Ekofisk pipelines undergo requalification

By 1998, Ekofisk center will be joined by two new platforms, one for drilling and another for processing and transportation (Fig. 1; photograph courtesy of Phillips Petroleum Co. Norway).

Requalification is a reassessment of an installation's design with changed design parameters so that the facility can continue to operate. In some cases, original design requirements may have been unrealistically conservative.

As an engineering service, requalification will increase in importance and extent in the future, as the installations get older.

The Ekofisk field experienced seabed subsidence as a result of compaction of the producing formation. A total of 25 pipeline systems with diameters of 6-36 in. now run within the subsidence bowl. The existing pipelines and the risers had to be requalified because of new design conditions.

This requalification also included an investigation of how much of the nine existing pipelines that were to be rerouted in the redevelopment of the field could be reused.

Procedures were prepared for applying probabilistic evaluation because a simplified screening study performed earlier (called Phase I) showed that several of the pipeline systems did not comply with the then-used acceptance criteria.

During the present assessment, however, it became clear that use of the deterministic structural acceptance criterion in the Norwegian Petroleum Directorate pipeline regulation and of the allowable strain criterion in BS8010 was sufficient basis for the necessary actions.

Ekofisk subsidence

Ekofisk, in the southern Norwegian North Sea, is one of the world's largest offshore oil and gas fields (Fig. 1). It was discovered in 1968; first production was in 1971. Its original design life was 30 years.

The Ekofisk center complex serves as a production and transportation hub both for the Ekofisk fields and several other gas and oil fields.

Each day, approximately 85,000 cu m (530,000 bbl) of oil are being piped 354 km (220 miles) to the Teesside terminal in England. Another 56 standard million cu m (2 bcf) of gas are piped 441 km to the Emden terminal in Germany by the two Norpipe pipelines operated by Phillips.

When Ekofisk started production, the recovery factor was estimated to be 17-18%. To the present, the factor has doubled to around 35%. The original operating license for the Phillips group lasts until 2011.

The increased recovery will extend the producing life of the field; the license has therefore been extended to 2028. This implies that the Ekofisk installations are needed for significantly longer than originally designed.

Compaction of the producing formation has caused the seabed to move horizontally inward, towards the Ekofisk Center, in a movement called "subsidence-related horizontal movement" (SRHM).

The Phillips group decided partly to redevelop the Ekofisk complex. One new drilling platform (2/4X) will be installed this year and one new process and transportation platform (2/4J) will be installed in 1997.

Reasons, regulations

A requalification assessment is a re-evaluation of an original design. It may be triggered by change of the original design bases, mistakes or shortcomings that have been discovered during construction or operation, change of operational parameters, deteriorations that have exceeded design assumptions, lifetime extensions, and/or discovered damages.

The Norwegian Petroleum Directorate pipeline regulation Section 43 states: "Repair and modifications shall not impair the specified safety level."

This regulatory requirement implies that within the original design lifetime and without essential changes to the originally foreseen operation, regulations and design standards under which the pipeline was designed and built still apply for repair and modifications.

The pipeline design is based upon maintenance of a minimum safety standard throughout the lifetime. If deterioration is present, the safety level will be reduced with time, but the minimum reliability target must be met.

In case of a lifetime extension, the minimum reliability target must still be complied with.

Certain phases and decisions are common to requalification (Fig. 2 [20152 bytes]).

Pipeline information can be gathered from design documents or operation and inspection reports. Simplified structural calculations on the pipeline can assess whether it fulfills design requirements.

If not, corrective actions may be performed or a more precise appraisal may be executed if design requirements are complied with. If not complied with, execute corrective actions or try a more advanced calculation methodology, for example structure reliability methods.

Structural design, requalification

A multidiscipline team must be able to perform a complete requalification assessment of a pipeline system. Single discipline assessments are of limited value unless the assessments have been performed in close cooperation with the other disciplines.

If corrective actions are required, such as changing operational parameters or repair, different alternatives should normally be established. The final decision will be based upon an economical and technical evaluation of the different alternatives.

The primary requirement for structural design is that the capacity (including material factor) exceed the load (including a load factor). This is illustrated in a "Load and Resistance Factor Design" (LRFD) formula shown in an accompanying box.

Fig. 3 [26250 bytes] shows an overview of a requalification process.

Uppermost in Fig. 3 are some reasons for requalification and related deterioration processes, given in the center ellipsis. The two smaller ellipses ("Loads," "Capacities") illustrate two sides in the equation.

