This study presents the redesign and implementation of a towed-streamer geometry for a 6700 km² shallow-water 3D seismic survey offshore Tierra del Fuego, Argentina. To balance near-offset and spatial-sampling requirements with operational efficiency under severe weather and currents and within a short seismic season, a wide-tow triple source configuration was planned, but operational challenges required the source separation to be reduced from 100 m to 60 m. Using 90 m streamer separation resulted in a 12-streamer spread that could be placed on 480 m or 570 m sail-line separations, corresponding to 10- and 12-streamer configurations. This reproduced the short minimum offsets of the base configuration while the wider template compensated for the reduced efficiency of the narrower template. Early acquired lines, using the 12-streamer template, showed that the wider sail-line separation was difficult to maintain under the encountered circumstances. Delays during acquisition of the first part of the survey resulted in postponement of the second part to the following season. Upon return the 12-streamer spread was retained and placed on the 10-streamer template to reduce infill through streamer overlap. This resulted in the delivery of fit-for-purpose data within budget, demonstrating the operational value of a flexible acquisition geometry.
Introduction
In early 2024 Equinor acquired a 6700 km² towed-streamer survey off the southern tip of Argentina, near Tierra del Fuego. Located in a very remote region, this was the southernmost commercial 3D marine seismic survey in the world. The survey area lay in relatively shallow water, less than 30 km offshore from Río Grande, and covered the nearshore Austral 105 and 106 blocks as well as the adjacent Malvinas Oeste 121 block (Figure 1). Situated in the ‘Furious Fifties’, a region known for strong winds that build over vast distances, the area presented numerous challenges. Rough weather, strong tidal currents, and high exposure to marine life were compounded by the relatively shallow water depths.
With little previous seismic activity in the area and limited weather information available in this sparsely populated part of the world, it was difficult to estimate the length of the seismic season. But climatic models and other resources indicated a sufficient window of suitable conditions between October and May, albeit with a high likelihood of extended weather downtime. Compliance with local regulations, aiming to minimise impact on protected fauna, was expected to further limit the available time for data acquisition. Thus, a survey of this size located in this challenging environment necessitated an acquisition geometry that provided high efficiency and robust resilience against adverse conditions.
At the same time, the shallow water and shallow target depths required acquisition of data with offsets shorter than those usually associated with towed-streamer seismic, where sources are commonly placed between the innermost streamers and combined with wide receiver spreads to improve efficiency. This dictated a towing configuration that minimised the distance between the outermost source and outermost streamer, thereby constraining the spread width and reducing operational efficiency. The remoteness of the area introduced further uncertainty in the available bathymetric data, particularly near the shoreline.
Industry experience has demonstrated with many examples that these challenges can be addressed using wide-tow sources. However, after encountering operational challenges driven by adverse conditions in the survey area, a redesign was forced that maintained the required data quality while enabling acquisition of the full project scope. We present a solution that utilises a single acquisition geometry with two different sail-line separations, adapted to water depth and geological requirements and optimising utilisation of available equipment.
Challenging conditions and little time
The survey area was located close to the coast, where expected water depths ranged from 50 m to 100 m. Pre-survey modelling of the water bottom identified a critical offset of only 250 m (Figure 2). This indicated a high risk of a pronounced acquisition footprint and emphasised the need for offsets shorter than 250 m to allow for imaging of the water bottom and to minimise the expected footprint. Geological features such as iceberg scouring and other geological information suggested a relatively hard water bottom, which could generate strong multiples that contaminate the target response, requiring particular attention during processing. Together, these factors reinforced the need for high spatial resolution and therefore a dense acquisition geometry, with the associated reduction in survey efficiency.
