STATOIL STUDY CONFIRMS ADVANCED DESIGN FOR CONDENSATE PIPELINE

Jan. 1, 1990
F. Blaker Statoil Oslo In a study of the feasibility of installing a condensate pipeline over uneven seabed offshore Norway, Statoil has concluded that the use of advanced design criteria can give cost savings in the order of 30% compared with a design approach based on the permissible stress criterion and use of simplified calculation procedures. The cost saving is strictly related to intervention work which in traditional studies takes place over the total length of the route across irregular

R. Bruschi
Snamprogetti
Fano, Italy
F. Blaker
Statoil
Oslo

In a study of the feasibility of installing a condensate pipeline over uneven seabed offshore Norway, Statoil has concluded that the use of advanced design criteria can give cost savings in the order of 30% compared with a design approach based on the permissible stress criterion and use of simplified calculation procedures.

The cost saving is strictly related to intervention work which in traditional studies takes place over the total length of the route across irregular stretches.

The study, part of the Troll Phase I project development plans, investigated the possibility of installing a condensate pipeline through the deep waters of the Fensfjord to bring the condensate ashore at the Mongstad terminal, northwest of Bergen, Norway.

A description of the Troll-Mongstad condensate-pipeline concept, particularly the outer Fensfjord route section, leads to review of the implications of introducing advanced design criteria into a project in which the sealine must face extremely irregular seabed topography in deep waters (maximum 540 m).

OIL COMPANY EFFORTS

During the 1970s, considerable efforts were made by oil companies and research institutions to produce design criteria applicable to sealines in deep waters.

As a result, the main findings of the research were synthesized, and guidelines and recommended procedures were prepared for use in defining the minimum requirements for a submarine pipeline in deep water.

An excellent example of this work can be found in the rules issued by Det norske Veritas from which one can determine those fields which at that time were of major interest, particularly with respect to structural material problems during sealine installation. 1 2

At the same time, computer programs were developed to perform analyses of the complex structural behaviors necessary to satisfy the design requirements of these projects which were at that time opening new frontiers.

In particular, the subject of free-span analysis related to the crossing of uneven seabed areas became a major item in the design process.

This was due to its implications on the technical-economical feasibility of a crossing, which was strictly bound to the amount and quality of seabed preparation works and interventions necessary on the laid line in order to guarantee safe operation for the duration of its design life.3 In this context, the main aspects of free-span analysis gave rise to significant developments, mainly addressed to the following:

  • Experimental data regarding the behavior of suspended lengths of pipe under hydrodynamic loads in the near seabed scenario.4

  • Tools for decision making concerning methodology and extent of intervention works at various design stages, i.e., during design, after laying, and before and after commissioning.5

  • Review of guidelines currently adopted for the assessment of suspended lengths of pipeline under environmental loads and accidental loads.6

At present, the promising results from advanced studies of pipeline crossing in uneven seabed areas are viewed with great interest from offshore pipeline designers and operators.

REDUCING INTERVENTION

The introduction of design criteria based on the most recent results from research shows that there is a significant potential for reducing the quality and the extent of required intervention works and, as a consequence, for reducing related Costs.7

Furthermore, some recent works show the interest in the application of limit state and reliability-analysis methods to subsea pipeline engineering, aiming at establishing more consistent levels of safety and thus leading to more cost effective design.8

The experience gained during the compilation of the rules of the Design of Offshore Structures from Det norske Veritas 9 showed the role of reliability analysis in establishing the appropriate levels of safety according to statistical data mainly related to environment, loading conditions, and material properties.

For pipelines crossing irregular seabeds, loading conditions are mainly related to uneven profile and relevant exposure to on-bottom hydrodynamic field.

Regarding the Statoil study, some results presented here concern the following:

  • The implications of exceeding the permissible stress, as per DnV rules 2 and appealing to the elasto-plastic resources of the pipe section, thus referring to a permissible strain; and,

  • The implications of using advanced computer programs based on recent research projects regarding the free-span response to the hydrodynamic excitation.

The term "advanced criteria" is mainly related to pipe-line resting on the uneven seabed, from as-laid to operation (Fig. 1).