The various deterioration mechanisms will affect either load or resistance aspects; some both. Further, the figure shows that requalification does not necessarily include deterioration; an example can be changes in load or increased requirements for safety.

The lower box represents evaluation in the requalification assessment, that is, checking for the difference in the equation.

The result will often be neither "yes" nor "no," but a "yes" depending on, for example, inspection carried out to confirm that corrosion does not exceed a given value (varying with time).

Various alternatives for possible corrective action may also be considered.

The purpose of data retrieval is to obtain sufficient and reliable information of the pipeline condition. The retrieval is naturally divided into two parts, design and "as-is."

The pipeline to be requalified may be quite old. It is important, therefore, thoroughly to verify "design" data.

The "as-is" condition can be collected from such different sources as process monitoring, internal and external inspection reports, and damage reports.

When inspections are planned, it is important to specify the data to be obtained and detection tolerances to ensure proper selection of inspection methodology and equipment.^{2 3}

The major loads in pipelines result from pressure and temperature. Additional loads may derive from current and wave-induced vortex shedding loads in free spans, ground settlements, subsidence, or landslide.

Deterioration mechanisms result in degradation of the pipeline's structure.

The most common deterioration mechanisms are internal and external corrosion, erosion, damages from third-party operations (fishing or trawling, marine anchoring, and dropped objects), and fatigue developed from free spans (caused by seabed scour, upheaval buckling, or settlements of supports).

Two alternatives exist for documenting structural capacity: permissible-stress check or limit-state check.

Limit-state check may allow plastification and a higher utilization than the permissible-stress check.

In the permissible-stress check, the hoop stress and equivalent stress are compared to the yield stress with a certain use factor, all according to the design code.

For the limit-state check, all foreseen failure modes are checked individually. Typical failure modes are yielding criterion, leakage, buckling, ovalization, accumulated strain, unstable fracture/plastic collapse (UFPC), and fatigue.

Ekofisk requalification

The difference between a pipeline system with 30 years' design life and one with 50 years' is mainly capacity against deterioration. The two dominant deterioration mechanisms are corrosion and fatigue.

The Ekofisk requalification assessment covered the following:

• Loads caused by the subsidence-induced horizontal seabed movement

• Identification of possible mitigation actions

• Lifetime-extension assessment based upon corrosion inspections, results, and theory

• Establishment of the tie-in locations for the eight pipelines to be rerouted, based upon structural and material considerations

• New wave, current, and wind design data.

The requalification carried out for the Ekofisk pipelines was based upon original design information and data obtained by inspections.

Most of the pipelines on Ekofisk are trenched and have a natural backfill of soil. No data from soil samples taken from pipeline trenches at Ekofisk are available. The study therefore had to be based upon existing data of the original Ekofisk soil and on literature survey.

On this basis, design soil properties have been established.

The mean estimated values of soil parameters or properties have been used in the pipeline assessments.

Selected results from the study are given in Table 1 [6477 bytes], Table 2 [6676 bytes], and Table 3 [6526 bytes].

Material, corrosion

The Phase I study indicated that compliance with the Norwegian Petroleum Directorate's (NPD) permissible stress criterion could be difficult for all Ekofisk pipelines. The calculations showed that several of the pipelines would exceed the stress limitations.

However, the pipeline design codes allow use of plastic strain as an additional acceptance criterion.

Material samples from four available pipelines were tested in order to evaluate the influence of plastic strain on the mechanical properties. The base materials and longitudinal welds (where present) were tested up to 10% permanent strain.

The potential for strain aging was tested as well. It was concluded that base materials retain acceptable mechanical properties up to a strain level of 5%.

It is not expected, however, that the pipelines will experience a strain level exceeding 0.5% so long as upheaval buckling does not take place.

To enable fracture-mechanics calculations for welds with defects, design crack-tip-opening-displacement (CTOD) values had to be established. No girth welds were available for testing.

This omission led to a literature survey to establish the required design CTOD values, the CTOD value of 0.1 mm being considered conservative.

Internal corrosion and corrosion-like defects have been detected in several of the Ekofisk pipelines. The corrosion has occurred as scattered pits which sometimes overlap.

One gas pipeline has reported girth-weld corrosion.

The inspections performed in the past with reliable inspection vehicles showed that the corrosion rates were unacceptable in some of the pipeline systems. As a result, the process equipment and operating procedures have been improved.

The pipelines on Ekofisk were originally designed to be buried.

Some pipelines, however, have been reported to have exposed areas for many years. This may result from difficulties experienced from very hard or stiff clay in the pipe trench encountered during original burial.

