20102

PETRONAS TECHNICAL STANDARDS DESIGN AND ENGINEERING PRACTICE

REPORT (SM)

DEEP WATER PIPELINE, FLOWLINE AND RISER INSTALLATION SUMMARY REPORT

PTS 20.102 SEPTEMBER 1987

PREFACE

PETRONAS Technical Standards (PTS) publications reflect the views, at the time of publication, of PETRONAS OPUs/Divisions. They are based on the experience acquired during the involvement with the design, construction, operation and maintenance of processing units and facilities. Where appropriate they are based on, or reference is made to, national and international standards and codes of practice. The objective is to set the recommended standard for good technical practice to be applied by PETRONAS' OPUs in oil and gas production facilities, refineries, gas processing plants, chemical plants, marketing facilities or any other such facility, and thereby to achieve maximum technical and economic benefit from standardisation. The information set forth in these publications is provided to users for their consideration and decision to implement. This is of particular importance where PTS may not cover every requirement or diversity of condition at each locality. The system of PTS is expected to be sufficiently flexible to allow individual operating units to adapt the information set forth in PTS to their own environment and requirements. When Contractors or Manufacturers/Suppliers use PTS they shall be solely responsible for the quality of work and the attainment of the required design and engineering standards. In particular, for those requirements not specifically covered, the Principal will expect them to follow those design and engineering practices which will achieve the same level of integrity as reflected in the PTS. If in doubt, the Contractor or Manufacturer/Supplier shall, without detracting from his own responsibility, consult the Principal or its technical advisor. The right to use PTS rests with three categories of users : 1) 2) 3)

PETRONAS and its affiliates. Other parties who are authorised to use PTS subject to appropriate contractual arrangements. Contractors/subcontractors and Manufacturers/Suppliers under a contract with users referred to under 1) and 2) which requires that tenders for projects, materials supplied or - generally - work performed on behalf of the said users comply with the relevant standards.

Subject to any particular terms and conditions as may be set forth in specific agreements with users, PETRONAS disclaims any liability of whatsoever nature for any damage (including injury or death) suffered by any company or person whomsoever as a result of or in connection with the use, application or implementation of any PTS, combination of PTS or any part thereof. The benefit of this disclaimer shall inure in all respects to PETRONAS and/or any company affiliated to PETRONAS that may issue PTS or require the use of PTS. Without prejudice to any specific terms in respect of confidentiality under relevant contractual arrangements, PTS shall not, without the prior written consent of PETRONAS, be disclosed by users to any company or person whomsoever and the PTS shall be used exclusively for the purpose they have been provided to the user. They shall be returned after use, including any copies which shall only be made by users with the express prior written consent of PETRONAS. The copyright of PTS vests in PETRONAS. Users shall arrange for PTS to be held in safe custody and PETRONAS may at any time require information satisfactory to PETRONAS in order to ascertain how users implement this requirement.

CONTENTS LIST OF FIGURES LIST OF ABBREVIATIONS PREFACE ABSTRACT I.

INTRODUCTION

II.

PRIMARY CONCLUSIONS

III.

DETAILED CONCLUSIONS A. Design Aspects of Deepwater Pipelines and Flowlines B. Steal Catenary Risers C. Steep S Buoyancy Supported Risers D. Lateral Deflection of Pipelines and Flowlines E. Second End Flowline Pull-In to Satellite Wells F. Stab and Hingeover of Flowlines G. Final Alignment of Pipelines and Flowlines H. Repair Considerations

IV.

DESIGN ASPECTS OF DEEPWATER PIPELINES AND FLOWLINES

V.

SCALING RELATIONS A. Introduction B. Scaling Relations

VI.

STEEL CATENARY RISERS (SCR's) A. Introduction B. Dynamic Analysis C. Installation, Abandonment, and Recovery of Pipeline SCR's D. Installation and A&R of Flowline SCR's

VII.

STEEP S BUOYANCY SUPPORTED RISERS A. Introduction B. Installation C. Steep S Riser Interference

VIII.

LATERAL DEFLECTION OF PIPELINES AND FLOWLINES A. Introduction B. Small Scale Model Tests and Numerical Analyses C. Large Scale Wave Tank Tests and Numerical Analyses D. Procedures

IX.

SECOND END FLOWLINE PULL-INS A. Introduction B. Small Scale Modal Tests C. Numerical Analyses D. Land Tests

X.

STAB AND HINGEOVER CONNECTIONS A. Introduction B. Wellhead Overload Protection and Flowline Latching Mechanisms C. Stab-in Tests and Analyses

XI.

SPECIAL SUBJECTS A. Final Alignment B. Repair Considerations C. Load Limiting Devices D. Bundling of Flowlines E. Buoyancy Considerations and Submerged Weight Control F. Concepts for Very Deep Water

XII.

RECOMMENDATIONS FOR FURTHER STUDY REFERENCES

FIGURES Figure 1. Figure 2. Figure 3. Figure 4. Figure 5. Figure 6. Figure 7. Figure 8. Figure 9. Figure 10. Figure 11. Figure 12. Figure 13. Figure 14. Figure 15. Figure 16. Figure 17. Figure 18. Figure 19. Figure 20. Figure 21. Figure 22. Figure 23. Figure 24. Figure 25. Figure 26. Figure 27.

Riser Concepts for Floating Production System General Layout of Floating Production System (FPS) with Modal Risers in the Wave Tank Lateral Deflection Operation Lateral Deflection Tests Second End Flowline Pull-in Operation Small Scale (1:100) Modal Table Tests of Second End Flowline Pull-in Stab and Hingeover Operation Large Scale (1:10) Modal Tests of Stab and Hingeover Shop Tests with Bullnose Shop Tests with Two-Pin Sled Deepwater and Shallow Water Design Comparisons Steal Catenary Riser Analysis Steal Catenary Riser Spectral Analysis SCR Abandonment and Recovery by S-Curve Layback and Recovery Installation of Flowline SCR from a Central Subsea Well Riser Interference During FPS Movements Off-Bottom Lateral Deflection of 20-inch Pipeline On-Bottom Lateral Deflection of 12.75-inch Pipeline Nondimensional Curves to Evaluate Lateral Deflection Configurations Second End Flowline Pull-in Comparison of Modal Table Tests and Numerical Analysis Nondimensional Curves to Evaluate Flowline Pull-in Parameters Different Bottom Latch Options for Stab and Hingeover Tool Bullnose Comparisons Two-Pin Sled Alignment Rotational Alignment of Two-Pin Sled with Large Guidecones Repair During Installation Repair to Operating Line

ABBREVIATIONS

DPFRI

Deepwater Pipeline, Flowlines, and Riser Installation

FLB

Flowline Bundle

FPS

Floating Production System (Semi-Submersible)

GOM

Gulf of Mexico

PC

Personal Computer

RAO

Response Amplitude Operator

ROV

Remote Operated Vehicle

SAMI

Southwest Applied Mechanics, Inc.

SCR

Steel Catenary Riser

SMYS

Specified Minimum Yield Stress

SSBR

Steep S Buoyancy Supported Riser

TLP

Tension Leg Platform

TRED

Transportation Research and Engineering Department

WRC

Westhollow Research Center

PREFACE This is an overview of the work reported in the six sections which comprise the technical elements of the program. Following the Introduction and Conclusions is a more detailed description of the subjects covered in the volumes of reports. The reader should consult these individual reports for more complete information. The reports from this program are contained in the following sections:

SECTION

TITLE

AUTHORS

10

Summary Report

F. Kopp, C. G. Langner, R. H. Orlean, and R. W. Patterson

20

Scaling Relations

F. Kopp

30

Risers at Floating Production Systems

40

Lateral Deflection of Pipelines and Flowlines

F. Kopp, C. G. Langner, and R. W. Patterson F. Kopp

50

Second-End Flowline Pull-ins

C. G. Langner, R. W. Patterson, and C. D. Michalopoulos

60

Stab and Hingeover of Flowlines

C. G. Langner

70

Special Subjects

71

Final Alignment of Pipelines and Flowlines

F. Kopp

72

Bundling of Flowlines

C. G. Langner, R. W. Patterson

73

Repair Considerations

R. H. Orlean

74

Buoyancy Considerations

F. Kopp

75

Load Limiting Devices

C. G. Langner, R. W. Patterson

76

Concepts for Very Deep Water

C. G. Langner, S. Huffer, and D. W. Mc Millan

ABSTRACT A fifteen-month research and development study has been conducted to technically evaluate first-end and second-end methods for connecting (a)

deepwater pipelines and flowlines to subsea structures and

(b)

catenary risers to floating production system. Evaluations were based on physical model simulations of installation methods at Shell's deepwater test facilities, augmented by numerical analyses. Model scales varied between 1:400 and 1:4.

The model tests were conducted to assess potential problem areas and limitations associated with various tie-in methods, to provide solutions for these problem, and to compare installation configurations with numerical analyses. The numerical analyses provided estimates of installation forces, pipe configurations, and pipe stresses for each of the evaluated methods. Step-by-step descriptions of tie-in operations and recommendations for major equipment performance requirements were developed. All tie-in methods considered in this study are technically feasible without the need for long lead time development of new technology and most methods are compatible with existing vendor hardware. The choice of a particular method and procedure will often involve a trade-off between either a simple procedure, which may cause the pipe to be subject to fairly high bending strains during installation, or a more complicated procedure which involves keeping pipe bending strains well below yield. In the deepwater environment the choice of the simpler procedure will most likely be the least expensive and risky one, if there is an adequate safety factor against pipe buckling. Most deepwater tie-ins of flowlines and pipelines, as well as installations of steel catenary risers will require a vertical lay method for at least the final deepwater section of the line, using a dynamically positioned lay vessel. The results of dynamic analyses and fatigue calculations indicate that steel catenary risers suspended from semisubmersible type drilling and production vessels in deep water (greater than 1500 ft) regions of the Gulf of Mexico are technically viable and merit further consideration and study. Installation methods have been developed as well as abandonment and recovery methods. These risers are suitable for both trunklines as well as flowline bundles.

Deepwater Pipeline, Flowline, and Riser Installation: Summary Report An Industry Program to Evaluate First-End and Second-End Methods for Connecting Deepwater Pipelines and Flowlines to Subsea Structures, and Catenary Risers to Floating Production Systems by

F. Kopp, C. G. Langner, R. H. Orlean, R. W. Patterson

I.

INTRODUCTION

Pipeline, flowlines, and riser installations in deep water will require new diverless installation technology. Of major concern are the tie-ins, or connections. A need was identified for an overall evaluation of the technical feasibility and practicality of various vendor- and oil company-proposed methods of deepwater pipeline, flowline, and riser installations. The primary objectives of such an evaluation would be to objectively identify potential problem areas associated with the use of a given installation method, offer comparisons of risk with alternative methods, and provide specific recommendations for improvements in methods and equipment which may take advantage of some of the characteristics of the deepwater environment. To achieve these objectives the Transportation Research and Engineering Department of Shell has conducted a 15-month-long industry sponsored model and simulation test program to technically evaluate first- and second-end methods for connecting (a) (b)

deepwater pipelines and flowlines to subsea structures; and catenary risers to a floating production system. Twenty-four companies joined Shell in sharing the cost of this program. The main value of this program is that its results provide the participants with an objective technical basis to evaluate their specific pipeline/flowline installation proposals, plans or projects. Model simulation tests, augmented by numerical analyses, pinpointed critical problems during installation operations and helped to provide practical solutions, which should avoid costly full-scale problems in the field.

The approach followed for each of the methods considered was to: (1) (2) (3)

Conduct small and/or large scale model tests to assess potential problem areas and provide solutions to these problems or recommendations to minimize them. Use model tests combined with numerical analyses to make estimates of installation forces, pipe configurations, and pipe stresses during installation. Use results of these tests and analyses to prepare step-by-step descriptions of tie-in operations and recommend major equipment performance requirements.

The deepwater range for this program is 1,000 ft to 3,000 ft, but a number of tie-in methods evaluated in this program could also be used for deeper water. Generally, production in water depths over 1,000 ft requires diverless pipeline and flowline tie-in techniques, and the deepwater environment will make it difficult to merely extrapolate the experience gained in making shallow water tie-ins. Deepwater also provides areas of opportunities, however, to take advantage of certain characteristics of pipe mechanical behavior. Two of these characteristics stand out. First, in the vertical direction, the pipe is more flexible than a comparable pipe in shallow water. Second, thick-wall pipe which must be used in deep water to resist collapse offers the possibility of accepting greater bending strains than allowable for thin-walled pipe in shallow water. These two areas were prime determinants in selecting tie-in methods which appeared suitable for evaluation in this program. Two vertical tie-in methods were identified as promising because they take advantage of the greater pipe flexibility in deep water. They are steel catenary riser tie-ins to floaters and flowline tie-in to subsea wells by stab and hingeover. On- or off-bottom flowline or pipeline pull-in methods were identified as methods which could also benefit from the possibility of accepting greater bending strains during deepwater installation than commonly accepted in existing shallow water applications. These pull-ins are categorized into first-end axial pull-ins, and first- or second-end lateral pulls or deflections.

In addition, a hybrid vertical/catenary riser concept was developed, namely the steep S buoyancy supported riser which is made of composite pipe. This concept minimizes the need for other vessels than the floating production vessel itself for riser installation, production, and maintenance from wells drilled underneath the vessel. The following summarizes the types of model tests and numerical analyses conducted for the different tiein methods considered. Program sponsors receive a ½-inch VHS Video tape with a summary of all model tests.

