TGS–GTI seismic tie-up: the standards behind "shoreline to deepwater"

What the TGS–GTI agreement puts on the table

TGS and Geophysical Technology Inc (GTI) have signed a four-year strategic collaboration agreement to pursue seismic acquisition projects jointly across shallow water, transition zone and deepwater, reported on 14 August 2026. Work will be evaluated and pursued case by case rather than under a single blanket scope.

The split of capability is clear from the announcement. TGS brings deepwater acquisition – its deepwater vessels, ocean bottom node (OBN) assets, global supply chain and project management, plus financial strength. GTI brings its proprietary NuSeis nodal recording technology and manufacturing base, specialised vessels and recording assets, and land, transition-zone and shallow-water experience, with an established regional presence across the Americas, the Middle East and Africa. Carel Hooijkaas, EVP Marine Data Acquisition at TGS, described the goal as offering clients “a truly integrated seismic solution, from the shoreline to the deepwater, without compromising on quality or efficiency.”

That last phrase is the engineering claim, and it is the one we would take apart first. Combining two fleets and two node inventories is a commercial and logistics exercise. Delivering a single dataset that a processing team can migrate as one coherent volume, across three acquisition regimes and two contractors, is a standards exercise. In our experience the two are not the same problem, and the second is the binding one.

The real test: one specification across three regimes

Shallow water, transition zone and deepwater are not three points on a depth scale. They are three different acquisition physics, three different deployment methods and three different noise and positioning environments. A node placed by work-class ROV at several thousand metres, a node laid on a rope through the water column, and a node planted by hand on a tidal flat share a data format and little else.

“Without compromising on quality” therefore has a specific technical meaning that we think the ITT should force into the open. It means every node contributes traces to a common bin grid, referenced to a single geodetic framework, on a single time base, meeting one set of acceptance thresholds for positioning residual, timing drift, sensor coupling and noise. If any of those varies across the seam between environments or across the seam between contractors, the merged volume carries a discontinuity that processing can suppress but not remove.

The industry already has the yardsticks for this. Positioning data exchange runs on the IOGP P-formats – P1/11 for processed positions, P2/11 for the raw marine record, and the P6/11 bin grid definition that ties every source and receiver to the image space. Field node data is delivered in SEG-D, typically rev 3.1 for modern nodal systems, with processed output in SEG-Y rev 2. The question an integrated programme raises, in our view, is not whether these standards exist, but whether both parties apply them to one shared definition, or two parallel ones that meet only in a merge script.

Positioning and timing: where the seams actually form

Node positioning is the first place a two-environment programme can quietly drift apart. Deepwater OBN work positions each node acoustically during deployment, then refines it in processing by inverting the direct-arrival and first-break travel times – the technique commonly called first-break (or refraction) repositioning. The acceptance criterion is a horizontal residual expressed as a fraction of the bin dimension; with bins of 12.5 m or 6.25 m, node positions generally need to be resolved to a metre or better for the imaging to hold up. That standard has to apply identically to a node the ROV placed at depth and a node dropped on a rope in a hundred metres of water.

The geodetic framework is the seam we most often see open unnoticed. A programme that runs from the shoreline seaward will touch land RTK GNSS on the beach, kinematic and acoustic positioning through the transition zone, and USBL or inertial-aided navigation in deepwater. Each of those chains can be internally consistent and still leave a datum offset at the boundary if the coordinate reference system, epoch and vertical reference are not fixed once, for the whole job, before the first node goes down. A few tens of centimetres of unresolved datum shift between a nearshore block and its deepwater neighbour will not fail any single QC gate, but it will misregister the volumes when they are stitched.

Timing is the second seam, and it is particular to autonomous nodes. Unlike a streamer, a node has no telemetry back to a master clock. Each unit carries its own oscillator, disciplined to GPS at deployment and again at recovery, and the drift across a deployment that may run for weeks is corrected in processing against those two fixes. Sub-millisecond timing error maps directly into imaging error, so we treat the drift budget as a hard specification, not a nicety. When two contractors bring two node types to one survey, their clock discipline procedures, drift models and rejection thresholds have to be reconciled to a single tolerance, or the timebase itself becomes the discontinuity.

The transition zone is the hard interface, not the shallow end

Deepwater is expensive and slow, but it is procedurally clean. The transition zone is the environment we have seen defeat integrated programmes, because it is where the positioning chain, the source, the deployment method and the environmental constraint all change at once and none of them changes cleanly.