Inspections performed in recent years have revealed, in the subsidence "compression" zone, development of free spans possibly caused by temperature and subsidence-induced compression causing upheaval buckling. Erosion of the seabed is assumed to have little influence on the development of the existing free spans.

The ROV-operated pipeline survey has detected upheaval buckles on a 10 in. and a 12-in. pipeline. These lines have been reinspected by a geometry pig.

Pipeline strain and bending radii for the 10-in. pipeline are shown in Table 4 [7205 bytes].

Deterioration mechanisms

Two methods for assessing future corrosion have been used: theoretical calculations based upon the specification for the content in the pipeline and extrapolation of inspection data obtained from inspection tools.

Each method has its shortcomings. For predictions, engineering judgment had to be used.

Three corrosion rates which could be used for assessing the future lifetime were "lower bound," "best estimate," and "upper bound."

Selected results from the study are given in Tables 5 [6568 bytes] and 6 [6283 bytes].

If a free span exceeds a predefined length for each pipeline, rock dumping is performed as a corrective to stabilize the pipeline.

Subsidence effects

The subsidence-affected area at Ekofisk field forms the shape of an elliptic bowl with maximum north-south extension of 12 km and 9 km east-west. The center of the elliptic bowl nearly coincides with the Ekofisk center.

Horizontal motion caused by subsidence increases inward from the outer edge of the subsidence bowl (Fig. 4 [52380 bytes]). The maximum horizontal motion, occurring in a zone between 1.5 and 2.5 km from Ekofisk center, is calculated to be less than 2.5 m in 1995 for a vertical subsidence level of 6.5 m.

Fig. 5 [12031 bytes] shows maximum future horizontal seabed movement.

Inside the point of maximum horizontal movement, the seabed movement decreases until it reaches zero at the center. The pipelines within this region experience compressive forces (Fig. 6 [20844 bytes]). Conversely, outward from the point of maximum horizontal motion, pipelines experience tensile forces.

Depending upon the axial pipeline stiffness and the frictional relationship between pipeline and seabed, pipelines will not necessarily move at the same rate as the seabed.

The maximum pipeline force gradient is governed, then, by soil-friction capacity which, in turn, is affected by the burial depth. The pipelines will develop significant elongation in the tension zone.

Near the Ekofisk center, for example, friction forces are not large enough to prevent the pipelines from sliding relative to the seabed, and pipelines and risers are consequently pushed toward platform jackets.

Calculation of forces

All risers were structurally 3D modeled to determine loads and stresses. The global pipeline assessments were performed by non-linear finite element (FE) models to determine pipeline loads and stresses.

The aim of the pipeline-load calculation was to calculate the stresses for the capacity evaluation of the pipeline itself and to determine pipeline expansion towards the platforms.

The significant axial loads in the pipelines made a true nonlinear 3D pipeline material model necessary, which adjusted the actual yield stress based upon the circumferential stress.

The nonlinear soil-pipe friction, varying with burial depth, was included in the FE analysis. The lower parts of the risers were modeled in order to give representative pipeline-end stiffness and corresponding reliable riser-bend displacements.

Loads were established with input data for design temperature, present operating temperature, design pressures, present operating pressures, and the horizontal seabed movement corresponding to vertical subsidence of between 6 and 20 m.

The tension zones of the pipelines experience a more severe load at ambient temperature. As a result, the capacity evaluations of these zones were based upon loads at this temperature.

Analyses of riser loads included pressure, temperature, displacement of the pipeline, 100-year wave load condition with corresponding jacket displacement, wave slamming and slapping, and local wave.

The calculations showed that the major contributors to the pipeline loads were SRHM and temperature.

The riser calculations were sensitive to the modeled boundary conditions, especially the soil-support condition below the riser bend. The boundary conditions were based upon the latest inspection reports to reflect the as-built support conditions.

Capacities

Acceptance criteria for risers' capacities are the NPD-permissible stresses.

A limit-state approach has partly been adopted for the pipelines because some parts did not comply with the NPD-permissible stress criterion. Each failure mode was checked separately.

Because the failure modes are different for different parts of the pipelines, only the relevant ones have been checked.

The following have been considered:

• Yielding, considered for tension zone only

• Bursting, that is, reduced hoop stress capacity due to corrosion

• Leakage, not considered a structural failure but a pitting corrosion criterion

• Global buckling, that is, upheaval buckling, considered for the compressive zone only

• Local buckling, considered for the compressive zone only

• Ovalization, compression zone only

• Accumulated strain

• Fracture, unstable fracture, and plastic collapse (UFPC). Fatigue was not considered to be relevant.

