Improved marine 4D repeatability using an automated vessel, source and receiver positioning system

Transcription

1 first break volume 29, November 2011 technical article Improved marine 4D repeatability using an automated vessel, source and receiver positioning system G. Brown 1* and J.O. Paulsen 2 Abstract An automated and integrated, vessel, source, and receiver control system has been developed to improve the accuracy and repeatability of 4D surveys. The control system replaces the need for operator intervention with automated updates to all active seismic elements vessel, source, and streamer steering devices. These updates are derived from the real time integration of positioning information from in-sea equipment, environmental information, and the survey plan. Results from field trials and commercial surveys have shown that the new system leads to a step change in the accuracy of source repeatability (typically 2.5 m repeat accuracy for 95% of shotpoints) and also improved repeat receiver positions. Although streamer steering can apply up to 3 of correction against natural feathering, we show that as long as currents have a tidal-driven predictability, ± 9 of natural feather can be accommodated through careful survey planning. Introduction: the need for repeatability The motivation to build the control system for marine 4D surveys came from observations made by Landrø (1999) who showed that overburden heterogeneity had a significant impact on 4D results. He observed a relationship between the similarities of traces recorded in a 3D VSP and their relative source locations. As the receiver location was fixed in the wellbore, only the source position was changing. For a homogeneous earth, after suitable travel-time corrections, it should not matter where the shot is located above the downhole receiver; the recorded traces should be the same. The fact that there were trace differences, and that these differences were a function of the separation of the shot positions, gave a clear indication that overburden heterogeneity limits our ability to repeat 4D seismic data and deliver 4D signals with low background noise levels. Such background noise is called 4D noise. Calvert (2005) discussed this issue in great detail and it is now clearly recognized that repeating the source and receiver positions from the baseline survey is of the utmost importance, comparable to using the same acquisition parameters and the same processing parameters. Goto et al. (2004) demonstrated the effect of 4D noise due to structural leakage when 4D differencing is applied to poorly repeated datasets. In their example, the static geological image that was not cancelled out was sufficient to totally obscure the true 4D signal. The cause of the structural leakage was put down to high variability in shot and receiver positioning. These results clearly illustrated the effect improved repeatability has on the 4D signal and begs the question as to what level of repeatability is required for any given 4D survey. This question is addressed in Figure 1 where we show a cross-plot of trace similarity against source and receiver repeatability for a number of North Sea 4D surveys. In this example, the similarity of traces is expressed using the normalized root mean square (NRMS) attribute described by Kragh and Christie (2002). The scatter of data points suggests a trend similar to that observed by Landrø (1999). Lines connecting some of the data points are for repeated monitor surveys over the same oilfields and suggest that, as positioning technology has advanced over time, more recent surveys have achieved lower NRMS values. As more subtle 4D signals are sought, there is a requirement to reduce the NRMS value to 10% or lower, with a corresponding source positioning error plus receiver positioning error (Δs+Δr) of m, depending on overburden heterogeneity. The higher quality 4D survey data points in the figure indicate around m repeatability is required to produce NRMS levels of <10% for the surveys studied in the North Sea. However, we cannot conclude that this is a general repeatability requirement; we should expect that repeatability requirements will vary from location to location worldwide. System description The automated navigation control system forms an integral part of WesternGeco s high fidelity seismic acquisition 1 WesternGeco, Schlumberger House, Buckingham Gate, Gatwick Airport, RH6 0NZ, UK. 2 WesternGeco, Schlumberger House, P.O.Box 234, N-1372 Asker, Norway. * Corresponding author, brown13@slb.com 2011 EAGE 49