Long (2017) and Widmaier et al. (2020) have presented possible solutions for these challenges by introducing wide-tow triple sources. Moving the sources towards the outer streamers shortens the distance between outermost source and streamer and allows us to employ wider spreads without jeopardising near offsets, thus improving the subsurface coverage provided by each sail-line. Souissi et al. (2023) presented a survey design study for this acquisition that followed this approach but went further by introducing the concept of a limited number of empty CMP (common midpoints) lines to improve efficiency. The suggested solution (referred to as base configuration below) recommended using triple sources towed at a separation of 100 m. Combined with a streamer separation of 75 m and a total of 12 streamers, this configuration provided near offsets to the outer streamers of 328 m (Figure 3) with a bin size of 6.25 m x 12.5 m. Although this configuration provided a relatively wide natural sail-line separation of 450 m, additional efficiency could be gained by removing overlap between neighbouring sail-lines, thus extending the separation to 525 m at the expense of leaving some CMP lines empty. It was assumed that these could be filled in processing using regularisation and interpolation. Natural variations in streamer location were expected to further mitigate any coverage gaps, allowing us to maintain high efficiency.
However, upon implementation of this configuration in the field, adverse conditions in the survey area caused increased stress on the towed equipment, increasing the risk for technical down time and potential exposure of offshore personnel. It was therefore decided to reduce the source separation to a smaller distance, easing the strain on the towed gear. Infield testing indicated that the sources could be operated safely at a separation of 60 m.
Reducing the source separation moved the sources towards the centre of the spread, increasing near offsets. Nevertheless, this adjustment was necessary because the technical downtime associated with the original configuration was expected to outweigh the efficiency gains from wider source towing, while avoiding unnecessary exposure of offshore personnel remained paramount. The original survey geometry therefore had to be redesigned to achieve the best possible balance between the required data quality and completion of the dataset within the short seismic season.
Solving challenges with flexibility
Considering a source separation of 60 m was operationally feasible, a new configuration was designed that maintained good spatial sampling and short near offsets while maintaining good survey efficiency. The proposed new configuration combined a milder source wide-tow of 60 m with an increased streamer separation of 90 m, resulting in a natural crossline bin size of 6.25 m x 15 m, an increase of the crossline bin size by 2.5 m. Composing the spread of 10 streamers resulted in a maximum near offset of 358 m, close to that provided by the base configuration. Following the philosophy of the original survey design, sail-line separation was increased from the natural 450 m to 480 m, thus introducing two empty CMP lines between sail-lines but reducing the efficiency loss caused by a smaller sail-line separation to 9% (see figure 4).
While this configuration provided a suitable solution, the reduced efficiency made the acquisition vulnerable to downtime as any delay would risk terminating the acquisition before the entire survey was acquired. In addition, a 10-streamer spread underutilised the onboard inventory of 12 streamers. For deeper water and target depths it was deemed possible to relax the constraints on near offsets. This allowed for an increased spread width and with that utilisation of all 12 available streamers. The resulting spread width of 990 m provided wider subsurface coverage for each sail-line and, by including two empty bin lines, placed the configuration on a 570 m sail-line separation. This measure increased survey efficiency by 9% compared to the base configuration, which could conveniently be used to compensate for the loss accrued by using the narrower spread (see Table 1).
Configuration | Maximum near offset [m] | Sail-line separation [m] | Efficiency gain/loss | Natural bin size dx [m];dy [m] |
Base | 328 | 525 | NA | 6.25 ; 12.5 |
12 on 10 | 359 | 480 | -9% | 6.25 ; 15 |
12 on 12 | 446 | 570 | +9% | 6.25 ; 15 |
Table 1. Comparison of the three proposed configurations. The base configuration was proposed by the survey design study presented by Souissi et al. (2023). The 12 on 10 configuration denotes the option of towing the 12 -streamer configuration on a 10-streamer template, providing shorter near offsets but at lower efficiency. The 12 on 12 represents the 12-streamer configuration towed on a 12-streamer template, achieving improved efficiency at the cost of near offsets. The maximum near offset denotes the distance from the centre of the outermost source to the first receiver on the outermost streamer on the same side and serves as a measure to compare offsets between configurations.