The tensile zone extends approximately from 1 to 2.5 km from the Ekofisk center and 5-10 km further outward.

There are three major failure modes for the pipelines in this area: leakage as a result of pitting, bursting from extensive longitudinal corrosion (gas pipelines only), and fracture as a result of weld defects in combination with preferential girth-weld corrosion.

Yielding

The calculated tensile stresses in the pipelines did not exceed the NPD-permissible utilization factor of 72% of the pipeline material yield stress (SMYS = specified minimum yield stress). This criterion was met for all pipelines when corrosion-free.

In addition, the axial stress has been checked for the same criterion for corroded sections.

Bursting

Actually a yield criterion or stress criterion for hoop-stress capacity, bursting has been considered separately, however, and corresponds to the reduced hoop stress criterion as a result of corrosion.

The ASME B31.G capacity formula has been applied as a conservative estimate since pipeline tensile stresses were considered favorable

Accumulated strain

The criterion ensures that material weakening does not become too great, especially on the toughness and ductile properties.

Material testing has been performed to document that the pipeline materials are acceptable up to 5% strain.

UFPC

If the pipeline is exposed to tensile stress, existing cracks may become unstable. The cracks may start to grow rapidly and result in brittle fracture or plastic collapse of the material.

The allowable crack size in a girth weld can be determined based on BS PD 6493 Levels 1, 2, or 3. The criteria for the Ekofisk pipelines were based upon Level 3.

The most detrimental scenario foreseen for an unstable fracture and plastic collapse (UFPC) failure would be girth-weld corrosion combined with a crack.

Even though the likelihood of the modeled event is small for one weld, it may occur because a long part of the pipeline is exposed to uniform tensile load in the order of 200-300 circumferential welds per pipeline (3-5 km).

A review of literature and of old welding specifications led to the conclusion that the most probable weld defect which could have passed non-destructive evaluation (NDE) was a lack-of-fusion defect of 4 mm height (one run) and 150 mm long.

The corresponding capacity for one pipeline is given in Fig. 7 [11876 bytes]. The UFPC generally only allows an axial stress in the order of 100-150 MPa.

Pipeline compressive capacity

The global analyses of the pipelines have for some of them shown that the axial strain will exceed the yield limit over a relatively long section.

Large axial compression strain may lead to local buckling of the pipe wall, even if the pipeline experiences no global buckling.

Local buckling

Local buckling is a failure mode in which the pipe wall fails because of large axial strain, ovalization, external pressure, or combination of bending moment and external pressure. Buckling covers many types of failure modes; all were checked.

Many formulas for buckling have been published and are also given in various design codes and regulations. Practice varies with regard to the use of these formulas, and therefore safety levels vary greatly.

The pure axial compression strain capacity mainly depends on the OD/W.T. (D/t) ratio and the material strength. The capacity was established according to the classification of the pipes into cross-sectional classes as per NS 3472.

The in-house experience of DNV in a joint industry study (JIP) on high-strength steel led to the conclusion that an NS 3472 cross section Class 1 corresponds to D/t <0.056e/smys and can take four times the yield strain.

A cross section of Class 3 corresponds to D/t 0.112E/SMYS and can take only one times the yield strain. Linear interpolation was assumed between Class 1 and Class 3. Internal pressure is considered to be favorable and was not included in the assessment.

Upheaval buckling

Global buckling may occur for axially and laterally restrained pipelines exposed to large axial compressive force. The pipeline may have small out of straightness along the route where lateral and/or vertical buckling may be initiated.

A traditional installation method, as for the Ekofisk pipelines, will leave the pipeline fairly straight laterally but deviating vertically. The lateral soil resistance is also normally larger than the upward soil resistance.

Vertical buckling (upheaval buckling) is therefore the most probable global buckling mode for the Ekofisk pipelines. The pipelines which are exposed to compressive axial forces have been evaluated to establish the susceptibility for upheaval buckling.

If upheaval buckling occurs for a pipeline, it will lift vertically and/or laterally to release the axial compression force. When the axial force is sufficiently released, the pipe will find a new equilibrium.

The postbuckling mode may not necessarily be critical for the pipeline provided acceptable stress or strain limits are not exceeded.

Permissible strain levels have been established for the relevant pipelines.

