Comment to article in Geophysics
Why Key’s Geophysics article does not apply to PetroMarker's technology.
A discussion by Terje Holten, PhD, PetroMarker
The article shows an interesting comparison between various acquisition configurations for marine CSEM. An electric transmitter dipole is directed in all three directions, and all resulting electrical and magnetic components are computed by forward modeling. After addition of noise, these fields are inverted, and the comparison is based on the recovered 1D resistivity profile. The author concludes that the inversions of the vertical field component have less sensitivity to the resistive target.
The article only focuses on variations of the traditional horizontal frequency-domain CSEM prospecting systems, which puts an unintentional bias on the article. The assumption that the source has to be moving during measurements is a prime example of this bias. For a stationary system of vertical source and receiver with accurately known positions, the relative noise of 1% ‘as a simple way to mimic the increased relative uncertainty in the navigated source receiver vector at short offsets’ can be seen as an exaggeration. During time-domain measurements, the transmitter is turned off, and the uncertainty in the transmitter vector does not translate into a random noise. All data sets in the article use information from receivers with offset up to 20 km, but the vertical electrical field from a vertical transmitter only gives useful information to receivers in the electric near field up to approximately 2 km (Holten et al., 2009).
PetroMarker has practical experience with stationary vertical transmitters and receivers in time-domain (Holten et al., 2009). We use a vertical transmitter with length approximately equal to the sea depth. In the paper the noise is the sum of the absolute noise of the receiver and a relative noise coming from the transmitter, which depends on offset. The electric field noise is normalized to transmitter length and current. The value used in this paper for the minimum absolute noise of 10-15 V/Am2 is a good guess for our noise levels. For instance in 1000m deep waters with 10m receiver and 5000A current, a noise of 50 nV is equal to 10-15 V/Am2. Our noise levels are usually between 5nV and 50nV, depending on measurement conditions and the estimation of systematic errors. An 1% relative noise gives about 10-13 V/Am2 amplitude at offset 500m, which is 100-1000 times larger than our observed noise. With this large, unrealistic noise level, the inversion of the vertical electric field component obviously fails.
The conclusions of the article must therefore not be taken at face value.
This article models a quite different situation from PetroMarker’s own technology, and does not apply to it.
Basically, the modeling is done for a wide range of offsets, corresponding to a stationary receiver, a moving vertical transmitter in frequency domain.
However, the PetroMarker’s technology uses stationary transmitter and receivers in time domain and the offsets are small (<2 km). There is a large difference at small offsets between time domain (where the response is measured while the transmitter is turned off) and frequency domain for which the direct EM signal is large.
References:
Holten, T., Flekkøy, E.G., Singer, B., Blixt, E.M., Hanssen, A., Måløy, K.J. [2009] Vertical source, vertical receiver, electromagnetic technique for offshore hydrocarbon exploration. First Break,27, 89-93.
Key, K., 2009, 1D inversion of multicomponent, multifrequency marine CSEM data: Methodology and synthetic studies for resolving thin resistive layers: Geophysics, 74, F9–F20.
Link to Kerry Key article:
© PetroMarker AS 2007 
(+47) 51 20 16 10 
(+47) 51 20 16 19 
petromarker@petromarker.com