Документ взят из кэша поисковой машины. Адрес оригинального документа : http://geophys.geol.msu.ru/shevnin/publ/109.pdf Дата изменения: Wed Oct 17 15:56:22 2012 Дата индексирования: Sat Feb 2 23:39:41 2013 Кодировка:

INVESTIGATIONS OF OIL POLLUTION WITH ELECTRICAL PROSPECTING METHODS
Modin I.N., Shevnin V.A., Bobatchev A.A., Bolshakov D.K., Leonov D.A., Vladov M.L. Moscow state university Russia, 119899, Moscow, MSU Faculty of Geology, fax: (+7-095)-9394963 E-mail: sh@geophys.geol.msu.ru
This paper was published in Proceedings of the 3rd Meeting environmental and engineering geophysics. Aarhus, Denmark, 8-11 September 1997. P.267-270.
The most widespread source of pollution in Russia is the oil pollution. It occurs at all stages of oil production, transportation and processing. Outflows, proceeding during decades result to formation of oil secondary deposits. In a near-surface zone, oil pollution becomes especially chemically active and reacts with geological environment, that results in the anomalies of various geophysical methods: SP, IP, GPR and resistivity. The oil pollution is an unusual object due to its ability to oxidation and mobility. The pollution causes processes, occurring with the speed, differed from natural geological processes. Changes of rock properties, caused by oil pollution are inconsistent. A priory, for example, the oil is an isolator, but frequently it causes anomalies of lowered resistivity. For engineering and environmental studies with electrical methods in B MN A 5m 1 A urban areas (including pollution studies) the main problem is the influence of 5 20 Ohm.m MN A geological noise. Upper part of cross-section includes many near-surface inho2 10 mogeneities (NSI), caused with artificial ground, asphalt cover, trenches, ca100 Ohm.m A MN 3 15 bles, tubes, etc. These inhomogeneities create strong distortions and influence 30 Ohm.m MN A like "the broken glass", preventing from clear seeing deep objects. NSI distort4 20 m ing influence can be canceled by the application of total electric sounding (TES) technology, developed in MSU. , Ohm.m 100 C In TES technology we apply 0 4 two-sided pole-dipole array with step 2 of distance growth equal to step be50 tween VES sites along profile. NSI NSI 3 1 creates distortions of four different 30 types (fig.1C), when movable (single 20 or dipole) element crosses the NSI, or unmovable (single or dipole) element is placed in NSI limits (fig.1B). Some AO, m 10 5 3 10 20 40 60 distortions of sounding curves are Fig.1. Example of NSI (A), types of conformable (fig.1C, 1-2), whereas Fig.2. The scheme of nonsurvey (B) and sounding curves ( C) others are non-conformable (fig. 1C, conformable distorting effect's origin 3-4). On apparent resistivity cross-section distortions display as vertical and for pole-dipole array. inclined strip zones (fig.2,3). For different types of array (Pole-dipole, polepole, Schlumberger, Wenner, dipole-dipole, etc.) distortions are Rmin R different (fig.3). For several NSI (fig.3, 3) interference of distortions 1 4 appears. We developed software for canceling such distortions. The DAS, AM AMN algorithm of median polishing was offered by J.W.Tukey (1981), Rmin ABmin and after modification made by E.V.Pervago was applied for proc5 essing of Total Electrical Sounding (TES) data. The algorithm's 2 operation is based on regularity of distortion effects and it allows MNB to cancel effectively these effects on apparent resistivity pseudoSCHLUMBERGER AMNB cross-section. J.W.Tukey described processing algorithm for data, 6 3 given in the table as following: at the first step - counting and subtracting median value for each column; at the second step WENNER counting and subtracting median value for every row. Then the 1st Fig.3. The schemes of distorting effects appearand 2nd steps repeat several times. ance on cross-sections for different arrays.
a