Table 7 [30699 bytes], which shows typical calculation results, does not include unstable fracture/plastic collapse as a limiting failure mode because the probability of having a weld with a relatively large defect at 12 o'clock of an upheaval buckle is small.

Accumulated strain, bursting

The global pipeline analysis indicated that several of the pipelines will exceed the yield limit for 12 m subsidence. If the limit is exceeded, the steel will accumulate plastic strain.

Extensive accumulated plastic strain may have detrimental effects on the steel material ductile properties which may reduce the pipeline's capacity to sustain other loads.

The pipeline material testing has shown that modern, fully killed steels (with low nitrogen content and fine-grain treated with nitride-forming elements such as Al) may very well sustain higher values of plastic strain than the DNV-81 pipeline code standard limit of 2%. This would come without significant degradation of the mechanical properties.

The test results led to the conclusion that the critical accumulated plastic strain can be set to 5%.

Evaluation of the capacity of pipelines with corrosion damages in zones exposed to subsidence-induced compression loads indicated that the normally used ANSI/ASME 31G code could not be used. This code is not valid if the pipeline is exposed to large compressive forces.

Formulas had to be developed for the new evaluation method and design. This was done as a part of the JIP "Residual Strength of Corroded and Dented Pipes" performed by DNV and in which Phillips Norway participated.

Riser capacity, results

The NPD-permissible stress criterion has been used as an acceptance criterion for the risers.

Several of the investigated risers will be over-stressed because of the SRHM-induced stresses. The riser inspections performed, however, indicate in most cases the pipeline movement to be less than indicated by the global pipeline calculations.

The main corrective action will be to loosen the lower riser guides to increase the riser flexibility and thereby reduce the induced stresses due the SRHM. The riser guides will be locked in the new position.

What was learned

The main conclusions from the requalification assessments are:

1. Design lifetime can be extended for all the investigated pipelines.

2. The compression zones of the gas and two-phase pipelines to be rerouted were not recommended for reuse.

3. The tension zones of the gas and two-phase pipelines to be rerouted can be reused.

4. The tension and compression zones of the rerouted oil pipelines may be reused.

5. The pipeline materials remain ductile and tough up to 5% strain.

6. Upheaval buckling is not a failure mode for a pipeline, provided the post-buckling stresses or strains are within limits which the present requalification assessment has established.

7. "Virgin" soil properties cannot be used for pipelines in naturally backfilled trenches.

8. Bases for proper planning of the future inspection program for the pipeline systems have been established through the requalification assessments.

9. State-of-the-art methods for assessment of corrosion defects in pipelines exposed to external compression forces had to be used.

Acknowledgments

For permission to publish this article, the authors thank Norpipe A/S, Phillips Petroleum Co. Norway, and the coventurers: Fina Exploration Norway S.C.A, Norsk Agip A/S, Elf Petroleum Norge A/S, Norsk Hydro A/S, Total Norge A/S, Den norske stats oljeselskap a.s. (Statoil), Norminol, and Det norske Veritas.

References

1. Collberg, L., Cramer, E., "Re-qualification of pipeline systems-Does this relate to the past or to the future?" (Norwegian; translation available upon request) presented at the Pipeline Conference in Trondheim, Nov. 2-3, 1995.

2. Henderson, Paul A., Total Oil Marine, "Re-examining pipeline pigging in light of the North Sea safety record; is the cost burden too high?" presented at Offshore Pipelines Design, Construction & Operation Conference, Aberdeen, November 1994.

3. Festen, L.M.M., KSLA and SIPM, The Netherlands, "Intelligent Pigging Development For Metal Loss And Crack Detection," Concawe Pipelines Integrity Management Seminar, Brussels, October 1994.

Based on a presentation to the Offshore Technology Conference, Houston, May 6-9.

The Authors

Tor G. Tangeland is senior pipeline engineer with Phillips Petroleum Co. Norway. Before joining Phillips, he was discipline leader, structural, in the engineering departments of Elf Aquitaine Norge A/S and Aker Offshore Contacting A/S. He joined Phillips in 1990 in the regulatory engineering verification unit.
In 1993, he joined the pipeline and subsea engineering unit. Tangeland was educated at Stavanger College of Engineering as a mechanical engineer (1974) and since 1991 has been a Feani registered European engineer.

Leif Collberg is currently responsible for pipeline research work in Det norske Veritas and for the ongoing update for the Rules for Submarine Pipeline Systems (DNV '96). He joined DNV in 1982 and holds a master of science (1982) from Link"ping, Sweden.

Copyright 1996 Oil & Gas Journal. All Rights Reserved.