Positioning through the surf zone and tidal flats hands off between land survey and marine acoustic methods across a band where neither works well. Sources mix – airgun arrays offshore, land vibrators or explosives inshore – and the amplitude and signature matching across that boundary is a processing burden that starts as an acquisition decision. Tidal windows govern access, node coupling to soft or mobile sediment governs vector fidelity for multicomponent data, and obstacle density governs where a node can physically sit. A pre-lay hazard and bathymetric survey to IHO S-44 standard is not optional here; it determines both the safe lay corridor and the achievable node grid.

The transition zone is also where the energy budget bites hardest. Autonomous nodes must record for the full deployment on internal power, and the interval between deployment and recovery is set by the environment, not the battery. The endurance margin, the pressure-housing integrity and the thermal envelope of the node population are real acceptance items rather than catalogue figures, and the hard-won lessons around powering sealed instruments through long subsea deployments apply directly to a node inventory drawn from two manufacturers with different cell chemistries and duty cycles.

Where integrated programmes quietly go wrong

The failure modes on a job like this are rarely dramatic. They are seams that pass every local check and only show up when the volumes are merged. Five recur often enough, in our experience, to design against.

1. The datum that was agreed but never documented

Two contractors agree “WGS84” in a kick-off meeting and each implements it against a different realisation or epoch. Nothing fails on the vessel. The offset surfaces at merge, weeks later, when re-acquisition is no longer an option. Fix the CRS, epoch and vertical reference once, in an IOGP P6/11 bin grid and P1/11 exchange definition issued to both parties before mobilisation, with the transformation and its stated uncertainty written down.

2. Two node timebases, one dataset

Each node type carries its own clock discipline and drift model. If the drift tolerance and the GPS re-sync procedure are not harmonised to a single number, the timebase becomes environment-dependent. Specify one drift budget across all node types and require per-node drift correction reports as a delivery item, not a QC afterthought.

3. QC thresholds that differ by contractor

Positioning residual, coupling, and noise acceptance limits set independently by each party produce a dataset that is uniformly good within each block and inconsistent across the seam. One acceptance specification must bind both fleets, with a single QC authority holding veto over both.

4. The ROV interface treated as a given

Deepwater node placement by work-class ROV is a survey operation in its own right – node handling and deployment run under IMCA R 004 for ROV operations, and the positioning aiding chain (USBL, INS) and cycle-time assumptions drive the schedule as hard as weather does. Where node-on-a-rope is substituted for ROV placement to gain rate, the positioning-accuracy consequence has to be stated, not absorbed silently.

5. Interface control between two organisations on one image

The commercial framework is a collaboration, but the technical accountability for the merged product still has to sit with one authority. Where two separately contracted parties acquire adjacent parts of the same dataset, the governance questions are the same ones that arise when separately contracted contractors share a single worksite: who holds the master specification, who decides on a deviation, and who owns the seam.

What to specify before award

We read the collaboration as sound engineering logic – deepwater strength paired with shallow-water and transition-zone capability is a genuine capability match. Whether a given call-off delivers on the quality promise is decided in the specification, not the partnership. If you are the client or the survey authority writing that scope, the following belong in it.

• Fix the geodetic framework first. Issue a single CRS, epoch, vertical reference and P6/11 bin grid to both parties before any deployment. Require all positioning delivered in that framework with documented transformations and stated uncertainties. Treat any datum offset at a block boundary as a non-conformance.
• Set one node-positioning acceptance threshold. Express it as a horizontal residual against bin dimension and apply it identically to ROV-placed, rope-laid and hand-planted nodes. Require acoustic and first-break repositioning QC on every node regardless of environment.
• Harmonise the timing budget. Specify a single clock-drift tolerance (sub-millisecond), require GPS discipline at deployment and recovery, and demand per-node drift-correction reporting as a deliverable. Reject nodes outside budget.
• Bind both fleets to one QC specification. Define common thresholds for positioning, coupling, timing and noise, a single SEG-D revision and header convention, and appoint one QC authority with veto across the whole survey.
• Control the transition-zone risk explicitly. Require a pre-lay hazard and bathymetric survey to IHO S-44, define tidal-window and coupling assumptions in the schedule, and state source-matching provisions across the airgun-to-land boundary.
• Manage the two-contractor interface as an engineering interface. Name the single technical authority for the merged product, document the deviation process, and hold combined-operations procedures where both parties work one block.

The same discipline is spreading beyond hydrocarbon exploration. Nodal and OBN methods now sit inside renewables site characterisation, where the expanded site-survey scope now expected for offshore renewables pulls high-resolution seismic into shallow water and transition zones under a tighter integrity regime. In our view a partnership built to move cleanly between those environments is well placed to serve that demand – provided the standards holding the seam together are specified as deliberately as the assets are combined.

---

Based on: TGS and GTI Collaborate on Seismic Acquisition