MODEL TESTS (SCALE)

DESCRIPTION

Pipeline and Flowline Steel Catenary Riser Installation, Operation, Abandonment and Recovery

Wave Tank Tests (1:100 to 1:400)

Steep S Buoyancy Supported Riser Installation and Operation

Wave Tank Tests (1:100)

Lateral Deflection of Pipelines and Flowlines

NUMERICAL ANALYSIS

Finite Element Difference

and

Finite

Model Table and Wave Tank Test (1:12 to 1:120)

Finite Element Difference

and

Finite

Second End Flowline Pull-in of Flowlines to Satellite Wells

Model Table, Wave Tank, and Land Tests (1:4 to 1:120)

Finite Element

Axial Pull-in and Lay-away from Subsea Structure

Wave Tank Test (1:120)

Stab and Hingeover of Flowlines

Wave Tank and Deepwater Tank Test (1:12 to 1:120)

Final Alignment of Pipelines and Flowlines

Shop and Wave Tank Tests (1:10 to 1:16)

In addition, several other areas of interest were evaluated, such as: (1)

repair options;

(2)

load limiting devices;

(3)

flowline bundling methods; and

(4)

buoyancy control of pipelines and flowlines.

Analytical, Finite Difference, and Finite Element

II.

PRIMARY CONCLUSIONS (1)

All deepwater tie-in methods considered in this study are technically feasible without the need for long lead time development of new technology and most are compatible with existing vendor hardware. The choice of a particular method and procedure will often involve a trade-off between either a simple procedure which may cause fairly high pipe bending strains during installation, or a more complicated procedure which involves keeping pipe bending strains well below yield. In the deepwater environment, the choice of the simpler method generally will be the least expensive and least risky one, provided that there is an adequate safety factor against pipe buckling.

(2)

Most deepwater tie-ins of flowlines, flowline bundles, and pipelines, as well as installations of steel catenary risers will require the use of a vertical lay method for at least the deepwater section of the line, using a dynamically positioned lay vessel.

(3)

Dynamic analyses and fatigue calculations have been performed for steel catenary risers. These studies show that steel catenary risers suspended from semi-submersible type drilling and production vessels in deep water (greater than 1500 ft) regions of the Gulf of Mexico or other areas with similar environmental conditions, are technically viable. They merit further consideration and study as an attractive alternative to composite pipe risers ("flexible risers"). Installation and abandonment and recovery methods have been developed.

(4)

Steep S buoyancy-supported vertical risers, made of composite pipe, which are run to wells underneath a floating production vessel, and installed and maintained by this vessel have a number of serious operational problems. The main problem is riser interference during vessel movements from over one well location to another. Much more development, including large scale testing, is needed to prove the viability of methods which can avoid this interference.

(5)

Lateral deflection of pipelines and flowlines to subsea structures is a viable technique for making primarily second-end connections. Shallow water techniques of supporting a section of the pipe off the sea bottom to minimize pull-in forces and using multiple pull points and pull cables, may not be practical in deep water because of their complexity. Instead, a single cable, on-bottom deflection, especially of small diameter pipelines with low submerged weights, should be considered. If a tow method is used to install the pipe, it may not be possible to place the pipe and in a target area as small as has been achieved in shallow water installations. Resulting pipe and misalignments at the subsea structure may, for large diameter, single pipelines, require the use of a connector which allows for some angular misalignment, before being locked.

(6)

A second-end flowline pull-in to a satellite well leaving the flowline or flowline bundle in a large U-shaped loop on the seafloor, is a viable technique which avoids some of the alignment problems associated with conventional lateral deflection techniques. In particular, the U-turn does not require the pipe end to be placed in such a small target area prior to pull-in.

(7)

The stab and hingeover method of connecting a flowline to a subsea structure appears to be a fast and simple alternative to a more conventional first-end pull-in. Methods have been developed to successfully stab and latch a flowline into a receptacle on a wellhead, without the need for surface heave compensation.

(8)

Final alignment forces can be an order of magnitude greater than the basic pull-in force (prior to alignment) . If alignment forces are expected to be high, the main pull-in cables should not be used to attempt final alignment. Some other more effective mechanical means could be used (hydraulic rams, for instance) or some flexibility could be built into the system. Problems associated with axial rotation of the pipe should not be underestimated, especially if a two-cable pull-in system is used.

(9)

An alternative to the more conventional method of flow-line bundling (by wrapping tape around the bundle) is the helical bundle. Configuring the individual pipes which constitute the bundle into a rope-like helix during onshore make-up, will give the pipe bundle the necessary flexibility and integrity to be bent onto a reel for offshore installation by the reel method. Reeling of a helical bundle also will be easier and preferable to reeling a flowline bundle inside a casing pipe.

(10)

Pipelines or flowlines damaged during installation by a surface-lay method may be repaired by "reverse-laying" to the damaged section or by cutting, lifting, and re-laying a portion of the line. The ability to recover and handle damaged lines should be designed into pipe laying vessels.

(11)

For deepwater lines that are damaged or fail during operation (after laying), the most practical repair methods appear to involve cutting the pipe and lifting both ends to the surface. These pipe ends may be joined by a flexible spool or swivel and laid down offset to one side, or mechanical connections may be welded onto the pipe ends, which are then laid down and connected on bottom. Tools for cutting, lifting, lowering, and connecting pipelines subsea, need further development.

(12)

A remote on-bottom repair system will be necessary for damaged pipelines which cannot be lifted to the surface, i.e., for deeply buried lines or for damage near a subsea crossing or tie-in. However, fully automated on-bottom repair of an offshore pipeline has not yet been demonstrated. If and when such remote on-bottom spool piece repair systems are successful, these systems may be very difficult and expensive to operate, especially in Gulf of Mexico conditions of soft soil and poor visibility.

III.

DETAILED CONCLUSIONS A.

DESIGN ASPECTS OF DEEPWATER PIPELINES AND FLOWLINES

(1)

During investigation of several of the tie-in scenarios it was found that a mere adoption of shallow water guidelines to determine allowable bending strains either made the tie-in very difficult to execute under less than ideal circumstances or required complex intervention methods to maintain bending strains below yield. Deepwater interventions such as subsea relocatable winches and pull cables are expensive, can add risk to the pipeline or flowline operations, and are incompatible with the need to keep deepwater operations as simple as possible. Relaxing the requirements for allowable bending strain during installation, however, often results in greatly simplified operations and reduced risk.

(2)

For most deepwater pipeline or flowline installation scenarios, the thicker wall pipe which must be used to resist collapse offers the possibility of allowing greater bending strains during installation than is normally allowed in comparable shallow water installations, and at a lower risk of mechanical failure.

B.

STEAL CATENARY RISERS (Figures 1,2) Kopp,Langner,Patterson (31,32,33)*

(1)

Results of dynamic analyses and fatigue calculations for the sagbend region indicate that steel catenary risers (SCR's) suspended from semisubmersibles in deep water (greater than about 1500 ft) in the Gulf of Mexico (GOM) or other regions with similar environmental conditions, are technically viable. They merit further consideration and study as an attractive alternative to composite pipe ("flexible pipe") catenary risers.

(2)

For the conditions studied, cyclic stresses resulting from first order, wave-induced motions of the semisubmersible will not cause fatigue failure in the sagbend region of SCR's.

(3)

SCR response to wave-induced vessel movement is nearly linear for stresses in the sagbend and for pipe tension at the surface. Thus spectral techniques can be used to determine the extreme response of both stress and tension and to predict riser fatigue life.

(4)

Installation and Abandonment and Recovery (A&R) methods have been developed for pipeline and flowline SCR's. All installation methods will require the use of specialized dynamically-positioned (DP) vessels equipped to maintain adequate horizontal tension in the pipe while handling the pipe at near vertical departure angles.

(5)

The most practical first- or second-end methods of installing an SCR at an FPS involve a vertical lay method. Other lay methods are technically feasible, but generally are more complicated and risky. Combination methods, however, may be cost effective. For example, a vertical lay method for the riser section including a short section of the on-bottom pipeline, can be combined with an S-lay method for the remainder of the line.

*Numbers in parentheses refer to like numbered Sections in the Volumes of reports (see Preface for a complete listing).

Figure 1. Riser Concepts for Floating Production System

Figure 2. General Layout of Floating Production System (FPS) with Model Risers in the Wave Tank

A. Plan View

B. View A-A

(6)

Temporary abandonment of an SCR by keelhauling the riser underneath the FPS is complicated and risky, and therefore is not recommended if other methods are available. Abandonment of an SCR by layback on the seafloor in a U-curve, followed by S-curve recovery, however, is straightforward.

(7)

A subsea anchor may be needed for installation of flowline SCR's from subsea wells drilled underneath the FPS. In case of multiple flowline SCR's, careful planning of their layout is required to avoid seabed congestion.

(8)

In addition to their suitability for connecting pipelines and flowlines to floaters, SCR's can also be used to connect pipelines to other deepwater structures such as Tension Leg Platforms (TLP's) or Compliant Towers. Riser installation at those structures will even be more straight forward in the absence of mooring lines.

C.

STEEP S BUOYANCY-SUPPORTED RISERS (Figures 1,2)

(1)

The concept of installing Steep S buoyancy-supported risers (SSBR's) made of composite pipe, to wells underneath an FPS is feasible. However, riser interference may occur during lateral movements or excursions of the FPS, as, for example, if the FPS moves directly over one of the wells for work-over or other well operations. Riser interference may be avoided by controlled movements of the FPS vessel, but larger scale model testing is required to verify this.

D.

LATERAL DEFLECTION OF PIPELINES AND FLOWLINES (Figures 3,4) Kopp (41,42,43,71)

(1)

Lateral deflection is a feasible deepwater tie-in method for pipelines and flowlines. Pipelines with a diameter greater than about 12.75 inches, or cased flowline bundles (FLB's), will in most cases require an off-bottom buoyancy supported section. Smaller pipelines and flowline bundles may simply be laid on bottom throughout the deflection and pull-in operations, provided that the bottom is sufficiently smooth, since neither the pull-in forces nor bending of the pipelines will be limiting factors.

(2)

The main difference between shallow water and deep water application of this method is a greater need for simplicity in operations as water depth increases. A single cable pull from one pull point is preferred over multiple cable pulls, for example. Similarly, if bending strains somewhat higher than yield are accepted, a simpler on-bottom deflection is preferred over a more complicated off-bottom deflection.

(3)

There is excellent agreement between pipe configurations measured during the model tests and those computed numerically, especially with regard to final pipe configuration. Model tests and numerical analyses also complement each other. It is much easier to investigate solutions for potential problem areas by conducting small scale model tests than by numerical analysis, but the numerical analysis is better suited to predict pull-in forces, alignment forces, and pipe bending stresses.

(4)

Final pipe configuration is predominantly determined by pipe laydown geometry (offset and overshoot), and is not very sensitive to changes in seabed friction or pull point location. Pipe stresses during deflection, however, are very much dependent on these parameters. If stresses are high enough to cause significant yielding of the pipe, misalignment problems during final alignment may result.

(5)

Wave tank tow tests showed that significant target overshoot or undershoot problems can arise if the pipe is towed. This is primarily caused by dynamic-frictional interaction (stickslip behavior) between pipe and seabed during stopping and restarting of the tow operations. Solutions to these problems, other than allowing for a large target area, are cumbersome in the deepwater environment.

Kopp (32)

Figure 3. Lateral Deflection Operation

Figure 4. Lateral Deflection Tests

A. View of Small Scale (1:120) Model Table Tests

B. View of Large Scale (1:12) Wave Tank Test

E.

SECOND-END FLOWLINE PULL-INS TO SATELLITE WELLS (Figures 5,6) Langner, Michalopoulos, Patterson (51,52)

(1)

Second-end flowline pull-in to a satellite well which leaves the flowline in a large U-shape loop on the seafloor is a viable alternative to the more commonly used lateral deflection techniques. This U-loop method alleviates some of the alignment problems associated with conventional lateral deflection techniques which require the pipe end to be placed in a small target area prior to deflection. Two different procedures were developed from small scale model testing.

(2)

Straight Laydown Position [Figure 5 (a)] If a flowline or flowline bundle (FLB) has been installed by a surface lay or a tow method, one method of connecting the flowline or FLB to a subsea well is as follows: terminate the flowline end at an "offset" distance (Y) of approximately three times the flowline characteristic length (C), and at an "overshoot" distance (X) of about ten times the flowline characteristic length past the wellhead. The flowline can then be pulled directly back towards the wellhead using a remotely operated pull-in tool with subsea winch. The tests showed that this procedure may result in considerable yielding of the flowline close to the flowline and, which in turn may cause alignment problems during the connection at the wellhead.

(3)

Curved Laydown Position [Figure 5 (b)] An alternative procedure was developed which avoids alignment problems due to plastic bends in the pipe too close to the pipe end. This procedure involves lifting the pipe end up vertically using a DP vessel while moving longitudinally to a position about half way to the wellhead, then maneuvering laterally a distance Y and toward the well a further distance Y to form a U-turn, and finally laying the pipe down directly in front of the wellhead. The pull-in is then completed by a short pull with a wire rope from the wellhead. Even with this improved procedure, bending strains in the flowline bundle may somewhat exceed strains normally associated with pipe yielding, but this generally does not affect pipe safety against buckling and the region of pipe yielding will be sufficiently far away from the pipe end that this will not cause significant pipe and misalignment.