2 technical article first break volume 29, November 2011 Figure 1 The graph indicates a trend of improved repeatability with lower 4D noise levels. Each diamond represents a 4D survey in the North Sea. The average NRMS for each survey is cross-plotted against the average difference between pairs of source and receiver positions for a given offset. The graph suggests that, over the area of investigation, a combined source and receiver repeatability in the range m is required to achieve a final NRMS value <0.1 (10%). (After Smit et al., 2005). Figure 2 Schematic diagram showing inputs and outputs from the new automated 4D positioning controller. Planned positioning information forms one input along with real-time positioning of in-sea equipment and, crucially, information derived from an onboard current meter. Commands are then sent to vessel, source, and streamer controllers to steer for optimum match to the desired locations of sources and receivers. system and was built with the objective of enabling the repetition of source and receiver positions between baseline and monitor 4D surveys. Some existing positioning control hardware components were used, although the integration and automation components were completely new. The system comprises four main sub-systems. n Receiver positioning is achieved through control fins which steer the streamers both laterally and vertically. These control fins were implemented on WesternGeco marine crews for commercial use in 2001, as reported by Curtis et al. (2002). n Source positioning is achieved through a winch system that allows lateral movement of the source array(s). This method allows fine tuning of the source position relative to the required 4D baseline shot position. WesternGeco implemented source steering for commercial use in 4D surveys during n Automation and integration of onboard control systems performs intelligent strategic steering of vessel, sources, and streamers, and other in-sea equipment. n 4D planning and QC tool. In order to cope with the added complexity of 4D surveys, a planning tool forms part of the overall system. The software was designed to enable rapid analysis of existing source/receiver positions. Position information is then fed directly to the control software which automatically steers the vessel, sources, and receivers to the desired locations. The software tool also combines a monitoring and QC function allowing on-line, end of line, and areal QC displays of required versus achieved positioning information. Why automation? It is not possible for a manual positioning operator to continuously and accurately predict the behaviour of a towed source/receiver spread and vessel in a changing current environment. Once the operator is aware that the source/receiver EAGE

3 first break volume 29, November 2011 technical article spread is moving away from the desired track, it already is too late to avoid significant mis-positioning. Subsequent corrections via alterations to the vessel heading are likely to lead to over-correction and an oscillation of vessel and source/ receiver spread about the required track. Conventional control mechanisms take the form of a feedback system, whereas what is really required is a method of predicting, ahead of time, the likely corrections that will be needed when currents change. In the new system this is achieved by constant monitoring of current strength and direction using Doppler current meters. The system is designed to be able to take input from any current meter as long as the current data also contain the spatial position of the meter, i.e., it is possible to have a current meter on a chase vessel that feeds data back to the recording vessel. This information is then used to predict what will happen to the towed source/receiver spread. So far, the system has only been used in a mode where the current meter is placed on the recording vessel itself. Current meter readings are fed directly into the control system software. As a result, the vessel and spread can more easily follow the required track, and streamer steering becomes more effective: it also makes the positioning operator s job much less stressful. Field tests In 2006 WesternGeco performed source repeatability tests in the North Sea by attempting to match a predefined navigation plan. This plan called for matching a dual source configuration to a non-straight line (Figure 3). Matching crooked line source positions is a tougher challenge than straight line shooting. Nevertheless, the test results showed source positioning errors to be less than 2.5m for 95% of the shotpoints. This level of accuracy is not achievable with conventional, manually steered operations, and is an order of magnitude better than the errors seen in some conventionally acquired 4D surveys. In effect, source positioning errors are eliminated. In the Gulf of Mexico, where another test was performed in an area with stronger and more rapidly varying ocean currents, the same excellent source repeatability results were achieved. This test was carried out in parallel with an ongoing wide-azimuth survey. This was a four-vessel operation with each vessel towing a source. In addition, two of the vessels towed streamer spreads. The two streamer vessels were equipped with identical spreads of 10 streamers, each 7 km long, and exposed to the same environmental conditions. One streamer vessel (Figure 4, dark blue circle) had the automated control system installed and the other streamer vessel (Figure 4, red circle) was manually steered. In this way, a true comparison was achieved. The vessel with the navigation control system achieved errors of 4 m relative to the pre-plot for 95% of the observations, while the manually steered vessel achieved errors of 14 m. The results were consistent for the duration of the test, which lasted for a few weeks. Also seen on the graph is that the source positioning Figure 3 Dual source navigation plan used for the North Sea test line. The challenge was to match source positions over a 30 km long line with ± 60 m deviations. The navigation plan lines and achieved positions are plotted on top of each other: only in one area (in bottom ellipse) can deviations from planned positions be identified EAGE 51

