Introduction

1.1         Marine CSEM

Electromagnetic (EM) methods for hydrocarbon exploration can be divided into active and passive categories. The use of passive sources such as MT has shifted to active EM methods such as the controlled source electromagnetic (CSEM) method. MCSEM has been used to detect and appraise thin resistive hydrocarbon reservoirs in a conductive surroundings. Controlled source electromagnetic method is also known as, CSEMi, sea bed logging or SBL. CSEM detects electrical resistivity contrasts (Kong, 2002). The method was developed for deepwater studies of oceanic lithosphere in the only 20^th century (Cox, 1981) and the first commercial MCSEM surveys were completed in 1979 (Spiess et al., 1980). The original research combined both natural and active EM sources to image resistivity variations beneath the sea floor. Prior to MCSEM, MT identified resistivity variations on large scale structures at basin level. MCSEM now ascertains the shape and resistivity of thin resistive bodies (Peace, 2005). These early surveys concentrated on shallow water exploration targets (Chave et al., 1981) however the method has expanded onto both locating both deep and shallow water targets. In the last couple of decades the MCSEM method has rapidly expanded through the efforts of various university and industry researchers such as Scripps Institute of Oceanography, The University of Toronto, Cambridge, Southampton University, Schlumberger PGS and Fugro Electro Magnetic.

The traditional MCSEM method is performed using sea-bottom electromagnetic receivers and a large electrical bipole. The

[youtube_sc url=http://www.youtube.com/v/VBdzzYzJePM?fs=1&amp]

MCSEM detects resistivity rather than acoustic properties. This is beneficial because hydrocarbon saturation can be characterised more accurately than by using seismic attributes. The response in MCSEM correlates with the saturation percentage, which is unlike seismic which can only detect accurately between 0 and 10% with amplitude and velocity analysis techniques (MacGregor, 2006); this can be seen in Figure 1‑1. Marine CSEM can detect petroleum, natural gas, gas hydrates and other resistive zones provided a sufficient resistivity contrast exists. It is also able to predict the hydrocarbon saturation for a potential reservoir ahead of drilling (Phillips, 2007). Marine CSEM surveys are becoming more popular in deep water due to the increasing expense of seismic surveys and the cost of drilling prospects (Spiess et al.,1980). The method must be applied under suitable geoelectrical conditions. Commercial projects are always preceded by a feasibility study evaluating the electromagnetic responses of the expected target and background geoelectric structure. A survey plan can be designed from the results of the feasibility study.

The controlled source electromagnetic method is used when seismic method is restricted by low velocity zones such as magma chambers or high velocity layers in sub-basalt or sub-salt areas. MCSEM is not perfect, its use is restricted when applied beneath thick tabular salt/basalt targets or where the target is thin, deep and has low resistivity comparable to the surrounding geology (see Table 1‑1). MCSEM is useful in deep water turbidites and deltas, over stacked reservoir sequences or where there are large resistivity contrasts such as in shallow gas hydrates (MacGregor, 2006). Geological environments such as reservoirs at the edges of salt diapirs and in carbonates can be explored by the MCSEM method.

Marine controlled source electromagnetic surveys have become an integral part of deep ocean petroleum exploration and appraisal (Gribenko and Zhdanov, 2005) and future use for monitoring purposes may be possible. Some are now performed in 3D rather than 2D. For 3D MCSEM, planning has become increasingly important for the success of a survey (Kong et al., 2005).

Table 1‑1: The practicality of the MCSEM method in different geological environments. MCSEM works in a variety of geological environments however in some geological settings MCSEM may not work. (MacGregor, 2006).

Figure 1‑1: The effect of Hydrocarbon saturation on seismic and MCSEM electric field observations. MCSEM electric field response characterise the hydrocarbon saturation more precisely because the increase in signal corresponds to the hydrocarbon saturation, while the seismic method fails to distinguish between 10 and 100 percent (Modified from MacGregor, 2006)

1.3         The controlled source electromagnetic method

There is a typically a resistivity contrast between the conductive host rock and the resistive hydrocarbon. Saturated mudstone rocks, sandstones and shales with low resistivity dominate deep water environments. A hydrocarbon reservoir can be 10 to 100 times greater in resistivity (Eidesmo et al., 2002). This physical property can be exploited by the marine CSEM method by recording the disturbances from an active electromagnetic source.