(4)

The second-end flowline pull-in procedures described above were also analyzed using the finite element programs ABAQUS and SPAN. Results obtained with both programs agree well with the results of the model tests.

Figure 5. Second-End Flowline Pull-in Operation

Figure 6. Small Scale (1 : 100) Model Table Tests of Second-End Flowline Pull-in

F.

STAB AND HINGEOVER OF FLOWLINES (Figures 7,8)

(1)

Using 1:12 scale models, stab and hingeover connections were successfully demonstrated both in air and under water. These model-scale connections were made from heights of 15 to 75 feet, monitored only by closed-circuit TV, with the model flowline suspended and paid out from a moving platform which simulated typical ship motions. Loads applied to the model wellhead were found to be quite small, proved by a breakaway connection in the model flowline which did not break, even though no effort was made to heave-compensate the flowline at the upper end.

(2)

Langner (60,76)

These model tests also demonstrated the feasibility of a simple latch mechanism, activated by the hingeover operation, for locking the stab tool into the receptacle on the wellhead. A variety of potential latch mechanisms were designed, some of which are reversible.

G.

FINAL ALIGNMENT OF PIPELINES AND FLOWLINES (Figures 9,10)

Kopp (42,71)

(1)

Pipe alignment forces are predictable and some simplified design formulations have been developed to predict the resulting forces on typical pipe end fittings. Large scale prototype testing before deployment in the field is necessary, however, despite being able to predict these forces and moments. Testing is necessary, to avoid interferences and to insure proper mechanical engagement of parts. The ability to apply large alignment forces alone does not guarantee successful alignment and it is essential that potential misalignment of the pipe end be analyzed and tested in all six degrees of freedom (DOF). The most difficult DOF to analyze and predict is axial rotation of the pipe.

(2)

Final alignment forces can be an order of magnitude greater than the force required to bring the end of the pipeline or FLB adjacent to the alignment structure on the subsea template or wellhead. Due to large loads, use of pull cables to provide final alignment may not be practical Other mechanical means such as hydraulic rams, for example, may be more effective in applying the required forces. Alternatively, some flexibility may be built into the system, by using short segments of composite pipe or by using mechanical connectors which can be accept some angular misalignment before latching.

(3)

A two-pin sled can be used instead of a bullnose to reduce lateral alignment forces. If axial rotation is present in the pipe prior to final alignment, however, use of a two-pin sled may result in significant alignment problems. Although minor rotational misalignments can be corrected using just the assistance of the two pull-in cables and properly designed guidecones, other means of correcting rotational misalignments should be investigated if rotations are greater than about 15 deg.

(4)

Height and inclination of the alignment structure are critical parameters in the process of minimizing alignment forces and moments. Wellhead and template design must consider pipeline, flowline alignment problems at an early stage.

Figure 7. Stab and Hingeover Operation

Figure 8. Large scale (1 : 100) Model Tests of Stab and Hingeover (Lay Direction Reversed from Figure 7)

Figure 9. Shop Tests with Bullnose

A. View of Test Set-up

B. Detail of Bullnose and Receptacle

Figure 10. Shop Tests with Two-Pin Sled

A. Alignment without Pipe Rotation

B. Alignment with Pipe Rotation

H.

REPAIR CONSIDERATIONS

Langner, Orlean (73)

(1)

Pipelines or flowlines damaged during installation by S-Lay, J-Lay, or Reel method are most easily repaired by "reverse-laying" the pipe back onto the lay vessel to the point of failure. Pipelaying then resumes after cutting out the damaged section. Capability to recover and handle damaged lines should be designed into all deepwater pipelaying vessels.

(2)

For lines which cannot be repaired by reverse-laying (e.g., if completely fractured or if the damage is detected a long distance behind the lay vessel, etc.), a repair can be made by abandoning the pipe, moving to the damage location, cutting, dewatering, and lifting the line, and then re-laying a portion of the pipeline. The abandoned section of pipeline would be recovered later. This "repair method" has already been demonstrated in 2000-ft water depth, and is extendable to even greater depths without much additional development.

(3)

Damage incurred during installation by a tow method is best repaired by returning the tow string to shallow water. If possible, the pipeline(s) should be towed back to the launch site for repairs. Otherwise, the damaged line(s) may be abandoned in place and repaired by other vessels.

(4)

Failures in operating pipelines or flowlines may be repaired at the surface or at the seafloor, with the choice of repair method often dictated by the location and burial depth of the line. Surface repairs cannot easily be made if the pipeline is deeply buried or if the line is damaged near a subsea tie-in. On-bottom repairs may not be possible depending on the water depth and on the capabilities of remotely operated repair systems available at the time.

(5)

The repair method requiring the least development appears to be the surface-lift method, which involves cutting the pipeline at the damaged location and lifting both ends to the surface. These pipe ends are joined by a swivel or flexible spool and then are carefully laid down with an offset to one side. Alternatively, mechanical connections may be welded to the pipe ends at the surface, which are then laid down and connected on bottom. These surface-lift repair methods preferably would be performed from a single vessel having lifting, pipelay, and remote tool handling capabilities.

(6)

Repairs to lines damaged within one water depth of their termination would require retermination unless repaired entirely on bottom. Prudent design of termination points should include provisions for future terminations, and should require minimal special equipment.

(7)

Fully automated on-bottom repair systems still require a great deal of development. Although much work has been performed in this area, more is needed to demonstrate the practicality of the various tools, particularly in areas of soft bottom soil and poor visibility.

(8)

All repair options become more complex when bundled lines are involved. The method of bundling lines should be selected to allow easy identification of a failure, and to facilitate subsequent repairs.

(9)

Future developments should address surface vessels, repair tools, and termination designs, as well as methods for detecting and locating failures, both during installations and operations.

IV.

DESIGN ASPECTS OF DEEPWATER PIPELINES AND FLOWLINES

For many deepwater pipeline or flowline installation scenarios, the overall risk of damage may actually be reduced by allowing greater bending strains during installation than is normally allowed in comparable shallow water installations. While pipeline design is not a major part of this program, certain aspects of pipeline design can have a strong impact on selection of a tie-in method. One of the decisions which must be made for any tie-in method is what level of bending stress or strain is acceptable and will be allowed during pipeline installation. This decision has been relatively straightforward in shallow water installations, where the allowable bending strain is usually chosen to be less than or equal to the nominal yield strain of the pipe (typically 85% of yield). Certain exceptions to this rule have been allowed, even in shallow water, such as J-tube risers or pipeline installations by the reel method, where bending strains on the order of one percent are accepted. During investigation of several of the tie-in scenarios it was found that a mere adoption of shallow water guidelines for allowable bending strains either made the method extremely difficult to execute under less than ideal circumstances, or required a great deal of complex intervention (using multiple subsea pullcables, etc.) to maintain bending strains below yield. These deepwater interventions can be costly and inherently carry additional risks to the pipeline or flowline, which is incompatible with the need to keep deepwater operations as simple as possible. Deepwater pipelines are required to have thicker walls in order to resist collapse due to external hydrostatic pressure. A direct result of selecting pipes with thicker walls (or lower D/t ratio) is that the safety factor against buckling due to bending becomes very large if bending strains remain limited to less than the yield strain. Design formulations for predicting pipe behavior under external pressure, bending, and axial load have been established elsewhere (References 1,2). Only one of these relationships is discussed here to illustrate its importance with regard to safety factor against buckling. Figure 11 shows the combined effect of external pressure and bending on the buckling/collapse failure of pipe. The vertical axis gives the ratio of external pressure P to the pipe collapse pressure Pc in the absence of bending. The horizontal axis gives the ratio of the bending strain ∈ in the pipe to the buckling strain ∈b, which is the strain at which the pipe will buckle due to pure bending (i. e. , with zero pressure and zero axial load). An approximate empirical failure criterion is given by the straight line shown. The exact position of this line is affected by the yield stress, the initial imperfections in the pipe such as residual stress and initial out-of-roundness, and other factors. Four data points are plotted on Figure 11. These points represent the external-pressure/bending strain coordinates for typical pipes to be laid in water depths ranging from shallow to ultra deep water. These pipes all have the same safety factor (Pc /P = 2.22) against collapse due to pure external pressure, and they are all subjected to a bending strain equal to the yield strain (∈ = ∈y). One sees immediately that the overall safety factor (distance from the failure boundary) of say the deepwater pipeline is much greater than that of the shallow water pipeline. Thus, even though it may be prudent to employ a somewhat greater safety factor for a deepwater pipeline, in view of the economic consequences of a pipeline failure, such prudence generally still allows bending strain in the deepwater pipeline or flowline that is greater than the yield strain. Of course, a number of other factors play a role in selecting an allowable bending strain, such as the influence of cyclic loadings and the effects of residual plastic deformations on the pipeline on-bottom configuration. Nevertheless, Figure 11 vividly illustrates that this aspect of pipe design should be considered when evaluating deepwater installation scenarios. It is poor engineering practice to consider certain methods infeasible for deepwater pipelines merely because they do not satisfy commonly used shallow water guidelines.

Figure 11. Deepwater and Shallow Water Design Comparisons

V.

SCALING RELATIONS

A.

Kopp(20)

INTRODUCTION

This section describes scaling relations used to select properties of the models which were made to simulate the tie-in concepts studied in this program. Selection of the scale and properties of a pipe model is often a compromise between conflicting requirements. Specific requirements for each individual test dictate which parameters are of greatest importance and thus which should be scaled correctly. Mathematical formulations of static and dynamic scaling relations are well established, but this does not mean that model pipe and scale selection is straightforward. Often exotic combination of materials may be required to achieve complete similarity between prototype and the model pipe. In many cases, however, complete similarity is not necessary and partial similarity will not lead to significantly different results. The importance parameters of pipe characteristic length C and modified characteristic length c* are defined by the equations : C = (EI/W)1/3

(1)

C* = (EI/µW)1/3 = C/(µ)1/3 Where

(2)

EI = pipe bending stiffness W = pipe submerged weight per unit length µ = friction coefficient between pipe and seabed

The constant C is a characteristic length property of any suspended pipeline or flowline, whereas C* is a characteristic length property of a pipeline in contact with the seafloor. Characteristic lengths of model pipes used in the DPFRI Program ranged from a low value of 0.6 ft for speedometer cable to a high value of 19 ft for 3/4” Schedule 40 steel pipe as shown below. The values shown are nominal values; actual values may differ slightly from the values shown.

Unit Weight W (lb/ft)

Bending Stiffness EI (ib-ft2)

Charecteristic Length C (ft)

Speedometer cable, in air In water

0.034 0.030

0.0065 0.0065

0.58 0.60

1/4” PVC solid rod, in air in water

0.032 0.011

0.65 0.65

2.7 3.9

5/8” PVC solid rod, in air in water

0.199 0.066

24.97 24.97

5.0 7.2

1/4” Sch 40 steel pipe, in air

0.43

695

11.7

3/4” Sch 40 steel pipe, in air

1.13

7720

19.0

Spiral pipe bundle, 3-1.0” OD

3.0

18700

18.4

Model Pipeline Materials

B.

SCALING RELATIONS

For the primarily static, model simulation tests conducted in this program, the following summarizes the scaling considerations of importance: (1)

For pipelay problems, where correct scaling of the suspended span configuration is more important than modeling the pipe/seabed friction interactions, the parameter to be scaled correctly is the pipe characteristic length C, as defined in Equation (1).

(2)

For problems where pipe/seabed interaction is important, it may be better to scale the modified-characteristic length C*, which is similar to the pipe characteristic length C, but allows for differences between model and prototype pipe/ seabed friction coefficients.

It can be shown that static similarity between full scale prototype pipe (p) and model pipe (m) is obtained, when the two relations are satisfied :

λL =

λF =

Lp Lm Fp Fm

=

=

Cp Cm Wp Wm

=

(EIp / Wp )1/ 3 EIm / Wm )1/ 3

λL =

EIp EIm

λ−L2

(3)

(4)

where λL = length scale factor (ratio of prototype to model dimensions) λF = force scale factor (ratio of prototype to model forces) For a given prototype characteristic length and scale factor, Equation (3) thus establishes the required model pipe characteristic length. Equation (4) then establishes the ratio between prototype and model external forces. If one wants to simulate, for example, on-bottom pull-in of a pipe or flowline, the scaling relations can be modified to account for differences in the prototype and model friction coefficients between pipe and seabed. This is easily recognized if we replace the term "W" in Equations (3) and (4) by "µ times W", where "µ" is the friction coefficient between pipe and seabed. Equation (3) then becomes :

λL = *

(EIp / µ p Wp )1/ 3 (EIm / µ m Wm )1/ 3

=

C *p Cm*

(5)

*

where C p and Cm are called the modified-characteristic length of the prototype and model pipe, respectively. The concept of the pipe characteristic length has been used previously by TRED. Participants in earlier Shell Deepwater Joint Industry Program may recognize this as a scaling parameter, frequently used to prepare nondimensional curves. As an example of such a relationship for a suspended pipe in deep water, it can be shown that the minimum sagbend radius of curvature at zero bottom tension is about 1.4 times the pipe characteristic length, provided that the pipe bending strains remain elastic. Similarly, on-bottom type pull-ins can be nondimensionalized by relating pipe configurations to the modified-characteristic length. In other sections of this DPFRI study the concepts of pipe characteristic length and modified-characteristic length are discussed further.

VI.

STEEL CATENARY RISERS (SCR's)

A.