4 technical article first break volume 29, November 2011 Figure 4 Left: Acquisition geometry showing source and streamer vessel 1 (blue circle), source vessel 2 (green circle), source vessel 3 (cyan circle), and source and streamer vessel 4 (red circle). Right: Source positioning error graph shows 4 m error or less for 95% of observations with the automated navigation control system, and 14 m or less without the control system. results for the source-only vessels (green and cyan circles) fall between the results of the two streamer vessels. The source-only vessels were not towing a large seismic spread and were more manoeuvrable than the streamer vessels. Even so, the easy-to-control source vessels cannot match the performance of the automatically controlled streamer vessel, which is towing a large spread. Figure 5 shows how the individual system components affect the accuracy of source positioning. In each plot, the red line indicates the source position error relative to the pre-plot for the manually steered vessel (vessel 4, red circle in Figure 4). Similarly, the blue line indicates the source position error for the control system vessel (vessel 1, blue circle in Figure 4), firstly using manual steering, secondly using automated vessel steering, and thirdly using both automated source and vessel steering. An improvement in source positioning accuracy is clearly shown as each component of the integrated control system is added. System performance: receiver repeatability During the same North Sea test mentioned previously, with the dual source configuration, the ability of the system to repeat receiver positions accurately was also tested. An initial test line was acquired without any active streamer steering, i.e., the streamers adopted a natural feather due to the ambient, and variable, current conditions. Results are shown by the dark blue line in Figure 6 (left side). Feather angle in this case is defined as the difference between the pre-plot line bearing and the source to tail-buoy bearing. For the shotpoint range shown in the graph, the feather variation is from -3 to +1, i.e., a change of 4. Once the baseline was established, three different attempts were made to repeat the baseline streamer feather (green, red, and cyan lines). While the red and cyan lines show a good match to the initial (blue) test line, the green line is a very poor match. These results are typical for an area that is dominated by tidal currents; the green line shows that, in this case, the currents were too strong for the streamer steering to be able to compensate. The right graph on Figure 6 shows the estimated streamer feather as dotted lines; these are calculated from the measured ocean current. The solid green and blue lines on the right graph are the same as the left graph, i.e., the streamer feather calculated from the navigation data, while the red and cyan solid lines are omitted for clarity. The estimation of current-derived streamer feather assumes a linear relationship, i.e., streamer feather is estimated as the angle between the measured vessel speed and the cross-line component of the measured ocean current. The estimate can be quite noisy, so the estimate is filtered before plotting. Two conclusions can be drawn from these measured feather angles (solid lines) and current-derived estimates (dashed lines). Firstly, as the red and cyan dashed lines are very different to the blue dashed line, we conclude that a variable degree of streamer-steering feather correction must have been applied in order for the red and cyan solid lines to match the blue solid line. Secondly, the green dashed line shows that, in this case, the currents were too strong for the streamer steering to be able to compensate. During collection of this line, the streamers were steered with maximum available force towards the baseline line for the entire line but, because the streamer-steering devices were only able to compensate for approximately 3 of streamer feather, a good feather match was never obtained. The difference between the green solid and dashed lines was therefore a measure of the maximum steering capability of the system. No attempts were made to match the tidal cycles during collection of these test lines and so, with careful planning EAGE

5 first break volume 29, November 2011 technical article Figure 5 Cross-line source position errors. Test 1 shows both vessels give similar results. Test 2 (blue line) shows slightly improved positioning using automated steering. Test 3 shows dramatic improvement by adding source steering. of line start times in regions where the currents are mainly tidal driven, it should be possible to match feather to legacy surveys, even when total feather variation is well in excess of maximum steering correction. 4D planning: challenges to repeatability In any 4D monitor survey there are many challenges for achieving good repeatability. If a dedicated 4D baseline survey has already been acquired, then the challenges are considerably reduced. If, as is more likely, the monitor survey has to be matched to a legacy exploration 3D survey, then typical challenges could be: n Changes to the number of streamers since the previous baseline or monitor survey (e.g., heritage survey acquired with only two or three streamers). n Changes from dual to single source (or vice versa). n Infrastructure or platform now obstructing the survey area, necessitating under-shooting or dead-head sail lines. n The need to merge lines due to reshoots and infill. It is not the intention here to review all of the above factors and how they may affect repeatability. Instead, we focus on the environmental factors that control repeatability, namely current strength and variability. In Table 1 we define three type levels of repeatability challenge and propose solutions based on acquisition technology and planning strategy. We then proceed to illustrate each repeatability challenge with an acquisition case history. In terms of the acquisition strategy, an ideal situation is for a new baseline to be recorded with straight pre-plot lines and with constant streamer feather of 0, or a chosen fixed feather. A baseline recorded in such a fashion would be the easiest one to repeat. Unfortunately, this is rarely the case. Many 4D surveys are repeats of old 3D surveys which were never designed to be baselines of future 4D surveys. Consequently, these 3D surveys were recorded with a shoot-for-coverage strategy to increase efficiency by minimizing infill. Any survey shot in this fashion will have non-straight source lines, with the result that any 4D monitor survey will have to attempt to match crooked source lines as well as receiver lines with feather. In order to try and understand these various acquisition strategies, we present Table 1 to elucidate the conditions in which we can achieve our desired goal of m overall repeatability. Virtually all 4D towed streamer surveys can be grouped into one of the three survey types given in Table 1. Type I surveys are categorized as having a low level challenge. This is the ideal type of 4D survey referred to above: the baseline source track is straight and the receiver feather 2011 EAGE 53

