Transformation of CSEM data for determination of resistivities by seismic data processing methods.
Lead Research Organisation:
University of Edinburgh
Department Name: Sch of Geosciences
Abstract
INTRODUCTION
Most oil and gas is produced from porous underground or subsea reservoirs discovered by seismic surveys. Normally the pores are filled with salt water. Seismic signals do not distinguish oil-saturated from water-saturated reservoirs, so wells are drilled to determine the fluids present in the reservoir. Three out of four exploration wells find no hydrocarbons and are "dry." The cost of finding new hydrocarbon reserves can be reduced by reducing the number of 'dry' wells drilled - £1 million to £100 million per well, depending on water depth and depth of well.
Hydrocarbons are electrically resistive, whereas salt water is conductive. It follows that electromagnetic methods have the potential to distinguish between water and hydrocarbons before drilling and thus reduce the number of dry wells. Since 2002 a method for conducting offshore electromagnetic surveys has been developed for the oil industry, known as the controlled source electromagnetic (CSEM) method. It has become an accepted tool in the search for sub-sea hydrocarbons, although it is still immature and there are many technical problems to solve.
RESISTIVITY AND ELECTROMAGNETIC WAVE PROPAGATION
The propagation of the electric and magnetic fields is affected by the electrical conductivity of the medium. Electrical conductivity is the reciprocal of resistivity: the more easily a medium is able to conduct electricity the less resistive it is. As the electromagnetic fields propagate they lose energy because electric current flows in conducting material: the greater the conductivity of the medium, the greater the loss of energy. This limits the depth of investigation. Sea water has a conductivity of about 3 S/m (resistivity 1/3 ohm-m); North Sea Tertiary sediments have a conductivity of about 1 S/m (resistivity 1 ohm-m). A sandstone reservoir with a resistivity of 1 ohm-m can have a resistivity as high as 1000 ohm-m when saturated with hydrocarbons. This has been well known for years from well logs. So the resistivity contrast can be two or three orders of magnitude when hydrocarbons are present.
Seismic impedance contrasts are much smaller: for an oil/water interface the contrast is a few per cent. Gas gives a higher impedance contrast than oil, but a small amount of gas can give the same response as a large volume of gas. It is difficult to quantify the amount of gas from the seismic response. The CSEM method is complementary to the seismic method: it can provide additional quantitative information about the reservoir fluids.
THE PROBLEM
We are concerned with the determination of the resistivities from the CSEM data. Conventionally, EM data are interpreted by inversion. That is, the response of a resistivity model of the earth is synthesized in a computer using the same acquisition geometry as for the field data, and the result is compared with the field data. The resistivity model is adjusted until the synthesized data match the field data within an acceptable error.
What comes out of the inversion is only what has been put into the model by the geophysicists. It has not been derived from the measured data. This contrasts with the seismic method in which seismic velocities are determined from the seismic data by lining up seismic arrivals. This principle is fundamental to seismic data processing.
THE PROPOSAL
Seismic data obey the wave equation: the wave travels without losing energy. Electromagnetic wave fields obey the diffusion equation. The calculation of the electromagnetic data has been formulated as a weighted sum of waves. The waves are an intermediate step in the calculation. We propose to transform the diffusive electromagnetic data to this intermediate domain, where they may be manipulated just like seismic data, enabling the resistivities to be determined directly: the theory says the resistivities should be proportional to the square of the determined velocities.
Most oil and gas is produced from porous underground or subsea reservoirs discovered by seismic surveys. Normally the pores are filled with salt water. Seismic signals do not distinguish oil-saturated from water-saturated reservoirs, so wells are drilled to determine the fluids present in the reservoir. Three out of four exploration wells find no hydrocarbons and are "dry." The cost of finding new hydrocarbon reserves can be reduced by reducing the number of 'dry' wells drilled - £1 million to £100 million per well, depending on water depth and depth of well.
Hydrocarbons are electrically resistive, whereas salt water is conductive. It follows that electromagnetic methods have the potential to distinguish between water and hydrocarbons before drilling and thus reduce the number of dry wells. Since 2002 a method for conducting offshore electromagnetic surveys has been developed for the oil industry, known as the controlled source electromagnetic (CSEM) method. It has become an accepted tool in the search for sub-sea hydrocarbons, although it is still immature and there are many technical problems to solve.
