At the time of the survey, there were, apart from the platform, a drilling vessel, a completion vessel in the areas, and up to 10 other vessels. In addition, currents were strong, with a maximum of 2.5 to 3 knots. A streamer operation would not have been able to acquire adequate coverage. During this operation, a node was deployed practically underneath the platform. Shots were fired as close as 656 ft (200 m) to the surface obstructions.(Images courtesy of Fairfield, data from Smit et al, 2008)

During these times of downward-spiraling commodity prices, many projects are being put on hold, particularly those that can be revitalized quickly when the turnaround comes. But deepwater programs are a different breed of cat.

These high-risk, high-cost programs entail years from concept to production, so they tend to keep going without regard to the price du jour. Sophisticated technology that enables improved efficiency and cost-savings is key to making this possible.

In the all-important field development phase, ocean-bottom seismic (OBS) node technology is gaining recognition as the preferred application in many instances to acquire seismic data that will provide the best possible image of the target reservoirs.

Nodes are particularly attractive for use in the deepwater Gulf of Mexico (GoM), with its plethora of subsurface salt sheets. These salt bodies severely distort seismic signals, yielding an inferior image of the deep, complex reservoirs.

All-azimuth illumination using OBS node technology to acquire true wide-azimuth seismic data by recording in all directions is now a proven solution to this problem, according to Mike Spradley, acquisitions marketing manager at Fairfield Industries.

This approach contrasts markedly with conventional streamers that routinely record narrow-azimuth data with a single illumination direction or else re-shoot in several directions to acquire additional azimuth data — a pricey endeavor.

The Deimos field

In 2007, the Shell-operated Deimos field in 3,300 ft (1,000 m) of water in the GoM’s Mississippi Canyon area was the site of the second deepwater node survey worldwide. The successful 3-D program covered about 52 sq miles (134 sq km) utilizing Fairfield’s Z3000 node system that also was applied successfully in 2006 at BP’s Atlantis field.

Deimos sits beneath a salt overhang that had prevented adequate imaging of the field. So when Shell started developing the field about two years ago, the development folks recognized early on they needed a different sort of seismic data because of the many uncertainties about the volumes, where to put the wells, what field development system to select, etc.

“We felt we needed more certainty to determine a final development plan,” said Frans Smit, senior operations geophysicist at Shell E&P Co. “We did some testing with long-offset streamer data before the node survey, and it was established we needed very long offsets,” he said. “It would be very expensive to get these offsets (using streamers), and we had infrastructure in the area that would make it difficult to acquire a survey like that.

“When you want wide azimuth (WAZ) over a small area, it’s very hard to make it economically efficient,” Smit added. “We were automatically drawn to nodes because that tends to be more economical for a limited area.”

The first order of business for the Deimos program entailed a seismic modeling exercise to determine the optimal placement of the nodes on the ocean floor during acquisition. The modeling process used a wave equation modeling (WEM) method that handles multiples capably.

The modeled data reaffirmed that very long offsets acquired at a range of azimuth angles would be critical to illuminate the target, according to Smit. He noted that these long offsets effectively undershoot the salt body.

To ensure maximum illumination of the subsalt target area, a number of node deployment scenarios were modeled. The modeling gave the Deimos team considerable confidence that the program would be successful.

The plan was to cover the target area in a single patch using all 900 nodes available, according to Reagan Woodard, Fairfield’s operations supervisor for the Deimos project. Ultimately, 807 nodes were used.

Prior to the survey, a select number of nodes were deployed on the seafloor, acquiring a limited amount of data before being retrieved. Woodard noted this served a two-fold purpose:
• To determine the optimal gain setting for each type of sensor; and
• To verify the operational readiness of the nodes and the deployment/ retrieval methodology.

Field operations

Field operations at Deimos were implemented via a dual-source shooting vessel and a remotely operated vehicle (ROV) that deployed the self-contained autonomous nodes on the ocean floor, ensuring positional accuracy as well as repeatability.

Smit noted a major advance during the survey entailed the efficiency gained via use of a nodebasket — a tool newly developed by Fairfield for the operation. Once operational, the nodebasket reduced the strain on the ROV considerably by limiting the number of necessary ROV trips.

The acquisition logistics at Deimos were dictated significantly by the 60-day battery life of the nodes.

“All of the relevant data had to be recorded before the nodes were retrieved to recharge the batteries and download the data,” Woodard said. “The speed with which the nodes were deployed and retrieved proved critical to the rate of progress of the survey.”

Spradley emphasized that every node was deployed and retrieved as planned, documenting the operational flexibility of OBS nodes in highly congested environments.

Strong loop currents in the area would have made streamer operations chancier than usual, particularly long streamers, Woodard noted. Repeatability, which is a fundamental feature of node surveys, would have been impossible using streamers.

The pre-processing stage of the Deimos program focused on delivering four main products:
• Accurate source and node positions;
• Accurate timing of the recorded data;
• Removal of any shear-wave leakage on the vertical geophone; and
• Splitting the data in upgoing and downgoing wavefields.

A global process was used to deal with the post-survey positioning calculations and uncertainties in the time break of the recordings. Concerning shear wave leakage, a method was used where removal of the leakage was closely linked to the wavefield separation.

The pre-processed data were taken through mirror migration, initially using the same WEM algorithm utilized for the modeling, Smit noted. Using this approach, the downgoing energy is used for imaging.

Comparing the resulting OBS node data with existing narrow-azimuth data over the field revealed significant improvements on the nodal data in terms of signal to noise, multiple content, and structural definition, especially under the salt overhang.

“One of the big advantages for us on the interpretation and processing side was the ability to test a lot of alternative velocity models very quickly,” said Mark McRae, senior staff geophysicist on the Deimos field team at Shell. “That’s a function of the way OBS is acquired.

“When you process the data you do a reciprocal operation on it so you turn the nodes into shots in the processing,” McRae said. “We had 807 nodes, so when we process and migrate we only have to do 807 shot migrations, whereas with a conventional WAZ you do a whole lot more shot migrations. The very fast migration turnaround times were very beneficial. As an interpreter, I need to test many models.”

He said the well-to-seismic ties are substantially improved with OBS nodes, noting this is relative to the two different directions of narrow-azimuth streamer data they have over the area.

“A key point there is that in tying seismic to wells, we’re now for the first time in the Deimos field confident we can see the hydrocarbon fluid effects,” McRae said. “We can see an amplitude effect that is correlatable to where the hydrocarbons are in the reservoir, giving us a lot more confidence in future development of the field.”
McRae acknowledged Deimos non-operating partner BP for working closely on the project and aiding their ability to acquire good data.

“We get questions about node surveys almost every week,” Smit said. “There’s a lot of technical interest.”

A caveat: “These sort of surveys also are very expensive, and each asset must consider their commercial aspects,” Smit cautioned.