6 technical article first break volume 29, November 2011 Figure 6 Streamer feather from navigation data (solid lines) and estimated feather from vessel-mounted current meter (broken lines). The dark blue line is the baseline track where the streamers were allowed to follow the natural feather. Typical 4D challenges CHALLENGE LEVEL Baseline SOURCE Baseline RECEIVER FEATHER CURRENT VARIABILITY CURRENT STRENGTH SOLUTION Type I Low Straight Zero Predictable Low Streamer steering Type II Medium Crooked Non-zero Predictable Moderate Streamer steering, source steering, automation & tidal planning Type III High Crooked Non-zero Unpredictable High 2-vessel operation Table 1 Typical 4D repeatability challenges related to current variability and strength and proposed technical solutions. Three levels of challenge are proposed. is controlled (by steering the streamers) to zero (or fixed) feather. In order to achieve this, the current variability must be predictable, e.g., tidal-driven with low overall current strength. An example of this type of survey is shown later in this section. A type II survey, of medium level challenge, is one where, again, the current variability is predictable, but non-zero feather and non-straight source lines in the baseline survey have to be matched. In addition, the current strength is greater than in type I surveys. Whereas a type I survey may be shot with very little infill, a type II survey will require infill or reshoots if the current strength creates feather greater than the steering correction limit. In this case, it is assumed that overlapping streamers and streamer steering are both applied. However, as long as current variability is predictable, quite large feather variations can be tolerated if the survey is planned and shot in tidal slots, i.e., with real time planning to start each line at the correct state of the tidal cycle. An example of a successful survey acquired using this strategy was reported recently (McHugo et al., 2008). A type III survey represents the high level challenge when source lines are non-straight, feather is non-zero and highly variable, and currents are strong and unpredictable. In these situations, where tidal planning is not an option, the amount of feather correction required greatly exceeds that which can be supplied by available streamer steering technology, and alternative shooting strategies are required. One such strategy is dual-vessel acquisition, first proposed by Calvert (2005) and recently reported by Ridsdill-Smith (2008). In this method the source and streamers are deployed on two separate vessels. However, it should be noted that the navigation control system reported in this paper is equally applicable to the two-vessel operation. As we have shown, high source repeatability is still a requirement, and streamer steering and automation will always provide additional control in a dual-vessel operation EAGE