a


We present soundings data as the table of a logarithms, where each spacing corresponds to a row, and each sounding site - to a column. The effect of horizontally layered medium in such table will be displayed simultaneously for all sites, conformable distortion will be displayed simultaneously on all spacings, and non-conformable distortion - as lines, inclined under 45° or another angle to the left or to the right, depending on type of array and sounding technology. To this table we apply algorithm of median polishing, but in addition to rows and Fig.4. Geoelectrical model with columns we shall also include in layering, deep objects and NSI. processing inclined lines, corresponding to non-conformable distortion. The result of algorithm operation is the decomposition of an initial field into several components, connected with: a) position of movable and unmovable electrodes; b) horizontally layered medium and c) some rests. After decomposition each component is filtered from Fig.5. Interpretation results for the high-frequency noise, connected basically with near-surface inhomogeneities model from fig.4 with filtering geoand errors of measurements. Parameters of filtering (width of smoothing logical noise (b) and without it (a). window) are get out by user, depending on survey conditions and geological structure. At the last stage the reconstructing process is applied - the smoothed components are united back into a field. Thus, the technology of processing with the program Median allows to reveal effects of distortions caused by near-surface inhomogeneities and to remove them, to see effects from deep inhomogeneous objects and to separate the influence of horizontally layered medium. Layered medium can be subjected to quantitative 1D interpretation. Deep objects can be interpreted with the help of 2D forward problem account, including inhomogeneities and layered medium. The role of NSI or geological noise is shown on fig.4-6. Canceling geological noise allows to trace boundaries with greater accuracy (fig.5b) and smaller fitting error (fig.6b). That is true Fitting error, % for field data also. Electrical survey in urban regions can be fulfilled without galvanic contacts with the ground. GPR (georadar) is very useful instrument for that. GPR has small penetration depth and problems with depth estimation. We use GPR together with electrical sounding, which gives reliable depths and has penetration deeper than GPR. At small depth GPR has higher detaility. Traditional electrical sounding needs electrode grounding. In Russia during more than 40 years a non-contact resistivity technology is used. An instrumentation for such VES sites measurements is produced in Petersburg. ERA-V instrument can operate withFig.6. Fitting errors for interpretation out galvanic contacts at frequency 5 and 600 Hz. At 600 Hz we can use non-contact current and measuring lines, at 5 Hz only measuring lines. This results from fig.5 with filtering geomodification has two main advantages. 1. In theory the result of measurement logical noise (b) and without it (a). with alternating current can be equivalent to DC resistivity, when the dependence of measurements from working frequency is absent. Thus, maximal distance for frequency-independent area is the function of the earth resistivity and working frequency. (it is determined by the near zone condition). On frequency 625 Hz for example, when =1 Ohm.m, this zone limit is at r=10 m, when =10 Ohm.m, it exist up to r=32 m, when =100 Ohm.m, r=100 m. 2. In practice non-contact survey is similar to EM methods with their simplicity in field operations. Electrical sounding results shown on fig.9-10 were measured with non-contact resistivity technology. Ecological studies of oil pollution on the department of geophysics MSU were carried out for several years, including town Noginsk (1993), on oil refining factories in Moscow (1994), town N (1994-96) and town S (1995-96). The depth of polluted zone was from 0.5 up to 50 m and the time of its formations from several months till 50 years. A typical example is the Moscow oil refining factory in Kapotnya. On the factory in Kapotnya oil polluted zone is placed at the depth of several meters. For its study a complex of methods VES, SP, IP, GPR was applied (fig.7). A geological situation is typical for Moscow region. On depth about 10 m the waterproof layer of Jurassic clay is situated, and higher of it is sand, the lower part of which is filled


