Seismoelectrics
The seismoelectric effects, caused by a seismic excitation on a porous saturated medium, are based on an electrokinetic coupling initiating at the grain-fluid boundary (Fig.1).
Figure 1: Illustration of seismoelectric effects of the two kinds associated with seismic propagation in a saturated porous medium: (1) Seismically induced fluid flow from compression to dilation zone creating a streaming current, known as the coseismic electric field, concomitant to the propagation of the compression wave. (2) Heterogeneities in the coseismic field due to the incident wave front hitting an interface are adjusted by a vertical electric dipole created at the first Fresnel zone. G and D represent the typical layout for geophones and electrical dipoles during field acquisition.
Due to the periodical relative movement between the fluid and the matrix occurring in a saturated porous media under seismic excitation, a coseismic electrical signal appears whose origins take place at microscopic scale in the so-called double-layer. The coseismic wave is proportional to the seismic particles acceleration (as the similarity between the electrogram and the P-seismogram confirms) and depends among others on the pore-geometry and the fluid’s characteristics. The intensity of this EM-wave eventually combines the influence of various hydraulic parameters including: fluid’s electrical conductivity, porosity, hydraulic conductivity, viscosity, zeta-potential and tortuosity.
By the arrival of the common wavefront on a geological (as sand / clay) or hydraulic (as freshwater / saltwater) discontinuity, a charge separation perpendicular to the interface may occur. This charge separation generates a vertical electric dipole, which radiates an EM-wave called converted wave.
This wave, converted from seismic to electromagnetic at the interface, propagates with EM-velocity up to the earth surface, where it will be recorded almost simultaneously on every electric dipole. In that manner, it forms a levelled line only to be seen on the electrogram (Fig. 3). If this interface is consistent with a seismic discontinuity, the arrival time of the converted wave amounts to half the traveltime at shot point of the corresponding seismic reflection. As is typical for a vertical electric dipole, the amplitude decays in 1/r³ with the offset. This property tends to restrain the use of the seismoelectric methods to near-surface application.
Fig. 2: Field layout with overlapping dipoles D (length of 1.5 m, directly cabled to preamplifiers). The geophones G (spaced by 1m) stand at the corresponding dipole midpoints.
At the LIAG, seismoelectrics is being employed and enhanced at both field and laboratory scales. The field measurements, on which characteristics we will subsequently comment, have already delivered repeatable results. In the laboratory the first auspicious experimental signals were observed.
By field measurements, both geoelectrical and seismic signal are recorded simultaneously (Fig. 2). For the transient recording we used a Geode (24 bit) on which we cabled the geophones but also the electric dipoles. The seismic excitation was optimised using repeated hammer strikes with manual trigger. Thanks to self-made preamplifiers the electrical signal was lifted over the noise-level straight after its entry. Through there is no strict imperative to record the seismic signal during seismoelectric measurements, at this development-stage it remains a necessity. As a matter of fact, the seismic signal provides us with both the key-data for the amplitude correction and the near-surface geometry, as well as with the necessary leads to the otherwise ambivalent electrogram.
As the similarity between the electrogram and the P-seismogram testifies, the coseismic EM-signal is in fact proportional to the seismic particles acceleration. The levelled signal by 10 ms appears solely on the electrogram: this constitutes a further evidence for the converted nature of this EM-signal. The analysis of its amplitude-distribution may unveil some characteristics of the vertical electric dipole, among other things the depth of the interface at which it was created.
Fig. 3: Results from field measurements after stacking and band pass filtering: the horizontal signal provided by the converted EM-wave by 10 ms shows up on the sole electrogram.
The main objective of the laboratory measurements consists in a thorough study of the seismoelectric effects’ dependence on hydraulic parameters under controlled condition. For that purpose we equipped a 50 cm long Plexiglas column with loop-electrodes and seismic receivers. For the source we operate a seismic transmitter with a resonance-frequency by 40 kHz. As for the filling material, we use rounded size-sorted glass beads, with granulometry ranging from 50 µm to 500 µm. With this set-up we hope to generate efficient artificial interfaces with a view to eventually be in a position to study the influence of the pore-geometry on both coseismic and converted EM-signal.
Fig.4: Results from lab measurements after stacking and band pass filtering, using material mimicking middle sand‘s granulometry with particle sizes from 200 to 600 µm, and source frequency of 40 kHz. Water-top ranges 10 cm over the glass beads edge, insofar providing an interface. After filtering is the coseismic signal on both the seismogram and the electrogram to be seen. The origin of the instantaneous signal on the electrogram is still under investigation.
Products & publications
Seismoelectrics Group
Julia Holzhauer
Frank Oppermann
Wolfgang Südekum
Products and Publications
Retrospection to workshop 08./09.02.2011
The workshop contributions
A project of the topical research field ground water systems - hydrogeophysics