1-s2.0-S0029801815006733-main

Ocean Engineering 112 (2016) 153 –17 1722

Contents lists available at ScienceDirect at ScienceDirect

Ocean Engineering journal homepage: www.elsevier.com www.elsevier.com/locate/oceaneng /locate/oceaneng

Simulated in-line deployment of offshore rigid �eld joint – A testing concept Ahmed M. Reda a,b, , Ali Mothana Saleh Al-Yafei a, Ian M. Howard b, Gareth L. Forbes b, Kristoffer K. McKee b n

a b

Qatar Petroleum, Doha, Qatar Department of Mechanical Engineering, Curtin University, Perth, WA, Australia

a r t i c l e

i n f o

Article history:

Received 17 September 2015 Accepted 8 December 2015 Keywords:

Offshore � eld joint Subsea power cable In-line deployment Radial water penetration test High voltage alternating current (HVAC) cable Deployment simulation

a b s t r a c t

Failure of submarine power cables have been shown to be attributed to cable � eld joints in 18% of cases. This high failure rate at the joining location indicates that the current acceptance testing of these joints is inadequate. The failure mode of these joints is believed to be entirely due to water ingress at the � eld joint location. Submarine cable � eld joints are required during installation due to cable repair, joining of insuf �cient manufactured cable lengths or when the lay process has to be abandoned due to the sea state. Current Current design design guidance guidance for cable �eld joints suggests sea trails to determine if the proposed � eld joining technique is acceptable. Sea trails, however, are often of ten prohibitively expensive, such that a set of standardised onshore testing regimes which improves both the reliability and affordability of these tests would be advantageous. This paper outlines a recent cable � eld joint onshore testing regime to ensure the cable joint integrity during the laying process as well as serviceability in operational life. The paper outlines a process for simulation of the calculated cable laying tension and bend radius with a set of physical tests developed for mechanical and water ingress protection. & 2015 Elsevier Ltd. All rights reserved.

1. Introduct Introduction ion

According to Worzyk (2009) a study on subsea power cable failur failures es was undert undertake aken n in 1986 by CIGRÉ CIGRÉ (Confe (Conferen rence ce Inter Inter-nationale des Grandes Reseaux Electriques). This study indicated a typical failure rate of 0.32 failures/year/100 km. Furthermore, 82% and 18% of the failu failure ress occu occurr rred ed in the the cabl cables es and and the the joint joints, s, respecti respectively vely.. The study did not specify specify the exact failure failure causes, causes, however, the three major fault causes for submarine cable joints were were know known n to be: be: 1) Inad Inadeq equa uate te joint joint desi design gn,, 2) Poor Poor joint joint assembly work onboard the vessel, 3) Adverse weather conditions during jointing, 4) Inadequate installation procedures. The other failures to subsea power cables are due to many other factors such as �shing, anchors, dredging and other activities. According to the International Cable Protection Committee (2009) anchors (2009) anchors represent the largest portion of submarine cable dama damage ges. s. Cont Contac actt betw betwee een n a cabl cablee and and an anch anchor or is often often Corresponding author at: Qatar Petroleum, Doha, Qatar. Tel.: þ 974 5576 5276; fax: þ 974 4013 9058. E-mail address: [email protected] [email protected] (A.M. (A.M. Reda). n

http://dx.doi.org/10.1016/j.oceaneng.2015.12.019 0029-8018/ & 2015 Elsevier Ltd. All rights reserved.

disast disastrou rouss as the forces forces applied applied by a movin movingg anchor anchor may be extremely high. The anchoring hazard may result from:

 Emergency anchoring (where an anchor is deployed to prevent collision or grounding).  Negligent anchoring.  A vessel being anchored inadequately and a resultant dragging session.  Accidental anchoring (where an anchor falls unexpectedly from a vessel due to equipment failure or operator error)  Insuf �cient protection for the cable  Component damage Fig. 1 shows a proportion of cable faults by cause, from a database of 2162 records spanning 1959–2006. It can be seen seen from from Fig. 1 that cable failure failure component componentss represents 7.2% of the statistical distribution of damages. Worzyk Wor zyk (2009) indicat indicated ed that that while while many many failure failure statis statistic ticss account for failures during operation, the statistics normally do not include damage to the cable that happens before commissioning. Cable damages during the installation installation might call for expensive expensive and time-consuming repair operations.

