Bilge keel refit, offloading re-arrangement allows Ramform to resume Banff service

The newly modified Ramform Banff (in foreground), back in the in field.

PGS was already experienced in this role through its Atlantic Power division in Aberdeen, and through various assignments in UK and Norwegian waters with the Petrojarl FPSOs, designed by Golar Nor. However, it wanted to broaden its remit to working with its own production concept.

According to Atack, Conoco and its partners had numerous reasons for choosing the Ramform FPSO. "We were a vessel provider that would extend to a full range of services, including provision of the subsea manifold. By doing this, we would reduce the operator's capital investment in their first phase four-well development to a minimum. We also provided a part-tariffed service contract, with a minimal bill for shutdown if the field didn't deliver. At the time, the oil price was not wonderful, and that form of contract was considered the lowest risk.

"Despite reported motion problems, it was in fact the oil export system that was the cause of many of the Banff's early difficulties. The Ramform for Banff is more a floating production unit (FPU) than an FPSO, in view of its storage capability of around 120,000 bbl. The FPU concept we had would have avoided storage altogether, with oil and gas pumped out through existing oil and gas export lines. In the event, Conoco secured an exit route for the gas into the CATS trunkline in the central North Sea, but oil export into the Gannet-Fulmar line was not available." Offloading to two 700,000 bbl dedicated shuttle tankers was therefore adopted, using APL's suction anchor loading (SAL) system.

Problems begin

The drawbacks of this approach became apparent soon after startup in January 1999. These included the maneuvering of the shuttle tankers onto the SAL system and their stationkeeping. But the arrangement itself proved unsatisfactory, Atack says. "Export oil is pumped from the Ramform through a subsea manifold and loading hose into the storage tanker. The hose and vessel mooring hawser are supported by a midwater buoyancy chamber, and further flotation devices.

Revised export system diagram.

"We experienced operating difficulties with the system's sensitivity to tanker positioning, hose handling, and hawser buoyancy. We had particular problems with the tanker over-running the subsea equipment, tangling the hose and hawser and potentially over-tensioning the system. On occasions, during the cast-off process in bad weather, when the tanker released the hawser/hose end, it would drop straight to the seabed, and we had to mobilize another vessel to pick them up. These issues caused four instances of lost production export. Two took over a week to remedy because we had to dig the hose end out of the seabed." These difficulties in 1999 resulted in approximately 38 days of production export loss.

Submerged turret loading

In January 2000, PGS decided to replace the suction anchor loading system in favor of an export route, using a permanently moored storage tanker with submerged turret loading, also provided by APL. This would limit the need for dedicated shuttle tankers, would make oil production less dependent on weather, and would boost the Ramform's limited storage capacity.

PGS took out a lease on Ugland's double-hulled, 131,000-metric-ton tanker, the Nordic Apollo, which had storage capacity of 950,000 bbl. The bow section was modified to incorporate a conical receptacle for an APL retrievable submerged turret loading (STL) buoy to moor the vessel. Once moored, the vessel would be free to weathervane and to receive Banff crude. A stern offloading system and helideck were also installed. The re-worked vessel was available for service in July 2000.

Oil enters the tanker through a flexible riser hitched up to the turret swivels. In turn, this riser is connected to the Ramform's export flowline via a pipeline end manifold and the existing SAL base. The STL mooring system is designed to keep the tanker inside a given offset in a passive mode. When significant wave height exceeds 7-8 meters, a thruster system comes into play, for heading control and surge damping. When the FPSO is unable to maintain position, it disconnects. Reconnection can then be effected in sea states with significant wave heights up to 4-5 meters.

Excessive roll motions

Adapting the Ramform design to production and onboard storage necessitated seastate response tests on a model at the Marintek tank in Trondheim. "Our designers decided to stick with the Ramform hull shape and fettle that we had used half a dozen times before," Atack says. "Originally the designers considered fitting bilge keels, but the early model tests indicated they wouldn't be needed for the stability of the vessel.

However in practice, Atack says, "it took us a long while to determine that the roll motion of the Ramform was too much in heavy seas. The media were very keen on reports that it was making our crew seasick, but that was really not the problem. Once the vessel moved beyond a certain range - 7.5 degrees of roll - operations had to be closed down. In fact, there were not many days when 7.5 degrees roll was exceeded.

Bilge keels being fitted to Ramform Banff.

"However, we determined that the behavior of the vessel still didn't fit our expectations. This was unacceptable, in considering the effects of severe weather on the structural integrity of the vessel's equipment. This called into question the original model testing, so we conducted more model tests, which proved that the roll motion wasn't good enough.