with water. The groundwater level is on the depth about 1.5 m. On that boundary the oil pollution is placed, being existed constantly for a long time. The top part of a crosssection with the thickness of 2-3 m is an artificial ground, which is typical for urban areas and has extremely changeable properties. Just in this layer several anomalous zones with intensive oil pollution are located. These zones are confidently mapped on IP, SP, GPR and resistivity methods' data. Polluted object is appeared similar to ore body, because it shows low resistivity, negative SP potential, high IP response. GPR shows low dielectric parameters in polluted area. Such low-resistive anomalies in the areas of oil pol- Fig.7. Results of electrical survey above shallow-depth lution were found also by H.Vanhala (Finland)(1994, 1997). oil pollution in Kapotnja. Why does oil pollution seem to be low resistive object? In accordance with investigations some scientists: Bailey N.J.L. (1973,1981), Evans C.R. (1977,1981), Rogers M.A. (1973, 1977), Dostalek M. (1975) oil-transforming bacteria are probably responsible for that. In upper part of the cross-section bacteria are very active, they transform upper layer of the oil film into some acids. Acids react with rocks and iron ions in water and result in high groundwater conductivity, karst processes and pyrite creation. Pyrite gives increased IP response. The GPR data were interpreted independently, and other methods -- in integration. The GPR data only after processing were possible to reveal layered structure and the sites of pollution. The total electrical sounding -- TES was used as a structural method for tracing ground-water level (GWL) and a clayey basement. With the help of an integrated parameter of pollution it became possible to find out its distribution in the area and on the depth. We used a priory information about the thickness and exact places of pollution at the reference profile. To characterize pollution we applied such parameters like the relative thickness of a polluted layer and its bottom's depth. These values were estimated on the regression dependence between these two parameters and some anomalous values of Fig.8. Results of polluted layer thickness the measured geophysical fields and their dispersions (fig.8). For estimation on regression with geophysical localization of pollution in plan IP method gave the maximum condata. tribution, about 80% of the all information. SP and resistivity gave less information. GPR information gave wider polluted areas than geological data. We suppose that GPR could see the whole area of pollution whereas the direct geological methods could see only the strongest pollution. In town N the study of territory around the oil refining factory was carried out for estimating ecological consequences of this factory activity. In upper part of a geological cross-section on depths up to 50 m during several decades of factory activity in the result of oil leakage from pipelines and storehouses the significant amount of oil

Fig.9. Town N, a field measured with non-contact electrical sounding technology before and after canceling geological noise

Fig.10. Results of TES interpretation with low resistive traces of oil pollution for data from fig.9.


Fig.12. Town S. Profile 4. VES results with traces of Fig.11. Results of TES interpretation over zone of oil oil pollution. pollution, town N, profile 13 products was collected. That can be considered as a rich secondary oil deposit. The pollution spread to significant distance (several km), despite extremely low values of rock permeability. The geophysical study in town N was carried out by VES in several stages due to great area for investigation. Each stage included several VES profiles with about 200 VES sites measured on TES (the total electrical sounding) technology with canceling geological noise. The study of the area (taking into account a priory data and results of drilling) has allowed to reveal layered structure of a cross-section and to find out paleovalleys, filled with loose deposits. The system of such paleovalleys serves the most probable ways of oil pollution spreading at significant distances. This example shows that not only background cross-section, but some anomalous features in it play the important role in pollution state and movement. TES is the main method for such pollution study. Town S. is not far from town N., and oil-refining factory is quite similar. Pollution on the surface is the same or even stronger, but secondary oil deposit is not found yet. Background cross-section is sandy with thin clay layers. We found traces of pollution spreading Fig.13. Town S. VES results with traces of oil downward, noticeable on TES method data, but not oil deposits. pollution on profile 2. REFERENCES 1. Bailey N.J.L., Krouse H.R., Evans C.R. Alteration of crude oil by waters and bacteria - evidence from geochemical and isotope studies. American association of petroleum geologist bulletin, N5, 1981. 2. Bailey N.J.L., Jobson A.M., Rogers M.A. Bacterial degradation of crude oil: comparison of field and experimental data. Chemical geology, N11, 1973. 3. Dostalek M. The action of microorganism on petroleum hydrocarbons. Microbiology, 1975. 4. Milner C.W.D., Rogers M.A., Evans C.R. Petroleum transformations in reservoirs. Journal of geochemical exploration, N7, 1977. 5. Vanhala H. Mapping oil contaminated sand and till by spectral induced polarization method. EAEG 56th annual meeting, Zeist, 1994. 6. Vanhala H. Mapping oil contaminated sand and till with the spectral induced polarization (SIP) method. Geophysical Prospecting, 1997, Vol.45, No. 2, p.303-326. 7. I.N.Modin, V.A.Shevnin, A.A.Bobatchev, D.K.Bolshakov, A.A.Gorbunov. Investigations of oil pollution, caused by oil-industrial plants with electrical prospecting methods. Annales Geophysicai. European Geophysical Society. Part 1. Society Symposia, Solid Earth Geophysics & Natural Hazards. Supplement 1 to Volume 14. P.171. 1996.