CONDENSATE-PIPELINE CONCEPTS

The giant Troll field, located in Blocks 31/2, 31/3, 31/5, and 31/6 in the Norwegian North Sea sector contains 1,200-1,300 billion cu m of natural gas. Approximately 70% of the gas is found in the eastern province, which is currently being developed by Norske Shell as the first out of several phases of the Troll field development.

The Troll Phase I unit partners are A/S Norske Shell, Norsk Hydro, Saga Petroleum, Statoil, Elf Aquitaine, and Total Marine Norsk.

The Troll Phase I project involves gas production, processing, and compression of 22.5 billion cu m/year. One of the field development prospects is based on an integrated platform with a two-stage condensate stabilizing process.

The condensate, which in this context should be considered a by-product with a peak production of 15,000 b/d, will be transported ashore via a pipeline, either to the Mongstad terminal or to Sture where Norsk Hydro has built an oil terminal.

Statoil, responsible for the engineering and construction of the condensate pipeline and later operator for the Troll Phase I platform, has undertaken several field surveys and technical studies to identify alternative pipeline routes, their complexity, and associated costs.

In order to accommodate both the Troll Phase I condensate capacity and condensate and oil production in the later Troll phases, a total transport capacity of 80,000 b/d has been considered. For this purpose, a 12-in. pipeline will have to be installed.

One of the main work tasks has been the examination of installing the 12-in. pipeline through the extremely irregular terrain at the outer Fensfjord area and further into the very deep waters (540 m, maximum) towards the Mongstad terminal.

The pipeline, which will be approximately 79 km long, will for the first 50 km from Troll pass over even seabed consisting of a relatively thick layer of soft clay. At the Fensfjord threshold, a moraine ridge of very irregular nature has created steep slopes in both the axial and the transversal direction.

At the deepest section of the fjord the seabed is relatively even. However, a narrow corridor with heavy marine activity will require extremely detailed planning and supervision of the pipelaying operation.

Close to landfall at Hageskjaeret, the pipeline will be routed in steep slopes (maximum inclination 1:3) which again will require very accurate positioning of the pipe. Finally, a 4.5 km land pipeline section will take the condensate from Hageskjaeret to the Mongstad terminal.

Fig. 2 shows the sequence of rocky outcrops and deep depressions along the deepwater stretch.

ALTERNATIVES

In order to identify the route alignment in this area, four alternatives to the main line based on extensive and reported surveys are available. These represent an intermediate solution to be further improved through a thorough route analysis.

The premises for the study are:

  • Basic design data including soil characteristics, environmental data, pipe and process data, intervention topology (gravel sleepers).

  • Pipeline design under installation and operating conditions, performed according to References 2 and 10.

  • The lay barge taken as reference is Saipem Castoro Sei, where the relevant performance parameters are taken from Reference 11.

The main problem areas being investigated are:

  • Free spanning pipeline, The unevenness scale compared with the selected pipe stiffness gives rise to long sections of the line under suspension and consequently under exposure to on-bottom currents.

  • Laying in restricted waters under dynamic positioning. The distance between facing shores in the narrowest stretch of the route is such that conventional mooring of a lay barge will be cumbersome and may seriously interrupt regular ship traffic to the refinery.

To look thoroughly into the technical and economical feasibility of such a difficult crossing, a gradual approach has been found necessary involving two stages.

In the first stage, reference is made to a preliminary routing to give a technical-economical figure mainly bound to free-spanning pipeline and relevant remedial measures.

In the second stage, routing is refined and advanced criteria and calculation procedures are applied to quantify the possible improvement of cost figures according to both more realistic design procedures and rationally reduced margins of safety.

It is this second stage which the present article addresses in detail.

Two further steps have been proposed in the event of positive results from Stages 1 and 2:

In a third stage, a detailed route analysis will be performed in parallel with an oceanographic survey to gather environmental data at the most critical stretches of the route in the Fensfjord area.

In a fourth stage, design criteria will be reviewed in detail; multispan fatigue analysis will be performed based on more refined and specific environmental data; and laying simulation in restricted waters under dynamic positioning will be carried out.

ADVANCED DESIGN CRITERIA

As a result of the current trend of offshore pipeline technology, a pipeline crossing uneven seabed areas should be allowed to exceed the permissible value of stress resulting from combined bending moment, differential pressure, and axial force loads, when certain conditions are met.