Typical marine CSEM surveys work by using a horizontal electric dipole to transmit a low frequency 0.1 to 5Hz square wave under high power up to 1000 Amps (MacGregor, 2006). A horizontal electric bipole is a 100-1000m long electric field transmitter towed 50m above of the ocean bottom (MacGregor, 2006). Receivers can be placed to record the perturbations in the electric and magnetic fields for all Cartesian directions. In modern surveys E_x, E_y, H_x and H_y components are typically recorded.

The transmitted wave diffuses through the through the water column and into seabed (as seen in Figure 1‑2). Electric fields attenuate less in resistive mediums. The increase in electrical field due to the reservoir can measured at the seafloor at offsets roughly double the depth of the reservoir below the seabed. The magnetic field circulates around the electric field and is responsive to secondary fields.

Figure 1‑2: A schematic of a MCSEM survey showing the path of the transmitted electric field. The electric field will channel above the resistive boundaries such as at air and hydrocarbon interfaces.

2                   Background

This chapter will overview the use of the marine controlled source electromagnetic method for hydrocarbon exploration.

1. The current state of the marine controlled source electromagnetic technique for hydrocarbon exploration.
2. Physics of low frequency electromagnetic
3. Instrumentation
4. Survey parameters
5. Factors affecting survey design
6. Possible survey designs

2.1         The current state of the marine CSEM method

2.1.1        Survey planning methodologies

Few papers have stated clear survey planning methodologies, the papers written by Peace (2005) and Flosadóttir et al. (1996) are overviews of the main areas of survey planning and MacGregor (2002) and MacGregor et al. (2006) only skim the topic surface. Currently there is no publicly available material that provides an in depth planning methodology, specifically in the area of 3D multispectral survey design and only Ridyard D. Et al. have provided insight into evaluating MCSEM survey designs.

2.1.2        CSEM component representation

It appears that the conventional representations of MCSEM data include 1D profiles, grids and in more extraordinary cases 3D contours, such as Peace (2005), Mittet et al. (2004), Phillips (2007). The use of 3D visualisation techniques is limited. Polarisation ellipses have been mentioned by MacGregor (2006) but only on a 2D

plane. Streamlines, isosurfaces, polarisation ellipses and polydata are scarcely applied to MCSEM data.

2.2         Physics of low frequency electromagnetics

2.2.1        Skin depth

The depth of penetration as a relationship of frequency and conductivity is described by the skin depth. From looking at previous MCSEM surveys MacGregor et al. (2001), MacGregor et al. (1998) and Ellingsgrud et al. (2002).

Skin depth is the effective depth of penetration of electromagnetic energy in a conducting medium when the amplitude of a plane wave has been attenuated to ¹/_e (or 37 percent) (Sherriff, 1991). This relationship is described by the skin depth in equation 1.(1)

The formula can only be used in an isotropic conductive environment. The skin depth equation can only provide an indication of decay lengths for more complex geoelectric models.

Sea water is generally considered highly conductive and acts as a low pass filter to incident MT source fields (Constable et al., 1998). Frequency also affects the depth of penetration because energy is lost for every cycle of the wave, causing higher frequencies to be attenuated more as seen in Figure 2‑1. The electromagnetic skin depth is shortest at high frequency.

Figure 2‑1: The relationship of skin depth for frequency and conductivity. The skin depth decreases as the conductivity or frequency increases.

2.3         Survey parameters

2.3.1        Transmitted waveform

The waveform dictates the frequency content. The frequency content influences the depth of investigation, resolution and sensitivity to the hydrocarbon body. Another attribute to consider includes the waveform stability. The waveform must be stable to ensure consistent signal phase measurements. To ensure this in the field, MCSEM waveform monitoring is used (MacGregor, 2006).