154

A.M. Reda et al. / Ocean Engineering 112 (2016) 153–172

Fig. Fig. 1. Proporti Proportion on of cable cable faults faults by cause Source: Tyco Telecommuni Telecommunicatio cations ns (US) Inc.

Worzyk (2009) stated (2009) stated that “The failure of some early installation joints during service shaded the reputation of submarine power cable joints. Failures in the joints were usually caused by poor engineering or inadequate installation procedures. ” CIGRÉ study (Update ( Update of Service Experience of HV Underground and Submarine Cable Systems, 2009) 2009) revealed that there were only four joint failures out of 49 failures in total in 7000 km of installed submarine cable. The study stated that 19 of the 49 reported faults could be repaired within one month. The ratio of joint failures changed from 0.22 to 0.095 failures/year/100 failures/year/100 km, from from 1986 to 2009 2009,, respec respectiv tively ely as per per Worzyk Worzyk (2009) (2009).. This This demons demonstra trate tess that that the design design of submar submarine ine cables cables has bee been n improved over the years and the cable joints are safer and more reliable today than in 1986. CIGRÉ (2010) & (2010) & Worzyk Worzyk (2009) highlighted (2009) highlighted that manufacturing of a reliable joint is often the most dif �cult undertaking during the development. During the installation or the deployment of the joints, they should withstand the mechanical stress and strains. CIGRÉ TB490 (CIGRÉ (CIGRÉ TB490) TB490) emphasized the importance of paying attention to repair joints as part of the AC submarine cable system. DNV-RP-J301 system. DNV-RP-J301 (2014) indicated (2014) indicated that all joints and terminations shall be subjected to a test program in accordance with the applicable applicable standards standards.. Since subsea subsea installatio installation n �eld joints joints and repair joints connect the cable parts along the cable route to form one integrated cable, the joint has to withstand all the expected differ different ent loads during during the service service life the same same as the cable (Karlsdóttir, 2013). 2013). This paper focuses focuses only on the stiff joints which have a rigid outer casing. This rigid joint serves as a connection point for the armouring wires of each cable end. The deployment of a rigid joint on the sea bed is probably the trickiest and most complicated operation of the cable installation. This is in part because during the deployment operation the two jointed cables must be handled with the rigid joint. Neither over-bending nor over-tensioning must occur or the cable arrangement being stuck in other structures on board Worzyk (2009). (2009). The deployment of the rigid joints requires a complicated crane arrangement due to the stiffness of the joints as well as the increased diameter of the joint compared to the cable. The stresses experienced experienced by the joints during the deployment on the sea bed are the likely maximum stresses experienced by the joint during the service life. Therefore, it is important to verify the mechanical integrity and reliability of the rigid joint during deployment. Jointing operation is challenging and requires valuable vessel time, good planning, highly quali�ed personnel, proper equipment for deployment, jointing facility container load on the vessel as well as good coordination between the jointing crew and the vessel crew. Jointing operation typically requires from one day to several days, depending on the cable joint and joint design, by which which a good good weathe weatherr windo window w is requir required ed to comple complete te the