These conditions are a final, controlled strain and shall be below a permissible value; the pipeline configuration shall be almost frozen on the seabed; and occurrence of the most severe oscillations induced by hydroelastic instabilities shall be avoided.

In this context, it necessary to define strain-based limit states. These are the following:

  • A serviceability limit state which is based on the feasibility of inspection pigging procedures, thus referring to a maximum level of ovalization achieved by the pipeline under load combination .2

  • An ultimate limit state which is based on the failure mode of pipe section under combined loads from installation to commissioning.

The failure mode of the pipe section for typical diameter-to-thickness ratios, i.e., 20-60, changes from plastic collapse, where the section possesses sufficient rotation capacity to form a plastic mechanism, to classical buckling, where only limited curvature and rotation capacity are exhibited before a rapid degradation due to ovalization.'2

Indeed, strain-based criteria require a detailed analysis of the non-linear relationship between the bending moment and the curvature under combined load conditions which the pipe is subject to, including pipe material behavior and load history which play a major role when elasto-plastic resources of the pipe section are to be utilized.

In case of thicker pipes, i.e., diameter-to-thickness ratio less than 30, the pipe material where bending strains are in the plastic region is able to deform with no increase in stress, thus reducing the inclination of the pipe section to ovalize and allowing the fully plastic moment of the section to be achieved.

Failure resulting from ovalization is shifted to higher strains, which suggests thick pipes for applications where strain-based criteria are requested."

Moreover, to go beyond traditional criteria requires advanced calculation procedures for static-strength aspects regarding the identification of the state of strain the pipeline achieves when resting on the irregular seabed profile and for the dynamic response of the suspended lengths exposed to hydrodynamic loads.

NON-LINEAR EFFECTS

Static calculation procedures should take into account non-linear effects resulting from large displacements of the pipeline; nonlinear relationship between bending moment and curvature under differential pressure after laying, during water filling, hydrotest, and operating conditions; and nonlinear relationship between soil reaction and pipe embedment.5

Although the adequacy of the approach is not at the moment quantified by direct measurements of the strain, it is, however, qualitatively confirmed by the large amount of inspection records which currently are available.

Besides, new experimental results regarding the characterization of moment-curvature relationship under differential pressure for the analyzed diameter-to-thickness ratio are now available. This allows movement with certain confidence toward higher strains while still being safe against collapse. 15

Dynamic calculation procedures should take into account all the aspects which influence the calculation of natural frequencies.

These aspects 16 are realistic end conditions in accordance with span topology, either multiple or single, induced either by scouring or by unevenness; axial-flexural response coupling; static configuration, i.e., midspan sagging; and dynamic stiffening effects.

Moreover, experimental data relating to the response of the vortex shedding under steady current conditions and oscillatory flows induced by surface waves, available from the most recent research projects," should be included.

These data demonstrate the conservatism of current approaches for some aspects, mainly because experimental data, to which reference is made in current procedures, were obtained without taking into account the main characteristics of the near sea-bottom scenario, e.g., wall proximity, seafloor profile, free stream turbulence, etc.

STUDY STRATEGY

Implementation of the study is according to the following:

  • A reference design case which uses the most refined computerized procedures, applying traditional design criteria

  • Application of advanced design criteria to compare the resulting cost for the intervention works with that from the traditional case

  • A sensitivity analysis performed to achieve more-refined knowledge on the influence of the main design variables on the amount of intervention works and on the margin of safety which is related to the adopted criteria.

Mainly, diameter-to-thickness ratio was investigated.

Criteria identified for the study are shown in the accompanying box.

WALL THICKNESSES

The investigation included different wall thicknesses.

This was done to provide a measure of strain at collapse vs. an assumed permissible strain of 1% (W.T. 9.5 mm - minimum requirement for differential pressure; W.T. 14.3 mm and W.T. 15.9 mm - strain at collapse far from conceived permissible strain) and to identify the appropriate curve bending moment vs. curvature under various combined load conditions (differential pressure, axial load).

Fig. 3 shows the non-linear relationship for different wall thicknesses including the DnV '81 limitations for buckling.