The MCSEM method traditionally uses a continuous harmonic square wave signal with fundamental frequency (and odd harmonics) typically within the range of 0.01-10Hz (Pound, 2007). The Square and Cox are two commonly used waveforms as seen in Figure 2‑2. They are used due to phase stability and the wide frequency content (MacGregor, 2006). Unfortunately there is a trade-off between resolution and depth of penetration. In most cases high frequencies resolve near surface resistivity variations, which can be used to constrain the geoelectric model for both the inversion and final interpretation. The waveform should contain multiple frequencies so that both near and far surface resistivity variations can be resolved whilst also maximising the depth that can be detected.

Figure 2‑2: Typical transmitted MCSEM waveforms and the associated frequency spectra. The frequency spectrums are usually dominated by the fundamental frequency and its harmonics. Waveforms with both low and high frequencies allow both near surface and target geoelectric properties to be resolved (Modified from MacGregor, 2006).

2.4         Factors affecting survey design

2.4.1         CSEM noise sources

A number of noise sources that must be considered. Sources include sea water currents, magnetotelluric activity, bipole vibration, internal electrode and amplifier noise and cultural effects, as seen in Table 2‑1 and Figure 2‑3.

Internal amplifier and electrode noise limits the instrument’s high frequency noise floor. Vibrations of the dipole receiver arms can also induce an unwanted signal. A vibration of less than 1mm for a 10m dipole can produce an induced voltage comparable to the target signal. These movements can be limited by using bottom weights, weighting each electric receiver dipole arm with glass rods and by using farings on dipole arms (MacGregor, 2006). These issues are not as influential in deeper waters so the receiver dipole arm length can be increased.

At lower frequencies (<1Hz) spheric noise and ocean induced fields can contaminate the received data. The magnitude of spherics noise reaching the receivers varies with water depth. Increases in water depth improve noise recording conditions of surveys because seawater acts as a shield from external magnetotelluric signals (Kong et al., 2002). The motion of a conductor in an electric field will have a voltage induced in accordance with Faraday’s law. Seafloor water currents can dominate the intermediate frequencies in a survey. Therefore care must be taken when planning surveys in areas with fast moving water currents.

Table 2‑1: Noise sources to consider.

Figure 2‑3: Noise sources. The noises can be split into two areas, external and internal. Internal sources are within the instruments which are internal amplifier and electrode noise, whilst external noises are caused by seafloor water currents, magnetotelluric (MT) signals, dipole vibration and cultural noises. It is best to conduct a survey in deep water to reduce the effect of seafloor water currents and external atmospheric noise.

2.4.2        Target style considerations

The target style greatly influences survey planning. Considerations include water depth, target depth, geological type of target, reservoir style, resistivity contrast, resistive non-hydrocarbon geological targets, reservoir complexity and the purpose of survey (Peace, 2005).

Water depth and bathymetry influences the onset of the airwave. Water bottom channels, canyons and sloping or having the target sit above the water bottom when surveying off a shelf edge, all need to be considered. Secondly the target depth from the water bottom and aerial size of the target compared to the depth affects the detectability of the target. Targets smaller in aerial extents than their depth are harder to detect. Thirdly the geological type and shape of target influences the response receiver on the ocean floor. Resistivity contrast also influences a survey plan. If the resistivity contrast between the hydrocarbon and the host is insufficient, no survey could detect the hydrocarbon. Non-hydrocarbon resistive structures could also lead to false positives. Therefore it is necessary to model a range of geological settings. Lastly the purpose of the survey influences the final survey design. For example recognisance surveys utilise widely spaced receivers and transmitter lines to target hydrocarbons. 3D surveys use many closely spaced receivers and densely positioned transmitter line locations at multiple azimuths to characterise or appraise a known field (See Figure 2‑4).