The two configurations provided a means of balancing acquisition efficiency and data quality, provided that the survey areas assigned to each configuration were appropriately proportioned. In areas with shallow water and shallow target depth, the narrower 10-streamer configuration could be applied to ensure short near offsets, hence reducing expected footprint. Increased acquisition time due to narrower sail-line spacing could be compensated by utilising the more efficient 12-streamer sail-line separation in areas where near offsets were less constrained. Swapping between the configurations and between their allocated areas would be seamless as no reconfiguration would be needed. This enabled us to achieve the required acquisition scope within the expected short seismic season.
To implement this approach, the survey area was divided into three distinct areas (Figure 5). The narrower 10-streamer template was planned for the AUS 105 and 106 blocks, which are situated closest to the shore and where the water and target depths are shallowest. Block MLO 121, further removed from shore and exhibiting larger water and target depths, was to be acquired using the more efficient 12-streamer template. In addition, a narrow strip comprising six sail-lines closest to shore was assigned the 12-streamer sail-line separation.
The six westernmost sail-lines were to be acquired first and would allow for test implementation of the infill strategy. The team considered the option of acquiring on pre-plot and subsequent coverage assessment as described by Monk et al. (2009) and Strand et al. (2012) but concerns about extensive feathering due to tidal currents and uncertainties in the stability of a modelled base case rendered this approach impractical. Thus, it was decided early on to acquire on coverage, albeit expecting complications in the placement decision for neighbouring sail-lines due to empty CMP lines. The feasibility of this approach needed to be tested early in the acquisition over a low-priority area.
The same sail-lines were also used to understand the impact on data quality. The project employed Bluware™ Volume Data Store (VDS) technology, where compressed seismic data were sent onshore via an upgraded high bandwidth satellite channel provided by BGP. This facilitated direct assessment of any change in data quality induced by variation of the survey configuration with an extremely short turnaround time and allowed for nearly real time decisions based on field data. Findings from this test were used to inform the final template decision for MLO 121.
For the rest of the AUS blocks, the 10-streamer template was to be used. Since the approach did not require reconfiguration, instead placing a 12-streamer configuration on a narrower sail-line separation, overlapping streamers could be used to reduce potential infill as shown by Buchan et al. (2006). For this part of the survey the 10-streamer sail-line separation had to be strictly maintained on the first offset class to ensure short enough near offsets, but on higher offset classes contributions from the outermost streamers could be used to improve coverage. This was to be achieved by excluding contributions from the outermost streamers in the first offset class when generating coverage maps. For larger offset classes, contributions from all streamers were utilised to visualise coverage and steer decisions on infill. This approach decisively reduced the infill effort and improved survey efficiency. With these modifications, pre-survey modelling indicated that the entire survey area could be acquired without compromising data quality by too large near offsets and without reconfiguration.
Implementing a flexible approach
Testing the approach on the westernmost sail-lines confirmed the expected challenges when steering for coverage. While data sent onshore and processed revealed that the impact of the increased sail-line separation on footprint was limited, maintaining the 12-streamer sail-line separation posed some problems. Separation could be maintained for the first few sail-lines but with increasing variations in spread stability, due to currents and weather, the crew started to struggle maintaining the required separation and the 10 and 12-streamer templates started to converge. Coverage maps revealed that the expected efficiency improvement through the larger sail-line separation was difficult to maintain for wider swaths (see Figure 6).
Additionally, the currents, comprising tidal currents and a strong northward current in deeper waters, resulted in increasingly stronger spread variations as the acquisition moved eastwards into deeper waters. Eddies, spun off from the northward current, disrupted the predictable tidal currents, complicating feather matching on adjacent sail-lines and resulting in increased infill, some of which could be compensated by overlapping streamers.