RESISTIVITY AND ELECTROMAGNETIC WAVE PROPAGATION
The propagation of the electric and magnetic fields is affected by the electrical conductivity of the medium. Electrical conductivity is the reciprocal of resistivity: the more easily a medium is able to conduct electricity the less resistive it is. As the electromagnetic fields propagate they lose energy because electric current flows in conducting material: the greater the conductivity of the medium, the greater the loss of energy. This limits the depth of investigation. Sea water has a conductivity of about 3 S/m (resistivity 1/3 ohm-m); North Sea Tertiary sediments have a conductivity of about 1 S/m (resistivity 1 ohm-m). A sandstone reservoir with a resistivity of 1 ohm-m can have a resistivity as high as 1000 ohm-m when saturated with hydrocarbons. This has been well known for years from well logs. So the resistivity contrast can be two or three orders of magnitude when hydrocarbons are present.
Seismic impedance contrasts are much smaller: for an oil/water interface the contrast is a few per cent. Gas gives a higher impedance contrast than oil, but a small amount of gas can give the same response as a large volume of gas. It is difficult to quantify the amount of gas from the seismic response. The CSEM method is complementary to the seismic method: it can provide additional quantitative information about the reservoir fluids.
THE PROBLEM
We are concerned with the determination of the resistivities from the CSEM data. Conventionally, EM data are interpreted by inversion. That is, the response of a resistivity model of the earth is synthesized in a computer using the same acquisition geometry as for the field data, and the result is compared with the field data. The resistivity model is adjusted until the synthesized data match the field data within an acceptable error.
What comes out of the inversion is only what has been put into the model by the geophysicists. It has not been derived from the measured data. This contrasts with the seismic method in which seismic velocities are determined from the seismic data by lining up seismic arrivals. This principle is fundamental to seismic data processing.
THE PROPOSAL
Seismic data obey the wave equation: the wave travels without losing energy. Electromagnetic wave fields obey the diffusion equation. The calculation of the electromagnetic data has been formulated as a weighted sum of waves. The waves are an intermediate step in the calculation. We propose to transform the diffusive electromagnetic data to this intermediate domain, where they may be manipulated just like seismic data, enabling the resistivities to be determined directly: the theory says the resistivities should be proportional to the square of the determined velocities.
Organisations
Publications
Werthmüller D
(2014)
Predicting controlled-source electromagnetic responses from seismic velocities
in Interpretation
| Description | We are working on the controlled-source electromagnetic exploration method. It is used to find subsurface resistivities in the Earth. When hydrocarbons are present in a rock the resistivity of the rock increases dramatically. So this method can be used to search for hydrocarbons, including shale gas. The difficulty is to extract the subsurface resistivities from the electromagnetic data. Up till now, resistivities have been found by a process called inversion, which is a search for an Earth model, whose response closely matches the measured data. The problem is to find the correct model. We have now found a way to extract Earth resistivities directly from the controlled source electromagnetic data, provided we can determine the complete Earth impulse response between the electromagnetic current source and the electric field receiver, for all measured source and receiver positions. |
| Exploitation Route | We intend to apply our new method to real data. If it works, it will enable the location of potential subsurface hydrocarbon reserves to be identified from the Earth's surface, without drilling. In the exploration for hydrocarbons, including shale gas, it will reduce the number of "dry" exploration wells that do not find hydrocarbons, thus saving millions of dollars and also, therefore, reduce the environmental impact. Our method can be applied onshore or offshore and can be used to monitor carbon dioxide sequestration, because the presence of carbon dioxide also increases the resistivity of the reservoir rock. |
| Sectors | Digital/Communication/Information Technologies (including Software),Electronics,Energy |
| Description | This small catalyst grant is helping us develop a data processing scheme for broad-bandwidth Earth impulse responses recovered from controlled source electromagnetic (CSEM) data, both onshore and offshore. The goal is to recover subsurface resistivities directly from the data. Our results applied to synthetic, or modelled data, are encouraging and we are now applying for funding to obtain suitable real data in land field experiments. We are also preparing a paper for publication and will submit it once we have completed our solution. |
| First Year Of Impact | 2014 |
| Sector | Energy |
| Impact Types | Economic |