Figure 2‑4: Geological considerations for MCSEM survey planning

2.4.3        The airwave problem

The airwave is a complex interference effect between the air and the water and seabed layering. The airwave poses problems for MCSEM surveys because it tends to dominate received signals at far offsets; masking seabed response (Eidesmo et al., 2002). It can be identified by a gradient break in inline electric field profiles, there is also a total phase lag which dependant on the offset and water depth (Eidesmo et al., 2002). The airwave is affected by the transmitter-receiver offset, transmitter dipole orientation, transmission frequency, resistivity structure and the water depth.

Eidesmo et al., (2002) and MacGregor (2006) have offered solutions to the airwave problem but none offer a ‘silver bullet’ solution. It is an ongoing problem for industry, EMGS and OHM that are working to resolve it by using various data processing techniques and also by limiting its effect by using novel acquisition practices (Eidesmo and Ellingsrud, 2002). Possible solutions to the airwave problem include selecting acquisition parameters that limit the generation of an airwave, signal processing and even using information in the airwave (MacGregor, 2006; Eidesmo and Ellingsrud, 2002l Eidesmo et al., 2002). Forward modelling can provide information regarding the effect of the airwave prior to surveying (Mittet et al., 2004; Johansen et al., 2005). Forward modelling calculates the onset of the airwave response. Modelling of these transmitter-receiver geometries will indicate the influence of the airwave.

Constable and Weiss (2006) have suggested that by using a vertical electric dipole transmitter the airwave can be limited. The coupling of the air-water interface is influenced by the orientation of the transmitter. Horizontal current loops create PM modes which strongly couple with this interface, while vertical current loops create TM modes and have no strong airwave effect (MacGregor, 2006).

The choice of transmission frequency is important because it affects the onset and amplitude of the airwave. Not only do higher transmission frequencies amplify the airwave effect, but the onset of the wave arrives sooner as seen in Table 2‑2. Therefore the benefits of low frequency must be balanced against reduced resolution (MacGregor, 2006).

The maximum depth of investigation before contamination of the airwave can be calculated by using localisation (Eidesmo et al., 2002). The depth at which this occurs can be calculated by multiplying the scale factor from Equation 3 (alpha) with the water depth (Tompkins et al., 2004). For example if alpha=1.76 and the water depth is 1000m, then the maximum depth of investigation would be 1760m.

(3)

Table 2‑2: The distance from the source (in km) at which the airwave starts to dominate the overall response. The point at which the response is dominated by the airwave is represented as function of water depth and the signal frequency. (Reproduced from Eidesmo et al., 2002)

Figure 2‑5: The magnetic and electric field patterns from vertical electric (VED) and magnetic (VMD) dipoles. Horizontal current loops strongly couple with the air water interface resulting in a large air wave response. The air wave phenomenon can be minimised by using a vertical electrc dipole.

2.4.4        Seawater conductivity

Seawater constitutes a large portion of the geoelectric model. Determining the correct seawater conductivity is important for both forward modelling and inversion. The conductivity of the fluid varies in the water column and also in the seabed structure itself. Auxiliary instruments record seawater conductivity over the duration of most MCSEM surveys (an example can be seen in Figure 2‑6). Sea water varies in resistivity due to temperature, salinity and pressure as described in equation 4.

(4)

Sea water conductivity is around  at typical ocean floor temperatures, it reaches a minimum conductivity of . This large variation in conductivity has a significant effect on the bulk resistivity of the formation. Archie’s 1942 law can be used to find the saturated formation’s true resistivity in  (see equation 5).

(5)

Figure 2‑6: Typical seawater conductivity measurement over the duration of a typical MCSEM survey. The conductivity varies over time and position and the recorded data should be incorporated into the geoelectric model for forward modelling or inversion (Reproduced from MacGregor, 2006).

2.4.5        Hydrocarbon saturation and formation resistivity

When dry, reservoir formations extremely resistive. Interstitial saline pore fluid allows the electrical current to flow. Archie’s law relates the electrical conductivity of sedimentary rocks with its porosity and brine saturation. Hydrocarbons increase the formation resistivity when it displaces brine from a formation. Hydrocarbon saturation increases the overall formation resistivity. MCSEM measurements can detect the increase in hydrocarbon saturation.