In comparisons of improvement of bending moment at collapse for increasing thicknesses with related improvement of strain, it appears that a better exploitation of elasto-plastic resources of the pipe section, while being safe against collapse, requires higher thicknesses, i.e., diameter-to-thickness ratios in the order to 20-25.

FENSFJORD SITE

Feasibility of laying the reference pipeline in Fensfjord was investigated in order to identify possible implications of deepwater laying on extremely irregular and steep seabed morphology, of pipe laying in a narrow fjord using dynamic positioning, of laying uncoated pipe, and of technical solutions for shore approach.

Moreover, layability of the thicker and heavier pipe was investigated in order to quantify benefits in final span pattern and state of stress for the pipeline on the uneven seabed after laying.

Fig. 4 shows the spanning pipeline during laying along the steep slope close to the landfall, for the case of the reference pipe (12.7 mm) and the thicker pipe (15.3 mm).

The higher strength and submerged weight of the thicker solution, in spite of increased stiffening, make the equilibrium configuration of the pipeline to have shorter free spans and higher margin of safety against collapse than in the case of the reference pipe. Use of thicker pipe over the peak and reference pipe made heavy by concrete coating over the depression should improve the case significantly.

PIPE BEHAVIOR

The pipe behavior on an uneven seabed was investigated.

This was to make a tentative selection among alternative routes on the basis of minimum intervention work requirements and to determine for the selected route the implication of static equilibrium configuration achieved by the pipeline on the uneven seabed after laying, modified during water filling, and hydrotested under operating conditions.

Also pipe behavior on an uneven seabed was investigated to establish the effect of a thicker pipe wall solution on intervention work requirements with specific reference to local effects at contact points on rock outcrops.

Figs. 5 and 6 show the equilibrium configuration of the pipeline along the worst section of the main line.

Fig. 5 shows a comparison between linear and non-linear analysis, while Fig. 6 shows the modification of state of strain for the various load conditions to which the pipeline is subject.

The most severe condition occurs when the pipe sections are flooded and rest on rocky outcrops.

Fig. 7 shows the equilibrium configuration of the pipeline on a steep slope close to landfall. It represents the modification of state of strain under various load conditions to which the pipeline is subject. The thicker pipe solution was applied in three different arrangements. It resulted in a significant reduction of critical points.

Moreover, analysis of local effects of concentrated reaction forces at contact points shows that these effects are negligible for the relevant pipe diameter and selected wall thicknesses.

From the performed analyses, it can be concluded that utilizing elasto-plastic resources of material at strains lower than 0.5% produces no substantial modification of technical results and relevant cost benefits.

FREE-SPAN BEHAVIOR

Free-span behavior under hydrodynamic excitation was investigated to determine maximum allowable suspension lengths for which resonant flow-induced oscillations are a remote hazard or their effects in terms of cumulative damage are limited during pipeline lifetime.

Furthermore, calculations were carried out to establish the minimum intervention work requirements based on both traditional and advanced design criteria.

The main results demonstrate that free-span assessment is strictly bound to environmental design data and to design parameters such as boundary conditions, damping ratio, and lock-in criteria.

Traditional design criteria give maximum allowable suspension lengths of about 20 m, related to in-line lock-in condition. When a realistic damping ratio of 3% induced by pipe-boundary, interaction at free-span shoulders is introduced, in-line lock-in condition disappears or gives rise to a negligible fatigue damage.

This represents the main point resulting from this analysis of in-line oscillation; the damping ratio for in-line oscillations is by far the main parameter, more important than lock-in criteria (first instability region included or not).

Suspension lengths, which are susceptible to vortex shedding cross flow oscillations, are longer than 1 00 m. These are significantly larger than the expected lengths because of the effect of dynamic stiffening in the vertical plane produced by mid-span sagging .16

These results refer to a structural model for isolated free spans.

INTERVENTION

The amount of intervention works due to the uneven seabed have been established according to both traditional and advanced design criteria.

Comparison of the results based on traditional (Combination No. 1) and advanced (Combination No. 2) design criteria shows that the saving of gravel for the advanced case is 73% of the overall volume of 28,750 cu m required for the traditional case. Combination No. 3 does not give any substantial differences.