INTRODUCTION

A pipeline SCR consists of steel pipe suspended in a catenary curve between the water surface and the seabed. In this study it was assumed that the riser is hung from an FPS, but installation of an SCR to a TLP or compliant tower is also possible. Maximum stresses near the top of the riser are limited by allowing the attachment point to rotate about axes orthogonal to the longitudinal pipe axis. The catenary shape of the riser is an effective means to absorb the motions of the FPS without the need for special equipment, such as heave compensators. Maximum riser sagbend stresses near the seabed touchdown point are limited by horizontal tension applied to the riser. This tension varies as the FPS moves in response to wind, wave, and current loadings. Analysis done by TRED indicates that, for certain water depth ranges and pipe sizes, the resulting cyclic stress variations are sufficiently small to prevent pipe fatigue failure over the design field life. Flowlines from satellite wells or from wells drilled directly underneath the FPS can also be connected to the FPS using the catenary riser concept. It is assumed that several individual flowlines and control lines are assembled into a flowline bundle (FLB) which is held together by taping or twisting the individual lines. For a satellite well, the FLB is simply laid from the well to the FPS in a relatively straight path, and the final length of the FLB is suspended from the FPS as a catenary riser. In case of flowline SCR's from a well underneath the FPS, the bundle must be laid on-bottom for some distance away from the FPS before looping back towards it. The length of this straight: line section must generate sufficient bottom friction to maintain the proper catenary shape (i.e., it must prevent the riser touchdown point from slowly creeping towards the FPS). This long flowline section on the seafloor adds to material cost, may cause seabed congestion in case of multiple FLB'S, may cause reel capacity problems on existing reel vessels, and may require a pilot-operated instead of direct hydraulic subsea control system. To reduce the length of this on-bottom section, a subsea anchor may be installed on the seafloor, and the FLB may be looped around this anchor to maintain the required horizontal riser tension. Although the risers are designed to remain connected to the FPS under all design conditions, there may be special events which will require disconnection and abandonment of the risers, for instance, if the FPS has been damaged and must be moved off-site for repair. B.

DYNAMIC ANALYSIS

Patterson, Langner (31)

The dynamics of SCR's were studied using a time-domain finite element computer program developed by TRED (FLOSS or Floating Offshore System Simulator). Certain simple theoretical models were derived from the computer results. Fatigue lives and maximum forces, stresses, and displacements, useful for design were obtained by performing an extensive parameter analysis which simulated catenary risers suspended from a semisubmersible FPS in deep water areas of the Gulf of Mexico. Parameters investigated included: pipe size and weight, water depth, and bottom tension (static offset). Regular heave, surge, and combined heave and surge motions at various frequencies were applied to the upper end of the riser. Although the problem is generally nonlinear, aspects of the riser response to the applied motion proved to be nearly linear, and a linear solution bound was evident. Thus, the results could be characterized by riser transfer functions or Response

Amplitude Operators (RAO's) permitting the use of spectral analysis to predict extreme values and long-term fatigue damage. (Note that the RAO's are similar to transfer functions, except that the RAO's do not provide phase information.) The results suggest that SCR's are a viable, potentially cost-effective, riser alternative for deep water development. Figure 12 (a) shows a typical computer model of an SCR. To accurately model the bending stresses in the riser sagbend, and thus the element size was decreased in this region. RAO's were developed from the FLOSS output for displacement, tension, and bending stress for each nodal position of interest. Conservative general RAO's were derived, for instance, for the maximum bending stress in the sagbend region [Figure 12 (b)]. Figure 13 illustrates how these RAO's were then used to determine stress response spectra. These stress response spectra, in turn, were used to estimate riser fatigue life for the sagbend region. Given a linear system, the general relationship between input and output spectral density functions is given by : S(ω)output = S(ω)input RAO(ω)²

(6)

To obtain the riser response spectral density function directly from the sea state spectrum, two RAO's are required; one for' the semi-submersible vessel response to the sea and the other for the riser response to semisubmersible motions. Then :

S(ω)response = S(ω)sea RAO(ω)² semi RAO(ω)2 riser

(7)

For this work the Gulf of Mexico sea conditions were characterized by 27 Pierson-Moskowitz spectra. The four highest sea state spectra (S(ω)sea 24,25,26, 27) are shown in Figure 13 (a). The vessel RAO [RAO(ω)semi] shown in Figure 13(b) was obtained by using the heave RAO for a typical large semisubmersible. The riser RAO [RAO(ω)riser ] shown in Figure 13 (c) was obtained by fitting an upper-bound curve to the data of Figure 12(b). [Note that the radial frequency ω is inversely related to the period T by ω = 2π/T]. Thus the terms on the right side of Equation (7) are known and the stress response spectra for each sea state can be determined (Figure 13 (d)]. These stress response spectra were then used to determine the fatigue lives of various hypothetical catenary risers. The fatigue calculations involved the Palmgren-Miner theory, combined with the API-X fatigue curves, applied to the cyclic bending of the riser sagbend. In this manner the fatigue life for a catenary riser suspended from a typical large semisubmersible in the GOM is predicted to be more than 10,000 years. This estimate does not account for corrosion-fatigue damage or long term drift motion effects. Since this predicted fatigue life is so long, some other failure mechanism(s), such as corrosion or fatigue at the upper support point, will likely impose the ultimate lifetime limitation. Note that this result is dependent on both the sea state assumptions and the vessel motion characteristics. In an extreme case where the riser follows the motion of the waves rather than the vessel (for example, if the riser were suspended from a buoy) then the predicted fatigue life would be about six months instead of 104 years for the same sea conditions. Nevertheless, even for small semisubmersible FPS vessels, catenary riser fatigue lives should be acceptable for the GOM environment.

Figure 12. Steel Catenary Riser Analysis

Figure 13. Steel Catenary Riser Spectral Analysis

In a similar fashion, RAO's were determined for riser top tension and maximum riser displacements. These RAO's were then used to determine extreme values of the tension and displacement amplitudes. For example, for a gas filled steel riser (submerged weight 122 lb/ft) in 3,000 feet of water in a storm represented by a significant wave height of 40 feet and a zerocrossing period of 11.5 seconds, the maximum dynamic tension amplitude is 61.6 kips. This value must be added to the mean top tension to determine the expected extreme tension. These dynamic analyses were sufficient to demonstrate the technical feasibility of SCR'S, but further analysis still may be required in some areas. Recommendations for further work are given in Section XII. C.

INSTALLATION, ABANDONMENT, AND RECOVERY OF PIPELINE SCR'S

The scenarios tested and analyzed were : (1) (2) (3) (4)

first-end installation by pulling the pipeline towards the FPS as it is being made-up on a lay vessel, then laying away; second-end installation by laying the pipeline towards the FPS, then handing-over the pipe-end from the lay vessel (positioned close to the FPS) to the FPS; A&R by keelhauling underneath the FPS; and A&R by U-turn abandonment on the seafloor and S-curve recovery.

Unless a vertical lay method is used, first-end and second-end installations will be difficult, because they will involve having to add a certain length of pipe at the FPS itself. The amount to be added depends on the distance between the lay vessel and the FPS, and the horizontal component of the tension in the pipe. For other methods than a vertical lay method, the length of pipe to be stalked-on can be over a thousand feet, creating significant operational difficulties at the FPS. Candidate installation methods would thus involve either a DP reel vessel with a vertical stinger or a DP J-lay vessel (for instance, a converted drill ship). The first-end installation method consists of making up pipe at a lay vessel, stationed at some distance from the FPS, and pulling the pipe in towards the FPS. The pipe configuration during pull-in is determined by the pipe departure angle at the lay vessel, applied pipe tension, and the length of pipe on the seabed, if any. In a second-end installation, for example using a reel vessel with a stern pipe departure ramp, the lay vessel lays the pipe until it is close to the FPS, and then lowers the pipe and to just below the departure ramp to allow the vessel to turn around 180 deg and back-up towards the FPS. If the pipeline can be installed prior to installation of the FPS, a combined first- and second-end installation is possible, using a tow, S-lay, reel, or J-lay method to lay the pipe with the pipe end located in a predetermined target area past the FPS. After installation of the FPS, a special DP construction vessel retrieves the pipe end to the surface and makes a second-end connection at the FPS. This will require keelhauling of the riser under the FPS, however, which is considered a complicated and risky operation. Making an outboard second-and connection was practiced in the wave tank. The tests identified the need for heavy lifting equipment on the FPS. In case of second-end installation of an air filled 12.75-inch riser in 3,000 ft water depth, a maximum static lift load of about 100 kips should be expected. Contingency plans should be made, however, to lift the riser fluid-filled, wherein the lift loads will approach 250 kips. It will be necessary to rig extra steering cables to the pipe end, to correct out-of-plane movements during transfer.

A&R by keelhauling the riser involves passing the riser underneath the FPS to the other end and then laying the pipe down in a straight line. The most critical part of the operation is passing of the riser underneath the FPS. Even in the absence of other vertical risers underneath the FPS, this is a delicate operation because it involves simultaneous handling of two lift cables and because the pipe end cannot be seen during the keelhauling. The shallow angles of the lift cables will cause the riser end to make a large vertical drop if one of the cables breaks, or is inadvertently payed out too fast. This may result in buckling of the riser, which was simulated during the model tests by asynchronous winch operations. If this type of problems can be avoided, bending strains in the pipe sagbend remain moderate during the operation. A&R by U-turn abandonment and S-curve recovery consists of transferring the riser end to an auxiliary DP vessel, which in turn makes a U-turn away from the FPS, and then lowers the pipe end to the seafloor (Figure 14). The A&R cable should preferably be left attached to a buoy to avoid having to disconnect and reconnect the cable on the seafloor by some kind of diverless procedure. Leaving the cable attached to the buoy is only practical, however, if just one pipe is located in the area. If other pipelines are present, the A&R cables may foul, and the buoys may drag the pipelines over each other. To recover the pipe, the pipe is picked up to the surface and moved back in an S-curve towards the FPS. A second-end connection is then made at the FPS. The main advantage of this method over the keelhauling method is that only one vassal is used to handle the riser. To avoid overstressing the riser, the pipe departure angle must remain less than 90 deg, which requires careful operations. The model tests showed that the operation was straightforward and could be repeated several times for the same riser without significant change in the riser configuration. The radius of the pipe in the U-turn is determined by the friction between the pipe and the seabed, the pipe bending stiffness, and tension applied at the top. The characteristic shape of the U was repeatable during each test. A very slight S-shape remained after reconnecting the riser to the FPS. The radius of the U-curve may be small enough to cause yielding of a large diameter pipeline, especially if filled with fluids. This has two effects : (1) (2)

the radius of the U-turn decreases as the pipe starts to yield; and the fatigue life of the pipe may be influenced if this curved section of the pipe is part of the riser sagbend region.

Thus the possibility of having to abandon the riser at some point in time must be considered in fatigue calculations. D.

INSTALLATION AND A&R OF FLOWLINE SCR'S

The scenarios tested and analyzed were : (1) (2)

(3)

first- or second-end installation of an FLB to a satellite well at some distance from the FPS; first-end connection of one or more FLB's to wellheads directly underneath the FPS, followed by lay-around a subsea anchor, and second-end connection of the FLB risers to the FPS; and A&R of FLB SCR's.

In general, the governing factor in safely installing, abandoning, and recovering flowline SCR's, is not the bending strain in the flowlines, as was the case with installing pipeline SCR's, but avoiding interference of. multiple FLB's. Especially in GOM development scenarios, it is likely that several subsea wells will be drilled underneath the vessel, and if each FLB has to loop around, like shown in Figure 15, interference is unavoidable, unless the FLB configuration is carefully planned. Installation of a flowline SCR for a flowline from a satellite well is very similar to that of a trunkline, except that it is unlikely that any other installation method than a vertical reel method will be used to install the flowlines.

Figure 15. Installation of F from a Central

Installation of a flowline SCR for a flowline from a wellhead underneath the FPS is illustrated in Figure 15. An alternative to the first-end, two-vessel stab-in operation at the wellhead [Figure 15 (a)] was practiced during the model tests [Figure 15 (b)]. This alternative involves moving the floater sideways so that a DP vertical lay vessel can maneuver itself directly above a wellhead. The FLB can then be connected to the wellhead by the stab and hingeover method. In case of installing multiple FLB risers, several problems may arise : (a) (b) (c)

congestion at the riser touchdown points; congestion and overlapping at a subsea anchor; and interference during FLB abandonment or replacement.

These problems can be avoided by careful planning of riser layout and development of installation scenarios. A second-end connection must be made at the FPS, in a similar fashion as discussed previously for a trunkline connection. During the model tests, connections were made by lowering the riser end into supports on one of the FPS pontoons. The advantages of pontoon supports are : (a) (b) (c)

the steel risers are protected from the most severe wave action; separating the bottom and the top section of the riser makes it possible to just replace the top section; and in case of multiple flowline SCR's the SCR's which have already been installed are protected from contact by, for instance, the installation vessel.

Disadvantages are : (a) (b) (c)

the supports are underwater and thus diver intervention is required; the final lowering of the riser into its support cannot easily be observed; and the riser must be pulled towards its support with additional cables rigged underneath the FPS.

It is recommended to temporarily tie-off the riser end at the deck level of the FPS, make length adjustments if needed, install support brackets and valves, and connect the composite pipe hoses. Then the riser can be lowered down into the support on the pontoon, without the need to pull the riser inwards or outwards, provided the temporary and final support are in the correct position relative to one another. The model tests were done in this fashion and it was not difficult to lower the riser down to its pontoon support, after first temporarily tieing it off at the deck level. If the riser needs to be repaired or abandoned, or if some work needs to be done at the valves above the pontoon support, a similar procedure can be followed by lifting the riser support to a temporary support on the deck level.

