Testing carried out with the KV Svalbard in Storfjorden in Spitsbergen effectively represented a model test at 1:2.5 scale with real ice.

Floating production, storage and offloading vessels (FPSOs) that work where there is ice for all or part of the year face significant design and operational challenges. Where icebergs and ice ridges are present, it is extremely difficult to design a mooring system that can resist largest ice loads, which means a quick-disconnect system is required.

Furthermore, because there is virtually no data for typical ice load levels experienced by these large moored vessels, it is difficult to establish reliable guidelines for disconnection frequency and resulting downtime estimates.

Ice management experience primarily relates to drilling operations.

The effect of ice management on load levels for large vessels has not been well quantified. The occurrence of growlers, bergy bits, and icebergs within the ice pack in conjunction with ice and wave conditions presents new challenges for iceberg detection and management.

FPSOs are normally designed for a particular ice class, but this designation mainly applies to vessels in transit and is not totally relevant to moored systems. Possible iceberg impact needs to be considered separately.

To be effective, floating production systems (FPSs) will have to be capable of safe, efficient station-keeping. Otherwise, downtime could have a significant impact on the operability and economic viability of the FPS approach.

The FPSOs working on the Terra Nova and White Rose fields off the East Coast of Canada have achieved several years of safe operations. Even though the chance of iceberg impact is low, problems can occur with pack ice intrusion, which originally was not considered to be a significant threat on a year-to-year basis.

Ice on the White Rose field

A recent incident at the White Rose development highlights the problems with operating an FPSO in a hostile ice environment. The CBC reported on its Web site on April 8, 2008:Husky Energy is removing 32 workers from the SeaRose floating production vessel at the field, about 300 km southeast of St. John’s, with heavy ice causing a halt in production. The SeaRose is on standby to move if necessary.

“We’re fully prepared at the moment to disconnect should we decide to do so, and we have no firm plans to do so at the moment. But just to disconnect only takes about 20 minutes,” said Husky Energy Vice-President for East Coast Operations Ruud Zoon.

As well, Husky is having a drilling rig towed away from the field. The Global Santa Fe Grand Banks rig lifted its anchors on Tuesday and began moving south. The forecast is for southeast and southwest winds that are expected to disperse some of the ice. Until then, Husky Energy can only watch and count the losses. “On a good day, we produce about 130,000 barrels a day. With the price of oil $100 a barrel, you can do the math yourself,” Zoon said.

An ice map issued on April 8 illustrates the ice hazard in detail. The southernmost edge of the ice pack drifted as far south as 45?N. More than 62 miles (100 km) north of the FPSO, there was ice of up to 9/10th concentration, with maximum ice thicknesses of 27.6 in.

Information from a variety of sources, however, indicated that an ice-class anchor handling tug was able to enter into the pack ice and that it was able
to navigate the ice without difficulty through 10/10th coverage 16 to 20 ft (5 to 6 m) thick. It had no trouble in clearing away a swath of ice 49 ft (15 m) wide using a water cannon. The operator used this information to reach a decision to have several of these vessels on standby instead of moving offsite.

Reliability of ice predictions

The normal process of calculating global pack ice loads is to predict design ice conditions for a given probability of occurrence and, on the basis of theoretical models and ice tank tests, to translate ice loads into individual mooring line tensions.

Previous studies for White Rose indicated that the area experiences only short periods of 9/10ths ice coverage and that most of the ice coverage ranges from 1 to 3 ft (0.3 to 1.0 m). This appears to be consistent with the ice maps during the time of the incident.

Clearly, reliable field data is vital, and contingency plans are needed not just for the rare occurrence of iceberg impact, but also to cope with the greater likelihood of pack ice threats. In retrospect, a good understanding of the ice environment and reliable methods of coping with pack ice may have allowed production to continue.

Field tests for station-keeping in ice

Apart from the possible deficiencies in actual field data, there remains the problem of the accuracy of current global load models. Theoretical calculations are most often based on icebreaker model tests at typical icebreaking speeds. An FPSO has a different bow shape, and the relative ice drift speed is very low under typical station-keeping conditions.

The Research and Maritime branches of Det Norske Veritas (DNV) are carrying out field programs and desk studies on global ice loads for FPSOs and drillships. Earlier this year, the opportunity arose to join an ongoing field program that involved the Norwegian Coast Guard vessel KV Svalbard. Special instrumentation that had been previously installed was used concurrently with the Ice Load Monitoring (ILM) program managed by DNV Maritime to evaluate parameters for an extended station-keeping field program.

On Feb. 29, 2008, the vessel began testing in Storfjorden in Spitsbergen, Norway. Ice loads and vessel parameters were measured in stationary ice at very low speeds.

While keeping a constant speed presented problems, it was decided that constant power would be used while the vessel speed was continuously measured. Global loads were back-calculated from actual thrust measurements. The parameters for the KV Svalbard were then transformed into scaling factors for typical FPSOs, with adjustments made for bow shape and vessel dimensions.

Actual ice thicknesses and properties were measured in situ. These can also be scaled to conditions in other locations. Effectively, this represented a model test at 1:2.5 scale with real ice, as opposed to an ice tank test at 1:30 scale with model ice.

Plans are in place for additional tests to be carried out next year under typical pack ice and ridged ice conditions.

Conditions in the Barents Sea

While the FPSOs operating on the Terra Nova and White Rose fields are operating in sub-Arctic conditions, future production vessels operating in the Russian sector of the Barents Sea face much more difficult challenges.

Current development plans for the Shtokman field include only a surface control vessel or floating platform (natural gas will be transported to shore by pipeline). It should also be mentioned that the reliability of ice and iceberg data for the Barents Sea is still under review — these should be considered only as estimates.