jointing operation. Cable and joint repair is impossible during storm seasons as it is dif �cult to �nd suitable weather windows that hold up long enough. Good weather conditions are essential to ensure the workmanship workmanship of the joint and to obviate obviate cable fatigue damage of the hanging cable sections. It is essential to ensure the reliability of the offshore installation installation joints and repair repair joints. This is in part because any failure in the joint could lead to blackouts in offshore platforms resulting in �nancial and reputation impacts on the offshore platform operators. It should be highlighted that “installation joint”, “�eld joint” or “repair joint ” denotes a joint of the complete submarine power cable including the conductor, insulation system, armoring and all other intermediate layers. The rigid joint often requires complicated rigging arrangements for the deployment. This is in part due to the joint stiffness and increased diameter. Subsea cable offshore rigid � eld joints have to be designed and correctly installed. If not, the offshore �eld joint will represent weak points and frequently turn out to be the only source of seawater ingress. This seawater ingress will subsequently lead to electrical failures. For reliability, offshore � eld joints should be avoided where possible, as they are a potential source of failure. However, in some situations it is impossible to avoid offshore �eld joints especially in the cases whenever the subsea cable becomes damaged or the cable laying operation must be temporarily stopped. For the offshore pipeline installation industry, abandonment of the laying operation takes place when the weather conditions do not allow pipelay activities to continue or due to unforeseen circumstances in the pipeline � eld area. In this situation, normally a temporary abandonment head is welded to the end of the pipeline. The tension is then transferred from the tensioners to the abandonment and recovery (A&R) winch and the abandonment of the pipeline can begin. The barge is moved ahead a suf �cient distance to allow the abandonment to hard rest on the seabed as illustrated in Fig. in Fig. 2. 2. However However,, the situation can be different different with the laying of subsea power cables. Depending on the kind of emergency, there will be different procedures according to time availability for sealing the subsea cable. However However,, the cable will still be cut for each of the possible possible scenarios scenarios.. Once the weather weather conditions conditions improve, improve, the recovery recovery procedure will be undertaken in the reserve in order to continue the deployment operation. Then, the in-line jointing will take place place using using an offsho offshore re �eld joint joint (OFJ). (OFJ). DNV-RP-J301 DNV-RP-J301 (201 (2014) 4) recommends that the repair joint should, if possible, be laid in line with the cable not within an arc. CIGRÉ TB490 (CIGRÉ (CIGRÉ TB490) TB490) de �nes the �eld joint as the joint which is made onboard the cable installation vessel between the cable lengths. Whereas, it de�nes the repair joint as the joint used to repair a damaged submarine cable or jointing two delivery lengths offshore. In principle there is no difference between a � eld joint and repair joint. The subject joint can be considered as either a � eld joint or repair joint. The subject joint shall only be used in the situation where a repair joint is required during the cable installation, and Omega laying, shown in Fig. in Fig. 3, 3, is not viable due to the seabed con�guration or crossing conditions. In the offshore industry, the U shape that is deployed on the sea �oor is also known as “Omega”. Electra No. 189 (Electra (Electra 189) 189) indicated indicated that the aim of the type test is to “qualify the design and the manufacturing of the cable system against the conditions of the intended application ”. As part of CIGRÉ TB490 (CIGRÉ ( CIGRÉ TB490), TB490), type tests were conducted on a 132 kV power cable with unit capacity of 100 MW. This is inclusive of offshore rigid �eld joints (OFJ) as shown in Fig. 4. 4. Recently, a type test for an offshore rigid �eld joint was repeated three times. In the �rst and the second type tests, the OFJ passed all the mechanical and electrical tests successfully. On the


155

Fig. 2. Abandonment operations.

Fig. 3. Omega joint deployment.

Fig. 4. Offshore � eld joint (OFJ).

contrary, the OFJ did not meet the criteria de �ned in CIGRÉ TB490 (CIGRÉ TB490) for the radial water penetration (RWP) test. Therefore, it was decided to undertake a trial laying test with OFJ in addition to the third RWP test as required for part of the type test. It should be mentioned that after the investigations using the results from type tests 1 and 2, it was concluded that the OFJ during the two type tests were not dismantled carefully. Moreover, stresses were introduced to the pre-molded joint during the release from the compound �lling before conducting the RWP tests. The OFJ failed the two RWP tests due to one of the premolded joints containing incomplete �llings at two locations of the copper housings. During the third test, the OFJ was dismantled cautiously and it was ensured that no additional stresses would be introduced to the pre-molded joint during the release from the compound � lling before the execution of the RWP test. Also, it was ensured that adequate measures were implemented to ensure that the copper housing of the pre-molded joint was � lled completely.