7 first break volume 29, November 2011 technical article Results from commercial surveys In this section we show results from commercial surveys and how the automated control system performs when faced with differing current strength environments. Type I survey - North Sea In 2007, a 4D monitor survey was recorded in the North Sea that fits the type I challenge level described above. Current strength was low and predictable, and the baseline survey had been recorded with straight source lines and with streamer steering programmed to deliver zero feather. Figure 7 shows the survey repeatability results for source position relative to the desired (pre-plot) positions. For this survey, 95% of all shots were within m (cross-line) of their intended location. Similarly, Figure 8 shows planned versus achieved feather where 95% of the survey was shot with a feather mismatch of less than 2.5. Although feather matching represents a convenient onboard QC measure, it is not a direct measure of receiver repeatability. Receiver repeatability depends on the degree of feather mismatch combined with offset. For this reason the QC software has provision to display bin-based source and receiver deviations from planned positions: the concept is illustrated in Figure 9. For a single offset group, usually in the mid to far offset range, data are binned and within each bin the best match pair of baseline and monitor traces is selected. The pair of traces selected has the lowest Δs+Δr value. These values are then plotted for the entire survey, as shown in Figure 10. Here we see that 93% of the bins have a Δs+Δr value of less than 50 m a repeatability that could potentially result in achieving our desired goal of <10% NRMS in the final processed 4D data, based on the graph in Figure 1. Type II survey - North Sea In another 4D monitor survey in the North Sea, also shot in 2007, the same combination of automated vessel, source, and receiver control was used but strong, tidally driven (and hence predictable) currents gave rise to ±9 streamer feather variation. Figure 11 shows how the streamer feather, derived from legacy data, correlates closely to tidal periodicity. This observation leads to the possibility of achieving good repeatability in strong current environments, provided that the start of each line is timed to coincide with the correct phase of the tidal cycle. Through detailed analysis of the legacy positioning data it was determined that, to achieve the required receiver repeatability, the start of each line in the monitor survey should coincide with the same time in the tidal cycle of the legacy sail-lines within an allowed window of ±30 minutes. The repeatability achieved on this occasion for Δs+Δr at the 2500 m offset group was less than 25 m for 94% of the survey. Again, the values are within the desired range based on the graph shown in Figure 1. Two questions arise from these excellent repeatability results. How much of the success is due to planning, i.e., shooting at the optimum time of the tidal cycle? And how much of the success is due to the compensating effect of streamer steering. On completion of the survey, the forces applied to the streamer-steering devices were analysed and correlated with the predicted feather correction. Figure 12 demonstrates that streamer-steering correction was, on average, ~2 for the line in question. Analysis of other lines showed similar results. Figure 6 showed that using the steering devices deployed at the time, with forces ranging up to maximum, gave 2-3 of feather correction. Analysing streamer lateral forces allows us an alternative method to estimate average feather correction throughout the survey duration. Hence, the contribution of streamer steering to meet repeatability requirements can be quantified. Type III survey North West Shelf, Australia The high level challenge we describe occurs when the current strength is both high and unpredictable. Many areas other than the North Sea experience these conditions, e.g., the Gulf of Mexico and offshore West Africa, where strong Figure 7 Onboard QC plots of source position relative to pre-plot (desired) location. The map shows that most of the larger errors are associated with an obstructed area of the survey EAGE 55

8 technical article first break volume 29, November 2011 Figure 8 Streamer feather relative to pre-plot. Displays in this figure follow the same scheme as Figure 7 except streamer feather is plotted instead of source position. current eddies cause large and rapid changes to streamer feathering. One way to deal with large feather variations is to decouple the source from the recording vessel, i.e., implement a dual vessel operation as proposed by Calvert (2005). The advantage of this method is that the source vessel can repeat the shot locations while the streamer vessel can steer to place the streamers in the optimum location in order to repeat a desired offset. This principle is illustrated in Figure 13. For logistical and safety reasons, the source vessel has to be positioned at the rear of the spread. In doing so, traces are recorded in the opposite azimuth if the line was shot in the same direction as the baseline survey, and it is well known that this gives rise to poor repeatability. The solution to this issue is to shoot the lines in the opposite direction to that of the baseline survey, with the source vessel at the rear of the spread. This method of recording has become known as the reverse push technique. This method is not without some disadvantages, e.g., the loss of near offsets. In a survey recorded in 2008, the automated vessel, source, and receiver control system was used to perform a reverse push 4D monitor survey over the North West Shelf area, offshore Australia. As a previous 4D monitor survey had been recorded with conventional towed streamer technology, it was possible to compare repeatability of the two monitor surveys (conventional/reverse push) against the baseline. While the receiver repeatability in this example is Figure 9 For a chosen offset, each CMP bin is analysed for repeatability errors by searching for the nearest trace match to the baseline dataset. The black source and receiver represents the trace from the baseline survey while the blue source and receiver represents the best matching trace that would be used in 4D binning. Calculation of the values (Ds + Dr) for all bins provides a true metric for the assessment of repeatability. not as good as the results from the two North Sea surveys, there is a significant improvement in using the reverse push monitor survey with automated steering control compared to the conventional monitor survey. In this instance, cross-line target receiver repeatability was used as a QC measure. When using conventional technology, 90% of receivers were within 175 m of the plan. When the automated control system was used 90% of receivers were within 105 m of plan. The superior performance of the automated steering system can be attributed to the look ahead capability of the automated system, which allows the vessel to steer and compensate for current conditions in anticipation of streamer feather fluctuations. Final results 4D noise So far, we have only shown results of source and receiver repeatability without reference to the final 4D noise NRMS values achieved after processing. Final NRMS noise values are not available for all surveys illustrated. However, in another 2007 North Sea 4D survey, 95% of the Δs+Δr values were less than 50 m, whilst 67% of the mapped area had NRMS values <11%. Although it is not known where this survey area of the North Sea sits in terms of the complexity of the overburden, we can say that the Δs+Δr and NRMS values are close to the stated goal, as shown in Figure 1. Conclusions Repeatability is crucial for successful 4D projects, especially as we seek to resolve more subtle 4D signals where 4D noise is a critical factor. In this paper, we have reported on the design, implementation, and results from a new automated, integrated, navigation control system developed to improve 4D repeatability in single and multi-vessel towed streamer 4D operations. We have demonstrated that source repeatability of 2.5 m is now routinely achievable for 95% of each survey, a step change for marine 4D operations. In effect, source repeatability errors have been eliminated. Furthermore, EAGE