VII.

STEEP S BUOYANCY SUPPORTED RISERS A.

Kopp (32)

INTRODUCTION

The concept of the SSBR made of composite pipe [Figure 2 (b)] was developed to minimize the need for other vessels than the FPS for riser installation, production, and maintenance operations from wells drilled at a radius of about 300 ft underneath the FPS. This concept may find applications but will require considerable development. B.

INSTALLATION

If the FPS itself is used to install the SSBR, the following procedure is required (a) (b) (c) (d)

install a reel with the FLB on it, on the FPS; attach a bottom connection tool to the riser end (heavy enough to overcome the positive buoyancy provided by the buoyancy near the S-curve); lower the riser, while moving the FPS to obtain the desired steep S-shape; and stab the tool into the wellhead. This is a complicated operation, especially in the presence of other risers.

Alternatively, the FPS vessel can move away from the wellhead to make place for a reel vessel, like one of Coflexip's DP Flexservice vessels, which can lower the riser, make a stab-in connection at the wellhead, and hand the top end of the riser over to the FPS. This procedure is more complicated, and will require keelhauling of the riser underneath the FPS for inboard installation. C.

STEEP S RISER INTERFERENCE

Although the FPS vessel is normally maintained in a central position over the wellheads, it is sometimes necessary to move the vessel to an offset position for drilling or maintenance operations over a particular well. In addition, during severe storms, the FPS vessel may move about in a random fashion, with surge amplitudes as large as 5% to 10% of the water depth. Prior to model testing it was assumed that the vessel excursions would lead to formations of sagbend features in the risers as shown in Figure 2 (b), if the vessel would move directly from its neutral position over the wellhead. The model tests, however, indicated that the risers would rotate outof-plane (Figure 16) before the vessel movement was completed, resulting in riser interference. The parameters which were varied in the model tests were : (a) (b) (c) (d)

riser configuration and properties; riser manipulation; well pattern; and vessel movements.

Only the use of prescribed vessel movements proved to be effective in preventing riser interference. If this method is pursued during actual field development, however, testing will be required at much larger scales, to enable a better scaling relation between prototype and model pipe or bundle. Special care would also have to be taken in modeling the riser supports. A 1:10 to 1:20 scale test would be required, which in turn would require testing in a very deep tank, a deep lake, or offshore. Even if these tests are successful (and there would be considerable difficulty in determining riser configurations in deeper water, especially offshore, because of the difficulty in visually observing the entire system), substantial development work is needed to prove that the concept has practical applications. More than with any other riser test in the wave tank, these tests showed the value of small scale tests to quickly obtain three dimensional, visual information and to discover potential problems, before more extensive development and analysis work is undertaken

. VIII.

LATERAL DEFLECTION OF PIPELINES AND FLOWLINES

A.

Kopp (41,42,43,71)

INTRODUCTION

This section describes the results of small scale model testing and numerical analyses conducted to technically evaluate deepwater, diverless first- and second-end tie-ins of pipelines or FLB's made by using the lateral deflection method. This method consists of deflecting the pipeline or FLB from an initially straight laydown position, at some offset distance from and overshoot distance past a subsea structure (wellhead or subsea template), towards this structure, using one or more pull-in cables. The pipe may be dragged over the seafloor or, in case of large pipe sizes, the section of pipes to be deflected may be kept at a predetermined distance above the seabed by a combination of buoys and chains (Figure 4). This method has been used for deflection of single flowlines by dragging these laterally over the seabed. For pipeline sizes greater than about 12.75 inches or larger FLB's, it may be necessary to support a segment of the pipe off-bottom using a combination of chains and buoys to maintain a fixed elevation of this segment above the seabed and thereby reduce pull-in forces. The offbottom method has also been used before on some pipeline and FLB installations, but no commercial installation has been made yet in deep water (deepest trial connection was made in 1981, in a water depth of 820 ft), and probably none could be called truly diverless. At least one operator is planning to use this method in a GOM deepwater field development. The principal objectives of the testing and analysis part of the program were to : (a) (b) (c) (d) (e)

show a range of possible deepwater, diverless applications for the lateral deflection method; predict pipe configurations (during deflection and final); find methods to correct for pipe misalignments at the subsea structure, caused by initial over- or undershooting of the target laydown area; find methods to avoid overstressing the pipe during the deflection operation; and predict pipe pull-over and final alignment forces.

In addition, procedures were developed for special applications of this method, for example, how to install a pipeline underneath a mooring system of an FPS. These procedures are not discussed here, but further details can be found in Section 43. The objectives described above were achieved by conducting a large number of small scale tests (scale range primarily 1:60 to 1:120) on a model table [Figure 5 (a)] and in the wave tank at WRC. In addition, large scale tests (scale range 1:12 to 1:16) were conducted in the wave tank [Figure 5(b)] to study final alignment, rotation of the pipe during deflection, and the influence of pipe yielding. Comparative numerical analyses were also conducted. Parameters which were varied included geometric variables (initial offset and overshoot, pull point location), seabed friction, and pipe size (ranging from small FLB's to 20-inch pipeline). The results of the model tests and numerical analyses were used to develop tie-in procedures for the lateral deflection method. These procedures are discussed in Section 43. That section also contains nondimensional curves which can be used to make preliminary decisions on the selection of deflection parameters, without the need for extensive numerical analysis. An example curve is shown in this summary report (Figure 19).

Figure 16. Riser Interference during FPS Movements

A. Plan View of FPS Movement

B. Underwater View A-A

B.

SMALL SCALE MODEL TESTS AND NUMERICAL ANALYSES

Figure 17 shows the model table which was constructed for this program. The table top can be tilted to simulate the influence of cross-currents. The tests conducted were : (a) (b) (c) (d)

deflection of a 20-inch pipeline from a target area 270 ft offset from a template, varying the magnitude of the chain drag force, magnitude of bottom current, and the location of the pipe end with regard to the center of the target area; deflection of the same pipe from a target area 540 ft from the template; deflection of a 12-inch pipeline, varying the number of pull-in cables, the chain drag force, and the location of pull points; and deflection of a flowline bundle to a wellhead.

The results of the tests demonstrated (a) (b) (c) (d)

in general, the ease with which different concepts could be evaluated quickly by model table tests; the risk of overshooting or hitting the template with the pipe end, if the target area is too close to the template; the feasibility of using subsea winches, either deployed by cable or using a drill string to facilitate pull-ins; and the relative insensitivity of final pipeline or flowline configuration to variations in parameters other than the initial pipe geometry (location of pipe end in relation to the tiein point on the subsea structure).

Several aspects of the lateral deflection of a 20-inch pipeline (Model Test Series 1) were modeled with the general purpose finite element program ABAQUS. Agreement between the model table tests and the ABAQUS simulations was excellent [Figure 17 (b)]. The lateral deflection problem was also modeled with FLOSS. Agreement between the two finite element programs was excellent. The finite difference program LATERAL was developed for use as a simple tool for analysis of onand off-bottom pipeline and flowline deflections. This program, which is written in BASIC, can be run interactively on a PC and has excellent predictive capabilities for a number of important lateral deflection parameters [Figure 17 (b)]. Program sponsors are provided with the computer program listing. Tow tests were conducted in the wave tank on a 1: 120 scale to simulate towing a long pipeline or FLB into a small target area. The simulated full-scale water depth was 1200 ft and the simulated mean pull force about 200 kips. The tests demonstrated that it may be very difficult to place the pipe end inside a small target area (about 50 ft wide) without taking special precautionary measures. This is caused by : (a) (b) (c) (d)

soil "stick-slip" behavior which is caused by the difference between the static and dynamic soil friction coefficient, i.e. the starting and pulling friction; the difficulty of estimating the correct timing of when to stop pulling the pipe; the elasticity of a very long tow cable catenary; and the inertia of a long pipe string.

It will be risky to restart a pipe tow after the tow was stopped for whatever reason, if the pipe end is already close to the edge of the target area. A restart will likely result in overshooting the target area by a considerable distance. Further work is required to develop practical contingency procedures for target over- or undershooting in case of using the lateral deflection method combined with a pipe tow.

Figure 17. Off-Bottom Lateral Deflection of 20-inch Pipeline

A. Model Table Set

B. Comparison with Numerical Analysis

C.

LARGE SCALE WAVE TANK TESTS AND NUMERICAL ANALYSES

The primary reason to conduct these large scale tests in the wave tank was to enable us to correctly model the pipe deflection in all three dimensions and to realistically model the effects of buoyancy distributed along the pipe, in case of off-bottom deflections. Information was obtained primarily on the effects of axial pipe rotation and yielding of the pipe during deflection. The 1:12 scale model used to simulate the lateral deflection of a 12.75-inch pipeline, was a 3/8Inch (nominal diameter) size steel pipe, which was outfitted with buoys for the off-bottom lateral deflection [Figure 5 (b)], and provided with additional steel weights for simulation of on-bottom deflections with high seabed drag resistance. The 1:16 scale model used to simulate the lateral deflection of a typical FLB, consisted of two 1/4 -inch PVC rods and one 1/8-inch PVC rod, to simulate an FLB consisting of two 4.5-inch flowlines and one 2.375-inch annulus line. The Electro/Hydraulical control umbilical was ignored. Two methods were used to bundle these lines. The first one was the more conventional method of taping the individual lines, which created a composite bundle with different bending stiffness in different planes of bending. This was primarily done to demonstrate that such a bundle will start to rotate around its longitudinal axis during deflection. The second method of bundling consisted of forming a helical bundle. In this case, the combined bending stiffness of the bundle was simply the sum of the stiffnesses of the individual lines without composite effect. No bundle rotation was expected during deflection. A two-pin sled had been fabricated for some earlier model tests and this sled was also used for the deflection tests of the 12.75-inch pipeline model. The purpose of the two guidepins versus a single bullnose is to reduce alignment forces, especially in the horizontal (lateral) plans. The distance between the pins helps to create a moment arm, at least initially, to avoid high pull forces during lateral misalignment correction. The initial deflection was conducted with just one cable until the pipe end was close to the receiver. At that point the second cable was connected. The two-pin, two-cable sled has been used in most prior designs of lateral deflections of pipelines, although bullnose type fittings with a one-cable final pull-in have also found use. A simple bullnose with a one-cable attachment was used as end fitting for the FLB models. Most pull-ins of small FLB's have been completed using this type of one-cable pull-ins with a bullnose type fitting, although some flowline end fittings consist of a two-cable sled also. Pipe configurations in the horizontal plane (X,Y plane) were measured using a special measurement. probe [Figures 5 (b) and 18]. This probe could be moved to any point along the pipe and measure the coordinates of that point in a preestablished reference coordinate system similar to an acoustic transponder net which could be used in the field. The finite difference program LATERAL was used to compare model test results with numerical analyses [Figure 18 (b)]. The program also proved useful to make estimates of lateral misalignment moments in case of off-bottom deflections. The main value of the off- and on-bottom deflection tests was that it demonstrated the feasibility of on-bottom lateral deflections of small diameter pipelines (up to about 12.75-inch) and in some cases even larger diameters. Other findings, for instance the effect of axial pipe rotation and problems associated with vertical and lateral alignment of the sled with the alignment structure, are discussed in Section XI-A of this summary report. Most designs, thus far, for lateral deflection method of a pipeline or FLB with a greater diameter than about 10-inches have been off-bottom deflections to minimize pull-in forces. Most of these designs were for relatively shallow water applications with relatively high pipe submerged weights due to concrete coating. In deep water, however, the advantage of a possibly low submerged weight, combined with the very high cost of applying buoyancy to the pipe, may make an onbottom deflection an attractive alternative.

Figure 18. On-Bottom Lateral Deflection of 12.75 inch Pipeline

The results of the tests and the analyses demonstrate that on-bottom deflections are viable, provided one accepts minimum radii of curvature considerably less than commonly allowed for this type of tie-ins. In shallow water applications, the allowable radius of curvature is typically based on an allowable stress of about 85% of the SMYS. There are three reasons why yielding of the pipe during deflection could be undesirable : (1)

Yielding of the pipe could cause a plastic hinge which, in the worst case, could lead to pipe buckling but also could act as a torsional anchor.

(2)

Yielding of a pipe section could result in permanent deformation of the pipe which, if this pipe section is close to the pipe end, could cause misalignments at the alignment structure.

(3)

If the pipe is subject to plastic deformation, ordinary computer programs, based on linearelastic analysis may not be applicable anymore and more cumbersome and expensive elasto-plastic analysis might be required.

The previous discussion in Section IV already pointed out that the safety factor against pipe buckling due to bending may be much greater for a deepwater pipeline than for a shallow water pipeline, for identical collapse pressure ratios and bending strains. It may therefore be acceptable to accept bending strains somewhat higher than yield for deepwater lateral deflections. Unless the pipe segment which is subject to elasto-plastic deformation is straightened out during the final stage of the deflection, there should not be a problem with remaining deformation causing unanticipated lateral misalignments at the template. The need for more elaborate elasto-plastic analysis techniques was indeed found to be true, but the cost of engineering analysis is a relatively small part of the total pipe installation cost. Flowline Deflection The main conclusion from the tests was that deflection of the taped bundle, which had a substantial difference in bending stiffness in the two, vertical and horizontal, bending planes, caused axial rotation of the bundle This was expected and the tests showed that this rotation can be up to 90 deg. The exact amount is very difficult to predict and alignment equipment should be designed to at least handle this amount of rotation. D.

