Objective
The project will test and implement an innovative method to delineate and monitor reservoirs in a known field by inverting seismic reflection and multichannel transient electromagnetic (MTEM) data together. Seismic methods cannot distinguish between oil and water. The electrical conductivity of the rocks is very sensitive to variations in porosity and the conductivity of the pore fluid. Hydrocarbons are resistive and water is conductive. Multichannel transient electromagnetic (MTEM) methods can locate variations in conductivity. We are: (1) using the excellent vertical resolution of seismic and borehole data to define the major sedimentary interfaces; (2) using resistivity measurements from deep boreholes to determine the vertical trend of the electrical conductivity variation; (3) using the lateral variation of the conductivity determined from time-lapse MTEM data to monitor movement of water and hydrocarbons in a gas-storage reservoir.
Data were recorded at St. Illiers la Ville in October 1994 along a 4 km line across the underground gas storage reservoir with dipole transmitter spacing of 250 m, in-line electric potential field recordings of 125 m, cross-line potential and vertical component of magnetic field recordings at 250 m intervals. During EM transmission and recording, the cathodic protection of the network of pipes and wells was switched off, which reduced the cultural noise considerably. The signal-to-noise ratio on the recorded data before stacking was still poor, but the noise was not correlated with the signal and, after stacking appproximately 100 transmissions at each source position, we were able to achieve the desired signal-to-noise ratio about 100. Methods for suppression of cultural noise before stack have been applied, including notch filtering and prediction error filtering, but these are no more effective than stacking without pre-stack noise suppression.
Common-offset gathers of the stacked data show variations from trace-to-trace that are much greater than those that would be expected by one-dimensional modelling and changing the resistivity in the reservoir layer. We attribute these variations not only to three dimensional effects : the fields induced by the conducting pipeline and well network at the site must also have a very significant effect. These effects were expected. In the repeat survey of August 1996, the measurements were almost identical with those obtained in October 1994. In particular, the trace-to-trace variations that were so difficult to interpret in the first survey are repeated in the second survey. We expect them to disappear when we substract the data sets. In both surveys we were able to measure the response of the system, independent of the response of the earth. The data can be deconvolved for this system response to recover the earth impulse response, uncontaminated by the system. The detectable impulse response is only tens of milliseconds in duration. This has implications for the design of the next generation of equipment which we are evaluating.
The system response exhibits nonlinear variations of the order of a few per cent with transmitter current and load resistance. Since the system response was not measured at each transmitter position, but only in a separate experiment, this causes uncertainties in the deconvolved data of the same order. In comparing the 1994 and the 1996 data we have found errors of a few milliseconds in the time-break which vary with source position. Software must be developed to cope with these errors before substracting the two data sets.
Analysis of the well data at St. Illiers la Ville at the times of the two surveys shows changes in the gas pressure and water level that are not uniform over the site and may be detectable once the data processing can be made to cope with the new found faults in the acquired data.
It is a major concern to the oil industry to be able to detect lateral variations in porosity and fluid content and hence delineate and monitor the hydrocarbon reservoirs. The data acquisition and processing methods we are testing and implementing in this project have applications in many parts of Western Europe, as well as in other continents. If this method is successful it will improve the probability of delineating and monitoring the oil-bearing reservoirs in known fields using non-destructive geophysical methods alone. The risk of drilling dry holes is reduced and the number of holes required to be drilled is therefore reduced.
This method can also be used to detect the top and flanks of salt domes and is especially applicable where the seismic method fails because acoustic contrasts are too small or because the dips are too steep. In production of hydrocarbons, the electrical contrast between the hydrocarbons and water allows the movement of the oil-water contact to be monitored. The same principle can be used to monitor underground gas storage reservoirs.
An MTEM survey was carried out in October 1994. A repeat survey was carried out in August 1996. The difference between the two data sets is expected to be caused primarily by the replacement of the more resistive gas by the more conductive salt water. Modelling of the site indicated that the replacement of the estimated volume of gas by water between October and April would give an anormaly of about 5% of the measurement. Inversion of the difference data, using the known structure to constrain the model, will show where the resistivity variations are occurring. The results of this experiment will be checked against measurement made in the site control boreholes. We are now working on the data.
Programme(s)
Topic(s)
Call for proposal
Data not availableFunding Scheme
DEM - Demonstration contractsCoordinator
EH9 3JW Edinburgh
United Kingdom