9 first break volume 29, November 2011 technical article Figure 10 Areal plot of CMP bins for type I survey showing Ds + Dr attribute values in metres for traces in the group at 5000 m offset. Figure 11 Correlation of tidal cycle with streamer feather recorded in legacy dataset. such repeatability should be obtainable almost regardless of current conditions. Receiver repeatability is also improved but the ability to reach our target of m overall repeatability in Δs+Δr is constrained to areas where currents are limited in strength and fluctuations are predictable. In these situations we have shown that feather variation of approximately ± 9 can be accommodated with suitable planning effort. In areas of strong and unpredictable currents, two-vessel operations are a means of attaining the necessary repeatability. In this situation the control system described here still offers much improved control and stability to streamer operations. Acknowledgements We thank Statoil for permission to show results from the Kristin and Svale 2007 surveys, and the Blake Partners, BG Group, Talisman Energy UK, and Idemitsu Petroleum UK for permission to publish. We also thank the development team at WesternGeco s Oslo Technology Centre, the seismic crews on Western Regent and Western Spirit, and Andrew Thompson, Franck le Diagon, Johannes Hvidsten, and Patrick Smith 2011 EAGE 57

10 technical article first break volume 29, November 2011 Figure 12 Left: planned feather (blue), actual feather (red), and predicted feather as estimated from the current meter (green). Right: forces applied to each streamer control device throughout the duration of one particular line. For most of the line, the blue colours indicate that maximum forces were applied to port. This can be correlated with the separation of the green and red curves. Figure 13 The reverse push technique. In conventional single-vessel operations only the near offsets are repeated in the presence of severe feather difference between the baseline and monitor recordings (top). Using two vessels allows the streamer vessel to steer so that, for example, the mid offsets can be repeated (middle). In practice, the monitor line is recorded in the opposite direction with the source vessel at the rear of the spread (bottom). for useful discussions and for help in preparing some of the results shown in this paper. References Calvert, R. [2005] Insights and Methods for 4D Reservoir Monitoring and Characterization. SEG, Tulsa and EAGE, Houten. Curtis, T., Smith, P., Combee, L. and Olafsen, W. [2002] Acquisition of highly repeatable seismic data using active streamer steering. 72 nd SEG Annual Meeting, Expanded Abstracts, 21, Goto, R., Lowden, D., Smith, P. and Paulsen, J.O. [2004] Steered streamer 4D case study over the Norne field. 74 th SEG Annual Meeting, Expanded Abstracts, 23, Kragh, E. and Christie, P. [2002] Seismic repeatability, normalized rms and predictability. The Leading Edge, 21, Landrø, M. [1999] Repeatability issues of 3-D VSP data. Geophysics, 64, McHugo, S., Ramsden, P., Woods, N., Jones, C., Millington, M. and Wilson, A. [2008] Improved 4D seismic repeatability from dynamic spread control a case history from the Blake field, UKCS. PETEX 2008 Conference, London. Ridsdill-Smith, T. [2007] Benefits of two-boat 4D acquisition, an Australian case history. The Leading Edge, 27, Smit, F., Brain, J. and Watt, K. [2005] Repeatability monitoring during marine 4D streamer acquisition. 67 th EAGE Conference & Exhibition, Extended Abstracts, C015. Received 14 March 2011; accepted 4 September doi: / EAGE