LOD (µg/L)
LOD (µg/kg)
decane undecane dodecane tridecane tetradecane pentadecane hexadecane heptadecane octadecane nonadecane eicosane heneicosane docosane tricosane tetracosane pentacosane
0.989 0.999 0.999 0.999 0.999 0.998 0.999 0.999 0.999 0.998 0.996 0.991 0.997 0.992 0.997 0.996
57 + 71 57 + 71 57 + 71 57 + 71 57 + 71 57 + 71 57 + 71 57 + 71 57 + 71 57 + 71 57 + 71 57 + 71 57 + 71 57 + 71 57 + 71 57 + 71
3.8 3.3 2.6 6.4 5.5 5.8 4.1 2.6 3.3 4.5 4.7 1.1 0.38 5.2 4.9 2.6
54 48 40 33 30 25 20 15 24 17 13 10 8.6 7.4 6.7 6.5
3.0 2.7 2.2 1.8 1.7 1.4 1.1 0.8 1.3 1.0 0.7 0.6 0.5 0.4 0.4 0.4
a r 2 is the regression coefficient for the curves; m/z is the specific mass fragments used for quantification; and LOD is the limit of detection calculated as three times the standard deviation of the noise.
confidence level. The calculated values were compared to the ideal values (slope ) 1, intercept ) 0). Test Samples. Diesel components are believed to be located in the pore space between the mineral grains in the rock. Consequently, recovery experiments are impossible to perform due to the difficulties of spiking the samples and getting the analytes into the pore space. Artificial test samples were prepared in order to conduct laboratory experiments. The igneous rock used as test samples was from Småland, Sweden. It had a porosity of about 0.2% and consisted mainly of fine quartz and feldspar crystals. Oven-dried (150 °C) pieces of granite were placed in a desiccator, and vacuum (