2.4.6        Bathymetry

The marine CSEM method is affected by variations in seafloor bathymetry. The HED transmitter should be towed at a constant height above the receivers, following the shape of the bathymetry. Bathymetry should be taken into account because a seafloor rise can produce the same response as a reservoir and only quantitative interpretation such as inversion may be able to establish the true geoelectric properties (Mehta et al., 2006).

2.5          Multispectral instrumentation

Marine CSEM instrumentation has developed rapidly with ocean bottom receivers reaching higher levels of sensitivity while horizontal electric bipole sources are becoming increasingly powerful. The choice of instrumentation is a significant step in planning because it affects the noise-floor of the survey.

2.5.1        Multi component receivers

Technology during the late 1990s underwent major advancements notably by organisations such as Scripps Institution of Oceanography (Constable et al., 1998). The MCSEM method uses receivers modelled on marine magnetotelluric instrumentation from this period. The receiver records magnetic and electric field time series data. The receivers have a number of recording devices, including a digital magnetic compass/tilt-meter which records the orientation, a timing system which is synchronised pre and post deployment and magnetic and electric sensors recording multiple axial directions. Horizontal electric field receivers are composed of silver-silver chloride electrodes which are between 1 and 10m in length (MacGregor et al., 2006). A highly sensitive vertical arm can be used to measure the vertical field. The magnetic fields are measured by highly sensitive, light weight induction coil magnetometers (Keys, 2003). The typical components which are detected by commercial designs are x and y magnetic and electric field directions. The vertical electric field could also be used for commercial uses. Currently this orientation is not recorded because the vertical electric field dipoles are still too noisy to be used practically.

There are numerous receiver parameters to be considered when choosing the most suitable receivers for the survey (as seen in Table 2‑3). Key attributes to consider prior to surveying include the noise floor of the instrument, timing calibration, timing stability, battery type and energy use, recording capacity, response calibration, navigation and seafloor orientation (MacGregor, 2006). The dynamic range of the instrument is important because there is a large variation in the signal strength. Therefore 24-bit analogue to digital converters with pre-amplifiers enable high resolution data to be recorded at all source-receiver offsets without signal saturation.

The noise level of a receiver affects receiver selection during planning. The main source of electric field receiver noise above 1Hz is caused by amplifier and electrode induced noise rather than ambient noise (Hoversten et al., 2006). There is an inverse relationship between the noise level of electric field receivers and the frequency. For example OHM’s EFMALS III, has a noise floor of 1nV/m /  (MacGregor, 2002). In effect, the receiver can detect and resolve smaller amplitudes for higher frequencies. WesternGeco (2008) use receivers that have the capability to resolve values as low as  at 1Hz. Deep seafloor environments are typically electrically quiet. Most MCSEM receivers have a noise floor electromagnetically aroundfor electric fields and 4×10^-12T at 1Hz for magnetic fields.

To improve the signal to noise ratio of the receiver, the antenna length can be extended. This voltage is proportional to the receiver dipole length (Flosadóttir, Á et al., 1996). Synchronous stacking can be used to also recover a repetitive signal from the random ambient or instrument noise.

Technology is improving and the limits and features will drastically over next few decades. Hence the features seen in Table 2‑3 should only be taken as a guide. In summary the main features to consider when planning a survey are the noise threshold of both the E-Field and B-Field sensors and which electric and magnetic axial directions the device will record.

Table 2‑3: Typical receiver features to consider when planning a MCSEM survey. These values have been taken from two commercial contractors, OHM’s EFMALS III and WesternGeco receivers. The main attributes to consider when choosing the correct receiver when planning a survey are the electric and magnetic field receivers, their associated instrument noise and axial direction. (Reproduced from WesternGeco, 2008 and OHM Surveys, 2008)

Figure 2‑7: Diagram of a typical marine CSEM multicomponent receiver. The receiver is the Scripps Institute of Oceanography Mark III design only records inline magnetic and electric fields and broadside electric field (Constable et al., 1998; reproduced from Keys, 2003)

[nggallery id=4]

Gallery 2-1: Pictures I took from the Scripps boat when it was in Perth, 2009. You can see Kerry Key demonstrating how the navigation beacon float works. A few notes, the vertical Z electrode is made of kevlar so it is not conductive whilst retaining enough stiffness so ocean current's have limited effect in the electrode movement. The concrete base enables it to be sunken to the bottom, whilst once the signal is given, the charge is intiated to break the wire holding the receiver to the concrete pad, the yellow float enables it to float to the surface.