After completion of the acquisition of AUS 105 and 106, unexpected down time due to weather, shutdowns due to marine fauna observations and technical issues had accrued to cause significant delays. Tømmerbakke et al. (2026) provide a more detailed insight into the operational challenges encountered throughout the entire campaign. In addition, the start of the acquisition had been delayed, resulting in the seismic season nearing its end. Consequently, it was decided to postpone the acquisition of MLO 121 to the following season.
With the findings from the test lines, expected increased currents and exposure to weather as the acquisition moved away from shore, it was decided to reduce the sail-line separation for MLO 121 to the 10-streamer template. The overlapping streamers were to be used to reduce infill as any gains in operational time from the 12-streamer template were expected to be offset by increased infill. Despite not acquiring MLO 121 on a 12-streamer template, the flexibility provided by the configuration allowed us to leverage overlapping streamers and still achieve acquisition of the entire survey area within the originally allotted acquisition time, even though unexpected technical downtime forced survey completion into the next season.
Subsequent inspection of coverage maps revealed increased irregularity in sail-line spacing on block MLO 121, indicating that utilisation of the 12-streamer template would have had a detrimental impact on infill and most likely caused an increase in operational time. Data quality of the final data was fit for purpose. Expected footprints were observed on the raw data but could be successfully addressed using advanced footprint attenuation in processing, even for the six test sail-lines (see Figure 7). The flexible approach applied in this acquisition allowed us to acquire these data within the project budget and without excessive HSE exposure.
Conclusions
Facing challenging conditions for a seismic acquisition in a remote location offshore Terre del Fuego in southern Argentina, a variable sail-line separation approach was used to efficiently acquire a fit-for-purpose dataset within a short seismic season and after encountering operational challenges. The approach used a wide-tow triple source configuration placed on a 10-streamer template to reduce near offsets in areas with shallow water and target depths, while using a 12-streamer template in areas where efficiency could be prioritised. Although the majority of the acquisition area was acquired on the 10-streamer template, the flexibility provided by the configuration allowed us to leverage an efficiency increase provided by available inventory and overlapping streamers. The commitment and flexibility of all involved parties allowed Equinor and partners YPF and CGC to successfully acquire seismic data in one of the most challenging areas for seismic operations.
Acknowledgement
The authors would like to thank Equinor and partners YPF and CGC for allowing us to present this work. We would also like to thank BGP for its excellent collaboration throughout the project.
References
- Buchan, I. and Long, A. [2006] Exploring the Benefits of Streamer Overlap Shooting for 3D and 4D Seismic Surveys. SPG India 6th International Conference & Exposition on Petroleum Geophysics, Kolkata 2006, 175.
- Long, A. [2017] Source and streamer towing strategies for improved efficiency, spatial sampling and near offset coverage. First Break, 35(11), p.71-74.
- Monk, D. [2009] Fresnel zone binning: Application to 3D seismic fold and coverage assessments. The Leading Edge, 28(3), p.288-295.
- Souissi, K., Yenwongfai, H. and Brenne, E.O. [2023] Survey Design in the Austral and Western Malvinas Basins Offshore Argentina. First EAGE Conference on South Atlantic Offshore Energy Resources, Extended Abstracts.
- Strand, C., Campbell, S. and Dhelie, P.E. [2012] Reducing Infill – Real Time Application of Geophysically Based Coverage Specifications. 72nd EAGE Annual Conference & Exhibition, Extended Abstracts, P295.
- Tømmerbakke, R., Souissi, K. and Beitz, M. [2026] Operational Challenges Acquiring Austral 3D – The Southernmost Commercial Marine 3D Seismic Survey in the World. 12thCongreso de Exploración y Desarrollo de Hidrocarburos, p.43-53.
- Widmaier, M., Tønnessen, R., Oukili, J. and Roalkvam, C. [2020] Recent advances with wide-tow multi-sources in marine seismic streamer acquisition and imaging. First Break, 38(12), p.75-79.