PROCEDURES

Kopp (43)

A nondimensional graph was developed to aid designers in selecting preliminary lateral deflection parameters (Figure 19). The graph was prepared using the finite difference program LATERAL. One of the advantages of using this graph is that it allows a designer to very quickly see how a change in for example the size of the target area causes the misalignment angle at the template to change, for a given offset and overshoot. The diverging lines of constant approach angle θ indicate that increasing offset and overshoot distance will allow one to increase the size of the target area for a given allowable range in misalignment angles ∆θ. Of course, an increase in offset distance will also increase the pull force and this trade-off must be considered for each individual case.

Figure 19. Non Dimensional Curves to Evaluate Lateral Deflection Configurations

IX.

SECOND END FLOWLINE PULL-INS

A.

Langner, Patterson (51,52,76)

INTRODUCTION

Second-end flowline connections to satellite wells have quite often been made using some type of lateral pull of the flowline which was initially laid down in a target area some distance past the wellhead. As the water depth increases it becomes increasingly more difficult to place the pipe end inside a target area small enough to avoid excessive angular misalignments at the wellhead. An alternative method was developed, which involves increasing the initial overshoot distance so much that the pipe will be deflected towards the wellhead in a U shape, back in the direction it was laid from. Provided the overshoot distance has a certain minimum-length (about 14 to 16 times the flowline modified-characteristic length C*), the misalignment angle becomes practically independent of the length of the target area, which greatly facilitates the deepwater tie-in operation. Relationships were developed to predict the final pull-in direction (or misalignment angle) and the minimum radius of curvature during the pull-over. This was done by numerous small scale (1:100) tests on Shell's model table and by computer analyses. Some large scale (1:4) land tests were also conducted to study the affects of pipe yielding on rotational alignment of the pipe.

B.

SMALL SCALE MODEL TESTS

The primary parameters which were varied were : (a) (b) (c)

geometric variables (such as offset and overshoot distance); coefficient of friction between modal pipe and table top surface; and the pipe lay down and pull-in procedure [pull from an initially straight or curved laydown position (Figure 5)]. The model pipe used was 1/8-inch speedometer cable, which on a 1:100 scale represents a typical FLB.

Photographs were taken of each test (Figure 6) and results were nondimensionalized so that they can be used for any size flowline (using either the flowline characteristic or modifiedcharacteristic length). The table top surface was varied using four different surfaces (rubber, felt, plywood, and cork) to obtain a large range of variations in longitudinal and lateral friction coefficients. Section III-E of this report describes the two tie-in procedures which were developed during the model table tests.

C.

NUMERICAL ANALYSES

The second-end flowline pull-in procedures described above were analyzed using the computer programs ABAQUS and SPAN. ABAQUS is a general purpose finite element program widely used in the oil industry. SPAN is a program developed by Southwest Applied Mechanics, Inc. (SAMI) for use in pipe stability and pipe deflection analyses, and is particularly convenient for studying these flowline pull-ins since it will run on a PC. Results obtained with both programs agree well with the results of the modal tests (Figure 20). The parametric computer simulation using SPAN was carried out by SAMI and the results of the analyses are contained in appendices to Section 52. Both static as well as dynamic runs were made.

Figure 20. Second-End Flowline Pull-in-Comparison of Model Table Tests and Numerical Analysis

The main conclusions of this study are : (1)

Numerical analysis confirms the simple pull-in relationships derived by Langner from the small-scale model testing program.

(2)

Stresses may exceed yield during the pull-in process. However, plastic material behavior only affects the final alignment of the pipeline at the subsea structure when the initial onbottom position of the pipe end is close to the pull-in point.

(3)

Geometric parameters (i.e., position of the pipe end with respect to the termination point) govern the analysis. Nevertheless, friction effects are also significant, particularly for predicting pipeline stresses.

(4)

Results from dynamic pull-in simulations agreed with results from corresponding static simulations. This indicates that inertia effects and path history differences are not significant, at least for realistic pull-in rates. Therefore, for most second-end pipeline pullins, static, elastic analysis is adequate.

(5)

The analysis results were used to develop nondimensional curves to aid the engineer in preliminary selection of important parameters, without the need for computer analysis. Figure 21 shows a set of these nondimensional curves for an initially straight laydown position. It shows that for larger overshoot distances X (X greater than about 12C) the final misalignment angle is practically independent of friction and the overshoot distance itself.

D.

LAND TESTS

The land tests (1:4 scale) involved repeatedly pulling and twisting one end of a long pipe string over a grassy field, in both straight and curved configurations, using both single pipes and pipe bundles (1/4", 3/4”, and 1" steel pipe). Longitudinal friction coefficients varied between about 0.3 and 0.6, the higher values being associated with wet soil conditions. These land tests demonstrated the effects of developing a plastic hinge too close to the end of the pipe. Not only can this cause significant angular misalignment in the horizontal plane, but it can also require very large torsional moments to be applied to the end of the flowline in order to rotationally align the flowline after pull-in, because this plastic hinge acts as a torsional anchor. An example calculation illustrates this : Assume we have a helical FLB, with a bending stiffness of 5 103 kip-ft² , a torsional stiffness of about 3.8 x 103 kip-ft², and a full yield moment of about 60 kip-ft. As a result of the 180 deg U-type pull-in, assume that the flowline end has rotated approximately 180 deg (i.e., the flowline termination arrives at the wellhead upside down). Assume that the distance from the wellhead to the apex of the U-curve is about 200 ft (about 4 times the FLB characteristic length), and that a plastic hinge has developed at the apex which effectively anchors the flowline bundle in torsion. Then the design torsional moment must at least be equal to : Mtorque = Torsional stiffness x Twist angle/pipe length to anchor = 3.8 x 103 kip ft2 x ( π radians)/ (200 ft) = 59.7 kip-ft Since this torsional moment Mtorque is close to the torsional yield moment (which is approximately equal to the yield moment in bending), the alignment mechanism at the wellhead must be capable of applying the full torsional yield moment of the pipe. If the FLB had been tightly wrapped and thus had acted as a composite beam, the torsional alignment moment might have been as much as a factor 2 greater. The obvious way to reduce these high torsional moments, which are required to orient a flowline bundle after .pull-in, is to insure that any plastic bending remains a long distance away from the end of the flowline. This may be accomplished by laying well past the wellhead before pull-in, or by pre-laying the flowline in a U-curve as described in Section III-E.

Figure 21. Non Dimensional Curves to Evaluate Flowline Pull-in Parameters

X.

STAB AND HINGEOVER CONNECTIONS

A.

Langner (61,76)

INTRODUCTION

The stab and hingeover method consists of the following steps (Figure 7): (a) (b) (c) (d)

position the flowline installation vessel (which must be capable of vertically laying the flowline and must be DP'ed) over the well; attach a stab-in tool to the end of the FLB and start lowering the FLB from the vessel; stab the tool into a receptacle on the wellhead; and begin to move the vessel away from the wellhead. By continuing to pay out flowline from the vessel the flowline is then hinged over and laid on the sea floor.

Some of the primary advantages of this method are : (a) (b) (c) (d)

only one installation vessel is required for both pipelay as well as connection of the flowline to the wellhead; the connection operation itself is simple and fast, without the need for special subsea equipment such as pull-in winches; the connection operation can be accomplished either before or after installation of the production tree, although before is preferred; and loads on the wellhead are relatively small.

The problems addressed in this study were : a) b) c)

how to successfully stab the tool into the wellhead; how to protect the wellhead from heave-generated motions of the stab tool, and how to latch or lock the stab tool into the receptacle on the subsea wellhead.

B.

WELLHEAD OVERLOAD PROTECTION AND FLOWLINE LATCHING

The most severe loading condition which the stab tool can apply to the wellhead could occur just after the stab tool has been connected to the well, but before sufficient length of flowline has been payed out from the lay vessel, so that the FLB becomes taut as the lay vessel heaves upward. Theoretically, the full breaking strength of the FLB could be applied to the wellhead guidebase in this manner, causing permanent and irreparable damage to the subsea well. One possible solution is to "heave compensate" the vertical FLB from the lay vessel, so that the FLB remains stationary in the water during the stab and hingeover operation. However, this technique is not practical because : (a)

the system must continuously pay out a substantial length of flowline during the hingeover;

(b)

the system is only needed for a very short time period; and

(c)

the dynamics of such a free hanging system would be difficult to predict and control - it may well increase the heave motions instead of suppressing them.

A series of simple mechanisms were developed during this program which would latch the stab tool into the receptacle on the wellhead guide base and are activated by the hingeover operation. These mechanisms are designed to protect the wellhead from heave-generated loads in the FLB, since the latching will only occur after sufficient slack has been developed in the suspended FLB by payout from the lay vessel.

These latch mechanisms are cam-activated by the hingeover action of the FLB (Figures 7 (b), 8, and 22 show different latch options). In effect, the FLB acts as a large crank which is used to perform the latching operation. Typically the latch is not activated until the flowline departure angle reaches 45 deg, which provides the aforementioned heave protection. For further protection, for example to protect against a "drive-off" failure of the vessel's DP system, the hingepin of the stab tool or some other critical part of the latch mechanism can be designed to "break away" at a load that will not damage the well. Some of the latch mechanisms were designed to be reversible, which would allow the removal of the flowline from the wellhead by simply reversing the stab-in operation. C.

STAB-IN TESTS AND ANALYZES

Scale models (1:12 scale) were fabricated of a typical wellhead guidebase and several of the stab tools with cam latches, as discussed above. Stab and hingeover connections were then successfully demonstrated both in air and under water. These model-scale connections were made from heights of 15 to 75 feet, monitored only by closed-circuit TV, with the modal flowline suspended and paid out from a moving platform which simulated typical ship motions. Both high frequency (surge and heave) motions of the installation vessel were simulated as well as low frequency drift type motions. Loads applied to the model wellhead were found to be quite small, proved by a breakaway connection in the model flowline which did not break, even though no effort was made to heave-compensate the flowline at the upper end. During the stab-in tests it was essential to view the stab-in operation from the correct angle of view. In the field, it will be necessary to monitor the stab-in operation not only by means of fixed video ROV nearby. For initial location of the wellhead in deep water and as a back-up measure in case of poor visibility, acoustics also can be used to monitor relative position of the wellhead receptacle and stab-in tool. Further work is required to develop a practical means to provide axial orientation of the flowline. Either the flowline will have to be rotated at the surface, or some means must be found to rotate the stab-in tool near the bottom into proper orientation for stabbing into the wellhead. Preliminary numerical analysis has shown that the dynamics of the flowline stab-in operation are predictable using finite element programs such as ABAQUS. This analysis can be extended to augment the findings of the modal tests and to predict the behavior of the stab-in tool under a range of surface conditions (varying seastates and vessel response parameters).

Figure 22. Different Bottom Latch Options for Stab and Hingeover Tool

XI.

SPECIAL SUBJECTS

This section describes in more detail some aspects which are common to almost all tie-in methods considered, such as final alignment and repair considerations. Also discussed are load limiting devices, bundling of flowlines, buoyancy considerations and submerged weight control, and concepts developed by Shell for very deep water installations. A.

FINAL ALIGNMENT

Kopp (71), Langner (51)

For typical pipeline or flowline pull-in operations the majority of field problems will occur during the last phase of the tie-in operation, when the pipe end is within the final few inches of its receiving counter part on the subsea structure. These problems are often caused by insufficient design or development of the pull-in mechanisms, which may result in hardware failures, and also by underestimating the difficulties with aligning the pipe and in all six degrees of freedom. Simply designing the equipment to be very heavy and powerful may not always be the best solution, because this may cause installation procedures to be more difficult. A better understanding is therefore needed of what is involved in aligning the pips and what can go wrong. Because of the many degrees of freedom, numerical analysis of the complete problem is very difficult, and may not even point out some of the problems. This is an area which requires testing on a fairly large scale, with realistically modeled tie-in equipment. This section describes the results of primarily large scale model tests (scale range 1:16 to 1:10) and numerical analyses conducted to evaluate final alignment and pipe end fittings for pipelines and FLB's. Tie-in methods considered include first-end or second-and pull-ins to subsea structures. Experiments were conducted to make comparisons between results of numerical analyses and tests, and to discover and solve potential field problems with regard to final pipe and alignment. Guidelines were developed for the preliminary design of pipe and fittings, such as a bullnose or a two-pin sled. Axial rotation around the longitudinal axis of the pipe or FLB can result in significant alignment problems especially if a two-pin sled, with two pull-in cables, is used. Final alignment forces were found to be an order of magnitude greater than the force required to just pull the pipe up to the receiving structure on the subsea template or wellhead. The primary objectives of this test program were to : (a) (b) (c)

predict pipe alignment forces during the final alignment stage; find solutions to potential problems during final alignment; and provide guidelines for preliminary design of two typical pipe end fittings, namely a bullnose and a two-pin sled. Parameters which were varied included alignment variables (pipe orientation before alignment, pipe back-tension, and elevation and inclination of alignment structure) and geometry of pipe end fittings.

Causes of Misalignment It is often not difficult to align the pipe and as far as position is concerned (X, Y, Z), but problems arise with angular alignment : pitch (in the vertical plane), yaw (in the horizontal plane), and roll (rotation about the longitudinal axis of the pipeline or FLB). The following are causes of pitch misalignment : (a) (b) (c) (d)

height and orientation of the subsea structure are different than the design values; the seafloor slope is different than assumed; the back tension and pipe submerged weight during installation may differ from the specifications; and both tension and pipe weight will generally be different during installation than later during the operation of the well.