Water penetration tests can be conducted to measure the ability of the rigid joint to resist the water penetration up to the maximum water depth of the submarine joint. Water tightness is a crucial feature for a high quality power cable system. CIGRÉ TB490 (CIGRÉ TB490) recommends that actual sea trials be conducted to ensure the quality of the repaired joints. The loading conditions during the deployment of the rigid joint on the seabed are critical. Previously many joint failures occurred in the few days following installation/early operation. Hence, by controlling loading conditions during deployment operation, the joints do not have to be regarded as a weak joint anymore. Due to time constraints as well as logistic issues, it was decided to replace the sea trial tests with simulated on-land deployment. This was used to verify the mechanical integrity of the OFJ and to identify OFJ weak points under the deployment conditions. In order to determine the loads which should be applied on the OFJ during the on-land simulation, dynamic simulations using OrcaFlex software (Manual, 2014) were undertaken to calculate all the relevant loads/stresses expected during the over-boarding/ deployment procedure. OrcaFlex is a standard industry three dimensional non-linear time domain �nite element program speci �cally developed for marine dynamics and suited to the dynamic analysis modeling of cable catenaries. Similarly to the deployment simulation undertaken by this paper to mimic the deployment of the in-line rigid joint offshore, Woo et al. (2015) carried out an experiment to verify that the anchor collision caused no damage to the power cable covered by rock berm. Furthermore, Yoon and Na (2013) performed a safety assessment of mattress type submarine power cable protectors under the dragging forces of a 2-ton anchor through �eld tests on land.

156


This paper presents a new testing arrangement and testing procedure which can be used to simulate the deployment of inline rigid offshore �eld joints which is critical to the integrity of the OFJ. This focusses with particular interest on the weak point of the OFJ such as the plumbing point between the power cable metallic sheath and the copper tube of the pre-molded joint as shown in Fig. 5. The new testing arrangements employs the design loads, determined from the engineering simulations, to test the offshore rigid joint in air simulating the actual installation conditions. Once the load tests are completed, the joint including the plumbing area was subjected to the water penetration test. In the water radial penetration test, one pre-molded joint including plumbing areas was submerged in pressurized water for 24 hours followed by examination for seawater ingress. Upon completion of the radial water penetration test, a visual inspection was conducted for the plumbing area between the cable lead sheath and copper housing. This was to ensure that the plumbing area was clear of any cracks, or holes. The new testing arrangement can be simpli �ed to a load test of the joint in dry air and consequently performing the radial water penetration (RWP) test without loading. It is possible that the seawater ingress potential would be greater, while the joint is loaded and under external pressure, than the subject deployment simulation where the loads on the joint are applied in air and then the radial water penetration test is done without loading. Nevertheless, to overcome this issue during the radial penetration test, the external pressure employed in RWP corresponds to the maximum water depths along the pipeline plus 50 m. This is consistent with Electra 171 (Electra 171). Electra 171 (Electra 171) recommends to employ an external pressure in the RWP corresponding to the maximum water depths along the cable plus 50 m in case of the water depth less than 500 m and maximum water depth plus 100 m for water depths over 500 m. The new testing arrangement offers an alternative to subsea immersion testing for subsea cable joints and offshore deployment simulations. 2. Testing design process for deployment simulation

The testing design process for the deployment simulation involves a series of steps as shown in Fig. 6. 1. De�ne cable/ crane hoist/ eye-bolt/ joint acceptance criteria (i.e. allowable tension, minimum bend radius, crane hoist load,

eye-bolt load, allowable bending moment for the joint body). More details of this step can be found in Section 4. 2. Select the optimum layback and departure angle. In this step OrcaFlex (Manual, 2014) dynamic simulations should be undertaken to determine the layback which resulted in the highest dynamic workability (low tension / compression and curvature exceedance). Refer to Section 4.6. 3. Perform still water analysis at zero wave and current using OrcaFlex (Manual, 2014). The objective of this step is to investigate whether the acceptance criteria de�ned in step-1 are met with wide margin which then allows for wave/current action to be added. Refer to Section 4.10. 4. Perform dynamic simulations using OrcaFlex (Manual, 2014) to determine the limiting weather criteria. Refer to Section 4.11. 5. Perform a stress analysis using ABAQUS (Manual, 2012) for the OFJ body to ensure that the allowable bending is not exceeded during the deployment. In this step the bending moment from steps 3 and 4 shall be used. Should the stresses from ABAQUS be beyond the allowable stresses, the limiting weather criteria should be relaxed and the dynamic simulations run again to de�ne the relaxed weather criteria, as discussed in Section 4.12. 6. De�ne loads from dynamic simulations using OrcaFlex to perform the mechanical test on the OFJ. Refer to Section 5. 7. Conduct the mechanical test followed by visual inspection and RWP test as per Section 5. Should the radial water penetration (RWP) test and in-line test items not be acceptable after completing the mechanical test, then another joint should be prepared and the limiting weather criteria relaxed. If the radial water penetration test and in-line test items are acceptable then the OFJ design can be deemed to be acceptable. 3. Offshore � eld joint inline deployment procedure