Thicker pipe segments at contact points where over-stresses occur coupled with reference pipe segments for the suspended length possibly made heavy by concrete coating could be a suitable solution in the case where detailed information about route profile is available and the ability of the lay barge to make the pipeline follow a narrow corridor is guaranteed.

CONCLUSIONS

Advanced design criteria mainly regard the introduction of a permissible strain based on the non-linear moment-curvature relationship for the pipe concerned and the use of a damage-tolerant approach for free span assessment.

The use of advanced design criteria in new projects where sealines cross very uneven seabed areas can be appropriate if they are coupled with refined computer programs which allow a correct forecast of the state of strain which the pipeline may achieve when resting on the irregular seabed profile and of the related dynamic response of suspended lengths to on bottom currents.

The study for Troll-Mongstad condensate pipeline, mainly regarding the deepwater and irregular stretches of the pipeline route in the Fensfjord area and the landfall at Hageskjaeret, shows that the use of advanced design criteria can give rise to cost savings in the order of 30% compared with a design approach based on the permissible stress criterion and by using simplified calculation procedures.

The cost saving is strictly related to intervention works which in traditional studies characterize the total length of the route across irregular stretches.

REFERENCES

  1. "Rules for the Design, Construction, and Inspection of Submarine Pipelines and Risers," Det norske Veritas, 1976.

  2. "Rules for Submarine Pipeline System," Det norske Veritas, 1981.

  3. Berti, A., Bruschi, R., and Mattelli, R., "Some Aspects of Technology Relating to Submarine Pipeline Crossing of Uneven Seabed Access," in Case Histories in Offshore Engineering; ad. G. Maier, Springer Verlag, 1985.

  4. Bruschi, R., et al., "Submarine Pipeline Design Against Hydroelastic Oscillations," Proceedings of the 7th Int. Conf. on OMAE, Vol. 4, 1988.

  5. Bruschi, R., et al., "Deep Water Pipeline Design," OTC Paper No. 4235,1982.

  6. Raven, P.W.J., " The Development of Guidelines for the Assessment of Submarine Pipeline Spans-Overall Summary Report," Dept. of Energy-UK, 1986.

  7. Damsgaard, A., and Hermansen, B., "Cost Impact of Pipeline Free Span Design Criteria on Haltenbanken Oil Transportation System," Offshore Pipeline Technology, 1988.

  8. Ellinas, C.P., and Williams, K.A.S., "Reliability of Engineering Techniques in Subsea Pipeline Design," Proceedings of the 8th Int. Conf. on OMAE, Vol. Y, 1989.

  9. "Rules for the Design of Offshore Structures," Det norske Veritas, 1977.

  10. Statoil, "Design Specifications for Offshore Installation of Offshore Pipeline Systems," Spc. F-SD-101, 1987.

  11. Saipem, "The Pipeline System of Castoro Sei," Internal Document, 1981.

  12. Murphey, C.E., and Langner, C.G., "Ultimate Pipe Strength Under Bending, Collapse, and Fatigue," Proceedings of the 5th Int. Conf. on OMAE, 1986.

  13. Bruschi, R., "Methodologies for Adapting Submarine Pipelines to Very Uneven Seabeds, the MASPUS Project," JIRP, Snamprogetti, 1988.

  14. Bruschi, R., and Orselli, B., "Analysis of Implications of Exceeding the Permissible Stress and Appealing to Elasto-Plastic Resources of the Pipe Section," Internal Report, Snamprogetti, 1988.

  15. Kyriakides, S., and Yeh, M.K., "Factors Affecting Pipe Collapse," EMRL Report No. 85/1 for American Gas Association, 1985.

  16. Bruschi, R., and Vitali, L., "Large Amplitude Oscillations of Geometrically Non Linear Beam Subject to Hydrodynamic Excitation," Offshore Mechanics and Arctic Engineering Symposium, Houston, 1988, Additional Paper.

  17. Bryndum, M., et al., "Long Free Spans Exposed to Current and Waves-Model tests," OTC Paper No. 6153, 1989.

  18. Bruschi, R., et al., "Field tests with Pipeline Free Spans Exposed to Wave Flow and Steady Current," OTC Paper No. 6152, 1989.

Copyright 1990 Oil & Gas Journal. All Rights Reserved.