Especially the latter variations in pipe tensions and weight are often cause for problems and will require a. compromise solution. The pitch angle and tie-in height must be chosen such that excessive alignment forces are avoided during installation, but at the same time, allowable pipe stresses are not exceeded during operation. A particular problem arises with off-bottom deflections, because the pipe is floating several feet off the seabed during tie-in and after tie-in it will come to rest on the seabed thus causing a potentially large change in pitch angle. If this differential cannot be absorbed, for example by using a ball shaped mechanical connector which can absorb some angular misalignment, large vertical moments can result. The amount of yaw misalignment depends very much on the tie-in method used. If a first-end direct pull-in is used, the yaw misalignment is generally very small, resulting only from the lateral offset of the surface lay vessel, combined with the action of lateral currents. If the pipe is allowed to make contact with the seabed before being pulled-in, the amount of lateral misalignment can be reduced to within a few degrees, provided one can measure the orientation of the pipe axis close to the pipe end relative to the azimuth of the tie-in structure. Two transponders placed along the axis of the pipe, or a simple television camera placed in such a way as to monitor the alignment of the pullcable at the wellhead, will provide the required information. As discussed earlier, when a pipeline or FLB is connected using the lateral deflection method the lateral angular misalignment of the pipe end is directly related to the accuracy of initial placement of the pipe end inside the target area. Large angular misalignments (possibly up to 10 deg or more) are possible, and very little corrective action can be taken after the pull-in operation has begun. Rotational or roll misalignment is mainly caused by : (a) (b)

plastic deformation or "doglegs" in the pipe developed during laying or pull-in, which may cause unpredictable rotation often in one direction and then the other; and unequal bending stiffness in different bending planes of an FLB, which results in an interaction between bending and torsion.

In case of a single-cable pull-in with a bullnose as pipe-end fitting, there should not be a great problem getting the bullnose into the alignment funnel. However, problems may arise if the pipe must be rotated into alignment, as this may require large alignment forces and can create high stresses in the pipe itself. If a two-cable pull-in is used, with the pull cables attached to two guidepins, it is possible that the guidepins will not even align with the guidecones on the alignment frame because the required amount of torque to rotate the pipe can only be-applied by the cables itself, which may not be sufficient to rotate the pipe into alignment. Rotation of the pipe, so that the pipe end fitting does not arrive in the upright position at the alignment structure, can thus be a potentially serious field problem. A large portion of the tests with two-cable pull-ins was devoted to investigating rotational alignment problems. Pipe alignment model tests (scale 1:10 to 1:12) were conducted in our machine shop (Figures 9, 10) and in the wave tank. Conclusions drawn from these tests have been reported in Section IIIG. The following paragraphs discuss some of the considerations for design of pipe end fittings in more detail. Bullnose Design Figure 23 shows a simplified model of the forces acting on a typical bullnose. If the bullnose is aligned using a pull-in cable like shown, the pull force in the cable is the sum of the back tension T in the pipe and a term which depends on the moment M required to align the pipe. Contact between bullnose and guide funnel is idealized by forces acting at two discrete points, a distance L apart. The contact forces at those points create : (a) (b)

a couple sufficient to provide the alignment moment M; and friction which must be overcome by additional pull force.

If we neglect the lateral force R in the pipe, the following equation gives an estimate of the maximum cable pull force P : P = T + 2 f(α,µ) M/L

(8)

f(α,µ) is a function of the angle α of the bullnose and the friction coefficient µ between the surfaces of the bullnose and the guide-funnel. For values of µ ranging between 0.15 and 0.40, and α between 15 and 20 deg, f(α,µ) ranges between 0.4 and 0.7 (the larger µ or α, the larger the value for f(α,µ).

Sled Design Typical models of a two-pin sled and guidecones were fabricated (Figures 10, 24). Using a twocable pull in can reduce alignment forces compared with a one cable bullnose pull-in. The sled alignment tests proved very valuable in helping to visualize what happens during the alignment process. Figure 24 shows a plan view of a typical sled during correction of lateral misalignment. Initially a potentially large alignment moment can be applied due to the large moment arm ( = distance "a" between the guidepins). However, as soon as both guidepins contact the walls of the funnel, other mechanisms will start to act. During this intermediate alignment the clearance "c" between the guidepin and guide funnel is a primary determinant of the cable pull force (one can see the analogy with the pull-in of a pipe into a J-tube). Then, during final alignment, it depends very much on the particular dimensions of fitting and guidecone, whether complete alignment is obtained. The two graphs (prepared for typical prototype dimensions) show the dramatic difference in pull cable forces caused by just a 1/4-inch change in pin/guidecone clearance. A potential field problem may be caused by the fact that one might decide that the pipe end is aligned once a sharp increase in pull force is experienced. As the example graphs show, however, the maximum force may occur when the alignment is only 5O% complete. One way to avoid this clearance problem is not to use the customary cylindrically shaped guidepins, but to use cone-shaped guidepins and guide funnels, so that the sled merely combines two, narrow bullnoses.

Figure 23. Bullnose Comparisons

Figure 24. Two-Pin Sled Alignment

Pipe rotation can present great difficulties for a two-pin sled pipe end fitting. Figure 25 shows several stages during rotational alignment of a sled. The pipe was initially rotated over 180 deg. Buoyancy on the sled was simulated by a weight hanging from a cable which ran over an overhead pulley down to a pin attached to the sled. Note that the buoyancy provides maximum benefit at 90 deg rotation, but that this benefit decreases as the pipe is rotated into alignment. Until the guidepins make contact with the guidecones, the only other correcting moment is applied by the pull cables, and this may not be sufficient. There is a critical radius of the guidecones which is related to the distance between the guidepins and the amount of pipe rotation. If the radius of the guidecones is less than this critical value, the guidepins will overshoot the guidecones and no alignment is possible, no matter how hard one pulls. This is the reason that the guidecones shown in Figure 25 are so large (3-in. diameter or 3 ft on prototype scale) compared with the initially used guidecones [Figure 11 (a)]. Unless pipe rotation can be avoided or is small (less than about 15 deg) use of a two-cable pull-in can create significant field problems. If a two-cable pull is necessary anyway, one of the more practical solutions will be to allow the pipe end fitting to initially rotate freely around the pipe axis. After the fitting is aligned and locked, a mechanical device would then rotate the pipe, if necessary. In critical cases field measurements may be necessary, both prior to pipe installation as well as during installation. Relevant parameters of the alignment structure are height of the seabottom, and tilt angle and azimuth of the alignment funnel. These parameters can be determined prior to the tie-in, so that it is still possible to adapt the geometry of the pipe and fitting before offshore installation. Attitude of the pipe during tie-in can also be measured if alignment is a critical issue (for instance, by installing inclinometers on the pipe end fitting). Finally, visibility at the location of the tie-in point may be poor, so acoustic measures to back-up visual observations of alignment operations may be necessary.

Figure 25. Rotational Alignment of Two-Pin Sled with Large Guidecones

A. Sled Approach: 180 deg. Rotation with Buoyancy

B. Guide Pins Reach Cones

Figure 25. Rotational Alignment of Two-Pin Sled with Large Guidecones

C

D

E

F Final Stage of Alignment

B.

REPAIR CONSIDERATIONS

Pipelines or flowlines may fail during either installation or operation, from a variety of causes. During installation, damage can be sustained by improper handling, by loss of control of vessel position, or by overstressing due to equipment failure. During the operation of an offshore pipeline, failures may be caused by contact with anchors, or fishing gear (third-party damage), currents or mud slides, manufacturing defects, corrosion, or fatigue failure resulting from earlier undetected damage. Although much experience exists in repairing damaged pipelines in water depths within diver range, little or no experience exists beyond diver depth. However, by designing the ability to repair deepwater pipelines into the installation vessels, the pipe terminations, and the pipelines themselves, future repair activities could be simplified considerably. The repair options available for damage sustained during installations will depend on the installation method used and the capabilities of the lay vessel. For S-Lay, J-Lay, and Reel vessel installations, the simplest repair option is to "reverse-lay" the pipe back onto the lay vessel to the point of failure, and then to resume pipelaying after the damaged section has been removed. This "reverse-laying" technique can always be used for "dry damage", meaning damage that has deformed the pipe but has not allowed any ingress of water. On the other hand, "wet damage", meaning damage such as a wet buckle where the pipe wall fractures and water enters the line during installation activities, may be impossible to recover onto the lay vessel. Ideally, for surface-lay installation techniques, the vessel tensioner, stinger, and A&R winch should each have the necessary capacities to handle a flooded pipeline. If this is the case, then the repair can be effected by "reverse-laying" back to the damaged section as before. These requirements are not as severe for vertical J-Lay vessels, which can easily handle a heavy, sagging, suspended pipe span, but may be very severe for an S-Lay vessel since stinger loads in particular will be high. In the event that loads become excessive, or the pipe breaks in two, the line must be abandoned. Once abandoned, repair operations can be performed by deploying a tool from the lay vessel which cuts the line, allows it to be dewatered, and attaches a lifting cable for recovery by the A&R winch (Figure 26). Once back on the lay vessel, normal repair operations can continue. One such repair tool has already been developed and demonstrated in 2000 feet water depths. Extending the capability to even deeper waters is considered technically feasible. For a pipeline damaged during installation by a tow method, the repair technique also will depend upon whether the pipeline damage is wet or dry. For dry damage the pipeline should be towed back into shallow water, preferably to a location near the launch site, for repairs. Wet damage suffered during the tow operation probably will result in a pipeline that no longer can be positioned by the tow vessels. In this event, the line(s) must be abandoned in place, to be repaired by other vessels, using procedures described below for operating lines. The method used to repair a damaged operating pipeline will depend on the location of the failure. The repair method requiring the least development appears to be the surface-lift method, which involves cutting the damaged line, and (after adding pipe to one or both ends) lifting both ends to the surface. The two ends are then joined by a flexible spool or swivel and the repaired pipeline is laid down with an offset to one side, thus absorbing the slack (Figure 27). The design of the flexible spool will be critical to the control of pipe stress during the lowering operation, as will be the methods used to monitor the laydown operation. Should congestion or other conditions on bottom not allow space for the lines to be laid down to one side, then mechanical connectors can instead be attached to each end of the raised line, and the final connection can then be made subsea. The tools required to perform these subsea repair operations need further development, particularly for the on-bottom connections. It should be noted that the repair vessel should include sufficient lifting capacity, a precise dynamic positioning system, capability for subsea tool handling, and some pipe laying capability.

Figure 26. Repair during Installation - Abandon, Cut, Dewater and Recover

Figure 27. Repair to Operating Line - Straight Lift, Surface Swivel and Layover

The alternative to these surface repair options is a fully automated on-bottom repair system. Considerable work has been done on such repair systems both in Europe and the USA, starting with Shell's Submersible Pipeline Repair System design in the mid-1970's and culminating with actual systems being built and tested by Saipem (Italy), Comex (France), and Supra (Germany). While a great deal of progress is being made on these remotely operated repair systems, none of the systems under development has yet been demonstrated in a fully automated diverless mode. The performance of these very complex repair systems are considered to be most sensitive to equipment malfunctions. Much additional work will be needed to make these methods practical, especially for Gulf of Mexico conditions of soft bottoms and poor or zero visibility. Damage located less than one water depth from the termination point can be repaired by a fully automated on-bottom repair system, or by a surface lift method which abandons the pipe section between the damage and the tie-in point. After cutting and lifting the pipeline, the line is then relayed to the termination point, where it must be re-terminated. To avoid the need to remove the short damaged section remaining from the old line, it is advantageous to have alternative termination points available. Also, it is advantageous to design the termination hardware such that the re-termination can be made without the need for special equipment or vessels other than those used for the pipe repair. Of course, one other option exists for "repair" of a pipeline under these circumstances, namely replacement of the entire pipeline. Some other considerations for the successful performance of deepwater repairs include the development of tools capable of quickly detecting leaks and various other mechanical damages to lines. These tools are needed both for installation activities and during oil/gas production operations. Once detected, it is important to be able to quickly find the exact point of the failure. This task becomes complex in areas where the lines may be buried, or where visibility is poor, and will become especially difficult in very deep waters.

C.

LOAD LIMITING DEVICE

Langner (75)

Offshore structures must be protected against excessive tensile loads from pipelines or flowlines. These loads can arise from several sources, primarily mudslides, earthquakes, and other soil movements, and loads from anchor hooking by ships, construction barges, and other large vessels. Thermal and pressure expansion loads also can present problems under some circumstances. Overload protection can be provided by general geometry of the pipe as it is laid on the seabed or by installing a local weak link in the pipeline. Probably the simplest means to protect an offshore structure from large tensile loads developed by pipe movements (or compressive loads due to thermal expansion) is to lay the pipe along a curved route near the structure. Tension loads then will be absorbed in the process of straightening out the slack in the curved route. This same curved route also provides a means to absorb thermal and pressure expansion loads. For example, a FLB route having a 90-180 deg turn of a 200-1000 ft radius, located perhaps a half mile away from a subsea wellhead, would protect a wellhead from most mudslides and anchor hooking loads. Experiments conducted at WRC by TRED, have shown that a straight pipeline hooked by an anchor probably would buckle at the hooking point and may eventually also fail at that point if the pulling load continues to increase. However, it was found that the breaking load of the buckled pipe could be as high as the ultimate tensile strength of the pipe. Subsequent experiments on special breakaway loops in the pipeline demonstrated simple means of providing a weak link which can significantly reduce the failure load at the hooking point.