This section highlights the steps which should be used to deploy the in-line joints as an integral part of the emergency contingency procedure during cable installation. Figs. 7–10 show some of the required steps and a brief description of the operation. Fig. 7 shows the use of the vessel crane to lift the joint outboard. The tensioner pays out the cable to ensure suf �cient slack during operation. The joint is then lowered through the splash zone area. Fig. 8 shows that as soon as the joint is fully submerged, the tension on the crane wire will be removed. The crane wire will only be detached from the joint when the joint reaches the seabed. Until then nominal tension shall be maintained to ensure that the

Fig. 5. Plumbing area after in-line test and water penetration test.


157

Fig. 6. Testing design process.

Fig. 7. Lift the joint outboard. Tensioner to pay out cable to ensure suf �cient slack during operation.

crane wire does not become entangled with the joint. During this step, it is important to ensure that the departure angle is according to the dynamic simulations undertaken. Fig. 9 illustrates the lowering of the joint until it rests on the seabed. The remote operate vehicle (ROV) will con �rm proper set down. It is important to ensure that the departure angle is according to the dynamic simulations undertaken. The installation vessel then proceeds with the laying of the remaining cable as illustrated in Fig. 10.

4. Step 1 – De �ne Cable/crane hoist/eye-bolt/Joint acceptance criteria

Fig. 11 presents the cable lay conventions which are used in the paper. The results presented are parametric with respect to the

departure angle, layback length, cable, and chute and touchdown point. The simulations undertaken in this section were made using OrcaFlex (Manual, 2014). Fig. 12 illustrates the environmental conventions used throughout the simulations. The wave direction is also illustrated in Fig. 12. The environmental conventions are indicated as follows:

 Stern wave (0 degree)  Starboard (90 degrees)  Head wave (180 degrees)  Portside (270 degrees) 4.1. Cable data

The cross section and the mechanical properties of the submarine cable are highlighted in Fig. 13 and Table 1.

158


Fig. 8. Lower the joint through the splash zone area.

Fig. 9. Keep lowering the joint until it rests on the seabed.

Fig. 10. Installation vessel to resume the normal cable laying operations.

Fig. 11. Cable lay conventions.

159


Fig. 12. Environmental conventions.

4.2. Joint data

Table 1

The mechanical properties of 132 kV HVAC submarine cable.

The joint is comprised of the actual joint, armor pot and bend restrictor. The joint schematic is shown in Fig. 4. Tables 2–4 present the properties of the actual joint, armor pot and bend restrictor, noting that the joint is not a rigid body but a �exible member having prescribed bending limits. The allowable tension in the cable inside the bend restrictor varies with the angle of the bending. Table 5 illustrates the relationship between the bending angle and allowable tension for the cable inside the bend restrictor. 4.3. Crane lifting capacity

Table 6 highlights the maximum crane capacity limit used throughout the simulations.

Item

Value

Outer diameter 191 Weight in air 70 Weight in water 41 Axial stiffness 650 Bending stiffness 26 Allowable tension (Straight Pull) 160 Allowable tension (on Minimum Bend Radius (MBR) pull) 115 Allowable compression (Straight Pull)  17.3 Allowable compression (on MBR pull)  10.2 Minimum Bending Radius (For installation) 2.9 Allowable curvature (For installation) 0.345

Unit

mm kg/m kg/m MN kN m2 kN kN kN kN m Rad/m

Table 2

Joint body details.

4.4. Lifting aids capacity

Item

Value

Unit

Table 7 highlights the maximum eye-bolt and shackle capacity limits used throughout the simulations.