"We performed three series of model tank tests and found that through the simple addition of bilge keels, we could suppress roll by 40% - a huge degree of damping. This would apply not only to extreme weather, which occurs less than 1% of the time in our location, but would also calm the vessel in a regular seastate.

New bilge keels

"It looked simple on the model - but in practice, we were talking about fitting 1.4-meter wide bilge keels along about half the 120-meter-long FPSO. So this would clearly be a significant construction project. We did look at doing the work at sea, to avoid a shutdown, but this would have required a sustained period of calm weather. Then we would have to have put the keels on in sections, so it would never have been as good a job as if performed in a drydock. And it would also have taken longer to complete.

"In a drydock, we could have a go at a run of other things we wanted to change, such as the fairleads, thruster controls, and various minor process items. We had seen too much wear on the mooring fairleads, due partly to the vessel's motions, so we planned to re-work these to a different design. We had most of the engineering/specifications done well before the vessel arrived in drydock at the Blohm + Voss yard in Hamburg in October 2000." The yard would also carry out fabrication and structural improvements such as re-painting and pipe support reinforcements, which would have been implemented anyway after five years in service (the normal reclassification period for FPSOs).

Contract situation

Conoco and its partners agreed to halt production from Banff in October 2000, while PGS completed the modifications. There had also been a secondary contract with Ranger Oil to develop Kyle, some 15 miles away as a subsea tieback to the Ramform. With oil prices, soaring, Ranger's group decided to press ahead with an extended well test on Kyle. PGS was able to oblige by bringing in the Petrojarl I, which had just been released from Talisman's Blenheim Field. The strong test result led to Kyle being developed instead through Shell's Curlew floater. "We were sorry to lose that work," says Atack.

The Ramform was in drydock in Hamburg for around five months. "We did the bilge keel work on schedule, but other tasks took longer than expected. Also, unusually, we had taken the risers out of the field to purge seawater from the flooded sheaths. Having displaced seawater and inerted the annular space, we had to prove that no damage had occurred to the risers, so we undertook laboratory tests and also testing underneath the riser sheaths.

For the wire tests, we cut out sections from one of the old risers, which had been most highly stressed during its operation. We then had to re-certify the risers for re-use. We had to develop specific testing regimes rather than rely on a design factor to guarantee riser integrity, which would be more relevant for a new product."

Disconnection of the vessel had been effected during the first half of October 2000. "It turned out to be quite complicated getting the risers out without overbending. Getting the moorings off proved easier." Once the re-fitted vessel left Blohm + Voss in late February 2001, it headed straight back to its mooring location on Banff.

Mooring changes

The mooring system is comprised of 10 chains and 10 anchors. "As a modification, we decided to add a second anchor winch to the turret. We wanted to be able to routinely move the anchor chains across the fairleads to avoid excessive wear on single links. That had been difficult with only the one original winch."

Banff has been developed with two production wells and two water injectors. There are six risers connecting the vessel to subsea systems - two for production, one for water injection, one for gas export, one for oil export, and one control umbilical. Re-starting these systems proved to be straightforward. "We had gone through the process of determining the optimum throughput for the wells. In the event, we came on at the same rate as before, 25,000-30,000 b/d, in March 2001.

"Since then, we've been able to operate with 99% uptime, which is the same as our other FPSOs in the North Sea. The modifications were immediately obvious to the crew. Some of our crew had been skeptical that a 35,000 ton displacement vessel could be changed by a 'little welding', but everyone is now impressed by the different feel of the vessel, even in a calm sea. This has given us and the crew greater confidence." So far, roll motions of the re-configured vessel have not exceeded three degrees. Offloading to the Nordic Apollo has also worked well, with 100% uptime, as expected, and no problems reported.

PGS would like to see the Ramform continue to produce for a few years at a high rate, rather than a slow, protracted tail-off. Any further plans that Conoco might have, to accelerate production or tie-back area reserves, would help to pull back the current capacity surplus.

Despite the setback with roll motions, Atack claims that use of the Ramform for floating production has been shown to work, particularly as an FPU. "A lot of the time, operators don't want storage onboard. In the Gulf of Mexico, with all the infrastructure in place, it's conventional not to have storage onboard." When the MMS (US regulatory agency) decides to lift its ban on FPSOs, selected hulls are likely to be smaller than vessels operating elsewhere. These factors could work in the Ramform's favor, Atack believes.