2.5.2        Transmitter

The marine controlled source electromagnetic method uses a horizontal electric dipole as a transmitter. The transmitter consists of two electrodes separated between 100-300m. The Transmitter is towed 25-50m above the seabed at a speed of 1.5 to 2.0 knots (WesternGeco, 2008). The transmitter uses a powerful source generates around 1000A. There are numerous commercially available transmitters that can be selected prior to surveying.

Important attributes to consider when selecting a good transmitter include the peak output dipole moment and the stability of the output waveform. Waveform polarity transitions have a temperature-dependent latency (1-2µs) that decreases as the transmitter temperature rises and stabilises (WesternGeco, 2008). The output waveform’s phase must be stable to around 1 part in 10⁸ (1ms/day) otherwise the received waveform will be artificially out of phase with the transmitted waveform. Phase is controlled by GPS and consistently monitored during the survey.

The dipole moment is the length of the electrodes multiplied by the transmitted current. If the dipole moment is doubled the amplitude recorded at the receiver are also doubled.The source is deep-towed to optimise signal coupling with the sea floor and to reduce the conductive losses. Heights of around 25-50m are chosen but can be varied in accordance to bathymetry, ocean conditions and the coupling required. The altitude of the transmitter is constantly monitored by using an altimeter on the source. The tow speed of the transmitter should be considered prior to surveying as it affects the quality of the received data. A receiver records 50 transmitter cycles every 100m for a given frequency of 0.5Hz and a towing speed of 1m/s. Stacking reduces the noise by a 1/sqrt(N) times and improves with slower towing speeds (Hoversten et al., 2006).

Figure 2‑8: Typical schematic of a marine CSEM transmitter and HED (Reproduced from WesternGeco, 2008).

2.6         The use of forward modelling for survey planning

Forward modelling of a prospects geoelectric response prior to surveying fulfils several purposes.

• To optimise transmission parameters to reduce noise and maximise resolution
• Design transmitter-receiver geometries to enhance the response of the target reservoir

2.6.1        Optimisation of transmitter receiver geometry

Transmitter receiver geometry affects the response detected from the prospective hydrocarbon. For example, inline electric field receivers detect a larger electric field component whilst the magnetic field mainly influences the broadside geometry. There is a link between transmitter-receiver offsets and depth of investigation. Large transmitter-receiver offsets are sensitive to deep targets, while short offsets are responsive near surface resistivity variations.

Inductive effects dominate the broadside geometry while galvanic effects are much stronger in the inline receiver geometry. Errors due to azimuthal scattering (associated with surface heterogeneities) reduces with the increase in the number of receivers in accordance with  1/sqrt(n), where n is the number of data with independent transmitter and receiver locations.

2.6.2        Survey Design Options

The survey design includes both receiver and transmitter positions. This geometry determines how well the target can be detected and characterised. There are a number of options available to MCSEM practitioners ranging from simple 1D lines to a multi-azimuthal survey which ‘illuminates’ the target from all directions as seen in Figure 2‑9. The type of survey design depends on the objective of the survey. (McGregor, 2006). For example reconnaissance surveys identify resistivity variations on a regional scale, detection surveys need to identify commercial resistive targets and appraisals require a more detail constraint of an identified prospect. The purpose of the survey determines the receiver and transmitter line spacing.

Figure 2‑9:  MCSEM survey layouts of transmitter and receiver lines. The purpose of the survey determines the survey layout, including the transmitter and receiver positions. Each layout has its advantages, denser transmitter-receiver grids generally have higher relative responses, however the operations are more expensive to conduct.