Length Outside diameter Axial stiffness Bending stiffness Bending moment Weight in air Submerged weight

5580 605 2680 88000 125 744 449.3

mm mm MN kN m2 kN m Kg/m kg/m

4.5. Wave data

A range of Hs (Signi�cant Wave Height) was applied in the simulations, varying from 0.5 m to 1.00 m for WD (Water Depth) ¼ 38.4 m. A corresponding realistic range of Tp (Peak Period) is p p de�ned by γ H s :13 o T p o H s :30 as highlighted in (NDGL ), which is the relation for wind driven seas. In the analysis, 3 Tp ’s are considered, shown in Table 8, under categories of Upper Bound, Best Estimate and Lower Bound Time Periods. The best estimate time period is an average of the upper and lower bounds. The JONSWAP (Joint North Sea Wave Project) spectrum was analyzed with the wave steepness value formulated from (DNV ) Section 3.5.5.5, shown below,

ffiffi ffi ffi ffi ffi ffi

ffiffi ffi ffi ffi ffi ffi

γ ¼ 5 for T p = H s r 3:6;

p ffiffi ffiffi

T T γ ¼ exp 5:75  1:15p p for 3:6 o p p o 5; H s H s



ffiffi ffiffi

and

ffiT H ffi ffiffi



ffiffi ffiffi

γ ¼ 1 for 5 o p p : s

The in�uence of the steepness factor over the Hs can be summarized as shown in Fig. 14.

4.6. Step 2 – Select the optimum layback and departure angle

An intensive analysis was undertaken to determine the optimum layback distance and the departure angle of the cable. The main criteria in selecting the optimum layback among the various options was to choose the layback which resulted in the highest dynamic workability (low tension/compression and curvature exceedance). Figs. 15 and 16 present the cable curvature and the effective tension in the cable respectively. The results in these �gures are extracted from OrcaFlex dynamic simulations undertaken at the same water depth and different layback lengths. It can be seen from Fig. 15 that the cable curvatures at layback lengths of 44 m (with legend shown on Fig. 15 as L44) and 50 m (L50) exceed the allowable curvature. Fig.16 shows that the compression generated in the cable at layback lengths of 44 m (L44) and 50 m (L50) exceed the allowable axial compression limit. Also, the same � gure shows that the effective tension for the layback length of 59 m (L59) was greater than that from the layback length of 54 m. Based

160


Fig. 13. Con�guration of 132 kV HVAC submarine cable.

Table 3

Table 7

Armor pot details.

Lifting aids capacity.

Item

Value

Unit

Item

Value

Unit

Length per side Outside diameter Axial stiffness Bending stiffness Bending moment Weight in air Submerged weight

358 291 3695 36400 180 210.7 142.5

mm mm MN kN m2 kN m Kg/m kg/m

Shackle Eye-Bolt

9.5 17.3

Tonnes Tonnes

Table 4

Bend restrictor details. Item

Value

Unit

Length per side Outside diameter Axial stiffness Bending stiffness unloaded Bending stiffness loaded Bending moment Weight in air Submerged weight Minimum Bend Radius

2129 291 650 26 4600 100 281.4 213.2 3.0

mm mm MN kN m2 kN m 2 kN m Kg/m kg/m m

Table 5

Bend restrictor allowable tension. Angle Allowable tension (kN)

0 15 30 45 60 75 90

120 116 104 85 60 35 20

Table 6

Crane capacity. Item

Value

Unit

Crane lifting capacity at smallest reach Crane lifting capacity at largest reach Crane block lifting capacity

13 4.9 30

Tonnes Tonnes Tonnes

Table 8

Wave data. Hs [m] Lower bound (LB) [s]

Best estimate (LB) [s] Upper bound (UB) [s]

0.5 0.75 1.0

3.24 3.97 4.58

2.55 3.12 3.61

3.87 4.74 5.48

on that it can be concluded that the optimum layback is 54 m. This layback length resulted in low applied tension to the cable and allowable curvature. 4.7. Initial set-up

Fig. 17 shows a snapshot from the OrcaFlex model used to calculate all the relevant loads and stresses expected during the actual installation. These loads were applied on the OFJ during the on-land simulation. The initial set-up can be envisaged from Fig. 18. In this example, the vessel, at its draft, was positioned so that the layback length was approximately 54 m for a water depth of 38.4 m. It is worth mentioning that the water depth of 38.4 m was selected as it represents the maximum water depth along the cable route. Deployment simulations are considered at water depths of 10 m, 20 m and 30 m. However, it is concluded that the stresses experienced by the � eld joint increases with the increased water depth. OrcaFlex models were developed of the cable, joint, winch and vessel. The Cable and joint were modeled as lines objects in Orca�ex using input data described in 4.1 and 4.2. The winch wire was modeled as a simple winch. The end point of the cable was anchored on the seabed, whereas the end of the cable on the vessel was free. The segmentation of the lines in OrcaFlex has a considerable in�uence on the accuracy of the results. As such a sensitivity study was carried out to establish how small the segments had to be set before the results converged. This is


161

Fig. 14. Effect of Gamma and signi�cant wave height on spectrum.