Another option for overload protection is to install safety joints, which are pipe couplings designed to pull apart at a specified load, independent of pipe pressure, temperature or bending moment. About 150 commercial safety joints have been installed offshore and some 15 separations and subsequent pipe repairs have been reported, with separations mostly due to mudslides in the Mississippi delta area of the GOM. However, at least two inadvertent safety joint failures have been reported at tension loads well below their design loads. These latter failures point out a need for improved reliability in the mechanical safety joints, since such a separation would be very expensive to correct in deep water. D.

BUNDLING OF FLOWLINES

Langner (72)

In the past, flowline bundles have been installed utilizing a variety of bundling techniques, including : (a) (b) (c) (d)

lines laid separately, either one at a time or multiple lines laid simultaneously; parallel lines strapped together discretely at intervals, usually with steel banding material; parallel lines wrapped together continuously, usually with pressure sensitive tape; and lines placed inside a carrier pipe and then towed into position.

The first three bundling methods provide simplicity and low cost and are compatible with the various lay methods. The carrier pipe method adds buoyancy and stiffness to the bundle and offers excellent protection of the flowline bundle when the carrier pipe is water-filled on-bottom after installation. Problems may be encountered with multiple flowlines or FLB's as they are pulled-in and connected to a subsea structure. These potential problems include : (a) (b) (c) (d)

dynamic impacts between the lines, if laid separately; possible buckling of the smaller lines if the pipes are banded together or discretely wrapped; possible overstressing of some of the lines due to non-uniform sharing of the pull tension, and potentially large torque required to orient the flowline termination assembly before attachment at the wellhead. Continuous wrapping of the flowlines provides a solution to only the first two problems.

Placing the bundle inside a carrier pipe also has some problems and disadvantages. The method is relatively expensive and the increased bending stiffness is not desirable in certain situations. Furthermore, while this technique has been used successfully numerous times in shallow to moderate water depths, it will be more difficult to implement in very deep water, owing to the contradictory requirements for thin-wall to provide buoyancy and thick-wall to provide collapse resistance. These requirements can be met by internally pressurizing the carrier pipe with compressed air or nitrogen and then releasing same after installation. However, there may be safety concerns associated with handling such large volumes of compressed air, and there are risks associated with leaks and consequent premature loss of buoyancy. Another alternative bundling technique is to form the several individual pipes into a permanent rope-like helical configuration, prior to laying the flowline bundle offshore by the reel or tow method. This bundle may optionally be left as a simple spiral bundle, or wrapped continuously with tape, or embedded within an extruded plastic sheath, for added mechanical protection. A number of advantages can be obtained by employing such a helical flowline bundle. These are discussed below. The combined weight, stiffness and close proximity of the various pipes, give a helical bundle greater strength, integrity, and protection for the various pipes in a bundle than is possible if the lines are laid separately. Any problems associated with lateral buckling of the smaller pipes, due to either bending or thermal pressure expansion, which can occur for a straight-pipe discretely wrapped bundle configuration, are eliminated.

Composite beam behavior is encountered when a bundle is made by tightly wrapping together several parallel pipes; such composite beam behavior is eliminated for a helical bundle. Bendingtwisting stiffness of a helical bundle is simply the sum of the stiffnesses of the individual pipes. Hence, the bending moments and torques required to align the flowline termination assembly with the receptacle on a wellhead are smaller for a helical bundle than for a bundle of wrapped parallel pipes, using the same pipes. Spooling and laying of an entire flowline bundle as a single pipeline from a single reel becomes possible with a helical bundle. Thus there is no need for multiple reels, straighteners, and tensioners on the reel vessel. Shell has previously developed and patented a number of methods suitable for forming pipes into a helical flowline bundle. Interested parties can contact Shell's Patents and Licensing Division for further information. E.

BUOYANCY CONSIDERATIONS AND SUBMERGED WEIGHT CONTROL

Kopp (74)

Providing additional buoyancy to the pipe, for instance to support a section of the pipe off the seabed during a lateral deflection, is expensive. The material cost of buoyancy for deep water is high as is the cost of buoyancy removal after completion of the tie-in. Therefore, the trade-off, either use of buoyancy with resulting lower pipe bending stresses and pull-in forces and more complex operations, or a simpler on-bottom deflection with higher bending stresses and pull-in forces, must be carefully weighed in each individual case. Careful control and measurement of pipe submerged weight in case of on-bottom and off-bottom pull-ins are essential, especially with large diameter pipelines. The submerged weight of a large diameter pipe is the difference between two large quantities, the pipe weight in air and the water displacement of the pipe, and relatively small variations in either one, can result in large variations in pipe submerged weight. Measures which may have to be taken include : (a) (b)

specifying tighter pipe weight tolerances than given in the API Specification API 5L for Line Pipe and actually determining pipe submerged weight during pipeline construction. The latter involves measuring pipe outside diameter of each joint, and using mill data on joint length and steel weight to determine the most likely pipe submerged weight. Even then, some uncertainty remains, for example due to small variations in seawater density. If pipe submerged weight deviates from an acceptable range, trimming operations will be required and these will be time consuming.

If buoyancy is provided to the pipe and some type of surface lay method is used, difficulties may arise with passing the pipe through a stinger. Contingency plans need to be developed to cover accidental release of buoyancy during descent. Innovative methods have been proposed to reduce the cost of deepwater buoyancy. An example is to provide the buoyancy only when actually needed by lowering air bags from the surface down to the pipe, and filling these air bags using air supplied from the surface. Such a method may be practical in a low current environment, but a heavy burden is placed on ROV assistance and extensive field testing is required before pipeline construction. Contingency plans must be developed to handle either leaking or accidental overfilling of the air bags.

F.

CONCEPTS FOR VERY DEEP WATER

Langner (76)

Planning for possible production from Shell's very-deepwater Atlantic wells, which were drilled during 1983 and 1984, included studies of how to install, connect, and repair FLB's to these wells. Many of these studies are applicable to the range of water depths considered in this program (1,500 to 3,000 ft). Reference 76 is a book of drawings which illustrates many of the concepts considered by Shell for installing FLB's in water depths to 7,500 ft. Also illustrated are means of connecting catenary risers to an FPS. The reel method, using the existing reel ship APACHE or a similar vessel, and the J-lay method, utilizing a converted DP drilling ship, were identified as the most promising methods of installing long flowlines in these great depths. Three methods of connecting flowlines to deepwater subsea wells were found to be feasible : 1)

first-end one vessel stab and hingeover;

2)

second-end one vessel flowline pull-in; and

3)

first-end two vessel stab and pull-in.

The first two methods were evaluated in this joint industry program. Besides showing installation and tie-in concepts, the drawings in Reference 76 also show equipment and component concepts for different stab and hingeover equipment, bullnose type pipe end fittings, flowline bundle configurations, repairs, vertical stingers for J-lay, and catenary risers.

XII.

RECOMMENDATIONS FOR FURTHER STUDY

The work conducted in this program is sufficiently detailed to use in preliminary evaluations of field development scenarios. Once specific tie-in methods have been selected as prime candidates, more detailed evaluations can be initiated. The following summarizes the main areas of further study and development recommended for tie-in methods considered in this program. Ultimately, the selected method will have to be subjected to full scale tests before offshore field application. Design Aspects of Pipelines and Flowlines Some of the simpler and most promising deepwater tie-in methods will require accepting bending strains in the pipe which are greater than yield. A comprehensive overview study is needed to explain interactions between external pressure, pipe bending, and axial forces, and subsequent internal operating pressures in the pipe, and to provide guidelines for selecting safety factors against pipe buckling and collapse in deepwater. Effects of cyclic loading must also be considered in this study. Deepwater Pinelay Methods The vertical lay (J-lay) method has been identified as being required for most deepwater tie-in scenarios. At present there is no operational equipment to vertically lay pipe. As discussed in Section 76, however, the technology for vertically laying pipe is available, so the next step is to actually build and test the required equipment. Depending on the application, the lead time for such development may ranges from a few months to well over a year. Steel Catenary Risers Recommendations for further work include : (a) (b) (c) (d)

study of the effects of currents, vortex-induced vibration, and out-of-plane displacements; more accurate prediction of fatigue life by inclusion of second-order, long-term drift motions of the floating production facility; further study of the nature and location of the top riser support; and study of the influence of direct wave loading on the fatigue analysis of the riser near the surface. Moderate scale offshore field tests will eventually be needed to confirm the analysis results.

Since SCR's also look promising for applications at TLP's or compliant towers, installation methods for those structures must be studied also. If mooring lines are absent at those structures, installation may even be simpler than installation at an FPS. On- and Off-Bottom Pull-ins Four main areas which require further study and development are : (a) (b) (c) (d)

towing a pipe into a target area; ROV operations during deflection and connection; analysis of elasto-plastic pipe behavior; and diverless mechanical connectors which can accept angular misalignments.

Towing a long pipe string into a small target area can cause significant target over- or undershoot problems. Further analysis and large scale field testing is required to confirm our preliminary findings and develop solutions to avoid these problems. Pull cable handling is one of the most critical tasks during bottom pull-ins. Most ROV's today lack the capability to reliably transfer, connect, and disconnect larger size pull cables (say greater than about 1/2inch). The use of pull cables made of synthetic fibers may be promising, but this will require extensive testing and development of cable end fittings. Preliminary finite element analyses of elasto-plastic behavior of the pipe during a simulated lateral pullover, proved to be very cumbersome and time consuming. It may be possible to develop relationships which allow us to apply correction factors to linear elastic solutions to obtain the corrected elasto-plastic result.

Test and analysis results showed that even under almost ideal conditions, fairly large angular pipe end misalignments may occur at the tie-in structure. Some vendors have developed single pipe mechanical connectors which can accept angular misalignment prior to being locked, but no diverless commercial applications have yet been reported. Flowline Stab and Hingeover There is a need for further work in primarily three areas : (a) (b) (c)

axial orientation of the flowline during stab-in; control over position of stab-in tool during stab operations; stab-in operation simulator.

A practical method must be found to provide axial orientation of the flowline during stab-in. either the flowline will have to be rotated at the surface, or some means must be found to orient the stab-in tool at the bottom. The latter may be combined with an active system to control stab-in tool movements, such as thrusters, so that a stab-in can be made relatively independent of surface vessel movements. A simulator system may be developed which allows the operator on the lay vessel to stab-in the flowline by simply. moving a joystick while simultaneously observing the operation using both visual information (video cameras) as well as other positional reference systems (acoustics, for example). The simulator system could also be used as a training tool. REFERENCES (1)

Langner, C. G., Suspended Pipe Span Relationships, ASME, Proc. Third Int. Offshore Mechanics and Arctic Engineering Symposium, New Orleans, pp. 552-558, February 1984.

(2)

Murphy, C. E. and Langner, C. G., Ultimate Pipe Strength Under bending, Collapse, and Fatigue, ASME, Proc. Fourth Int. Offshore Mechanics and Arctic Engineering Symposium, Dallas, pp. 467473, February 1985.

DEEPWATER PIPELINE, FLOWLINE AND RISER INSTALLATION

CORPORATE PARTICIPANTS Amerada Hess Corporation

Saga Petroleum a.s.

Amoco Production Company

Saipem

Brown & Root, Inc

Santa Fe Offshore Construction Co

Chevron Corporation

Shell Development Co.

Conoco, Inc

Single Buoy Moorings of America, Inc

Elf Aquitaine

Smit International Marine Svsc (USA)

Exxon Production Research Co.

Standard Oil Production Co.

J.P. Kenny Offshore Engrg., Inc

Statoil

Mc. Dermott International, Inc

Sub Sea International Inc

Mobil Research and Development, Inc

Sun Exploration and Production Co.

Murdock Engineering Company

Tecnomare S.p.A

Norsk Hydro a.s.

Transcontinental Gas Pipeline Corp.

Petrobras

SHELL CONTRIBUTORS

R.R. Ayers, Program Manager F. Kopp, Assistant C.G. Lagner R.H. Orlean R.W. Patterson J.O Esparza, C. Ester*. S. Huffer*, B.M. Kuciemba, T.E. Long, D. McMillan*, J.F. Nimmons, R. Osborne*, H.M. Riggins, J.E. Smith, D.P. Thompson

______________ * Contract Personnel

CONSULTANTS Southwest Applied Mechanics, Inc

REVIEWED :

R.R. Ayers

APPROVED :

E.A. Milz

TECHNICAL PROGRESS REPORT - DISTRIBUTION

Shell Development Company 2-

Technical Files - BRC, Houston

Shell Oil Company 1-

Sr. Vice President, Exploration, Head Office, Houston

1-

General Manager, Drilling & Producing Operations, Head Office Production, Houston

1-

General Manager, Engineering, H.O. Production, Houston

8-

Manager, Civil Engineering, H.O. Production, Houston

5-

I&LS - Reports, Head Office, Houston

1-

M&T - Engineering - General Engineering Manager

1-

M&T - Engineering - General Engineering - Marine Project Manager

1-

Assistant General Counsel, Patents & Licensing. Legal, Head Office, Houston

Shell Pipe Line Corporation

1-

Manager, Pipeline Operations, Regs. & Maint. Stds.

1-

Manager, Pipeline Operation, Gulf Coast Division