Fig. 15. Cable curvature calculated at different layback lengths.

Fig. 16. Cable tension calculated at different layback lengths.

particularly important for the stress. The timestep size can be very important to the accuracy of the results, but also has a signi �cant impact upon the time it takes to run a simulation. For this reason it is desirable to maximize the time steps, but without compromising accuracy or model stability.

4.8. Rigging set-up

In this step the in-line joint was lifted using a pulley arrangement where a sling was used to pass the joint over the pulley neatly into the crane hook location as shown in Fig. 19. For the

162


Fig. 17. Snapshot from OrcaFlex Model.

Fig. 18. OrcaFlex Model showing the starting point of the simulation.

OrcaFlex simulations, a 17 m steel sling was used. However, the selection and the details of the slings were based on the maximum dynamic force results. In the simulations, the pulley was held 7.6 m above the in-line joint at the taut sling as illustrated in Fig. 20. Note that at this position, the joint is still located on the vessel desk with no lift from the cable. 4.9. Crane operation and vessel maneuvers

In this step, the in-line joint was lifted by the crane and the vessel was moved. The manoeuvers were done in such way that:

 The allowable tension, minimum bend radius, axial compression of the cable, etc. are not exceeded.  There was no chance of immediate collision of the joint with the adjoining metal structures.

Fig. 19. Lifting arrangement.

The total movement of the joint is displayed in Fig. 21. Additionally, Fig. 21 highlights the cable shape at different time steps of the dynamic simulation undertaken using the OrcaFlex software. This �gure starts at the moment the crane lifts the in-line joint until the moment the in-line joint is laid on the seabed. It is worth


Fig. 20. Initial position of the crane hook with reference to the in-line joint.

Fig. 21. Cable shape at different time line of events and important events marked.

Fig. 22. Cable tension over length 38.4 m water depth.

163

164


indicating that the slings were released from the pulley once the in-line joint was laid on the seabed. 4.10. Step 3 – perform still water analysis

The bending moment along the cable is shown in Fig. 24. It is evident that the bending moment requirement along the cable was not exceeded. 4.11. Step 4 – perform dynamic simulations

In this section, the analysis is undertaken ignoring the wave in�uence to determine if the requirements of tension, compression and curvature in the cable were met with some wide margin which will later allow for wave action to be added. OrcaFlex was used to extract the range of the tension values shown in Fig. 22. It can be seen from the results that compression was occurring at the joint. The joint has been designed for these small compression loads. The compression in the OFJ is in part due to the rigging set-up. Fig. 23 illustrates the maximum curvature in the cable. The �gure demonstrates that the curvature requirement in the cable was met. Moreover, the effect of the bend restriction was pronounced between the length (69.5 –71.6 m) and (78.2–79.5 m) as the curvature became reduced.

Upon completing the still water analysis, a range of dynamic environments were introducedin the OrcaFlex model to investigate the deviation in the results to the following environmental parameters:

 Wave height  Wave Period  Wave direction  Current velocity The dynamic simulations are required to determine the maximum seastate condition within which the vessel can deploy the joint and the cable giving regard to the following:

 Hoist wire tension.

Fig. 23. Cable curvature over length in 38.4 m water depth.

Fig. 24. Cable bending moment over length 38.4 m water depth.


 Maximum tension in the cable  Maximum bending moment in the cable.  Maximum curvature in the cable.  Maximum declination in the hoist wire The results of the maximum tension in the cable, maximum bending moment in the cable and maximum curvature in the

165

cable will be presented only in this section. The other results will not be presented due to the paper length. Fig. 25 presents the tension in the cable obtained from the range graph. In this � gure, the maximum and minimum tensions are given as a function of environment. It can be seen from the � gure that the signi�cant wave height of 1 m at the upper bound wave period shows tensions are beyond the acceptable limit of the cable.

Fig. 25. Cable tension value versus signi �cant wave height, peak period and steady current for the worst wave heading and current.

Fig. 26. Cable tension value versus wave direction and current for wave height ¼ 1 m and time 5.47 s.

Fig. 27. Cable curvature value versus signi �cant wave height, peak period and steady current for the worst wave heading and current.

166


Fig. 28. Cable curvature value versus wave direction and current for wave height

be determined from the following equation:

Table 9

Boundary condition and loading condition of analysis cases. Analysis case Boundary condition I

II

¼ 1 m and time 5.47 s.

Loads applied

Fixed at Upper Hole of 2 Main Gravitational Force of OFJ: Flange 5.5 t Bending Moment at Both Armor Pot: 125 kN m Fixed at Upper Hole of 2 In ter- Gravitational Force of OFJ: mediate Flange 5.5 t Bending Moment at Both Armor Pot: 125 kN m

From Fig. 26 the limiting weather criteria, for joint deployment, can be determined. Figs. 27 and 28 show the curvature of the cable as a function of environment. It can be seen from Fig. 27 that the curvature of the subsea cable is above the allowable curvature limit for signi �cant wave height of 1 m and upper bound wave period (5.47 s). In these �gures the red line refers to the cable allowable curvature and the green line refers to the bend restrictor allowable curvature. The curvature of the cable was also checked at the signi �cant wave height of 1 m and upper bound wave period (5.47 s) for different wave directions as indicated on Fig. 28. Fig. 28 shows that the curvature is beyond the allowable limit for the 60 and 90 degrees wave directions. As indicated earlier the extensive OrcaFlex simulations were undertaken to determine the limiting weather criteria during the actual offshore jointing as well as to employ the likely maximum expected loads in the on-land simulation. This is in part to verify the mechanical properties of OFJ under deployment conditions. 4.12. Step 5 – stress analysis for the joint

As the OFJ is subjected to a bending moment during the deployment of the in-line OFJ, this section presents stress analysis undertaken using the multi-purpose �nite element package ABAQUS (Manual, 2012). It can be seen from Fig. 22 that the maximum tension obtained from the still water analysis at the water depth of 38.4 m was 70 kN. The maximum dynamic tension from OrcaFlex dynamic simulation was 100 kN. This tension value is associated with the maximum seastate condition within which the vessel can deploy the joint and the cable. However, for conservatism this dynamic tension associated with the maximum seastate will not be used to calculate the dynamic ampli�cation factor. Instead a tension value of 112 kN will be used. This value represents the vessel ’s tensioner capacity. Based on that the dynamic ampli �cation factor (DAF) can

DAF ¼

Maximum TensionDynamic ¼ 112 ¼ 1:6 Maximum Tension for still water condition 70

It should be highlighted that this DAF will be used later in the hand calculations as part of the test procedure & requirements. Table 9 presents the two cases undertaken where the loads are applied to the joint as well as the boundary conditions employed in the two cases. The bending moments of the joint are applied using reference points at the right and the left ends that kinematically couples the relevant degrees of freedom of the inner joint faces. Reference points are constrained in the remaining DOF in which they are not kinematically coupled with the inner joint faces. Single node restraints are applied as per Table 9 to obviate rigid body motions. Linear elastic analyzes were considered to determine the stresses to be checked against the allowable stress criteria. The joint was modeled using S4R having 4 nodes (quadrilateral), with all 6 active degrees of freedom per node. S4R allows transverse shear deformation, where the transverse shear becomes very small as the shell thickness decreases (Manual, 2012). Fine �nite element meshes were used in regions of the joint where stress concentrations may occur. Figs. 29 and 30 show schematics for the two analysis cases undertaken using ABAQUS. As can be seen, the bending moment of 125 kN m was applied to the two ends of the OFJ. This bending moment value represents the allowable bending moment for the OFJ. Beside the bending moment applied to the two ends of the OFJ, a gravity load of 5.5 t was applied to the OFJ. This bending moment is greater than the dynamic bending moment obtained from OrcaFlex simulations. The gravity load was applied as a distributed load along the OFJ length. During the deployment, the crane will lift the OFJ using two main �anges as indicated in Fig. 29. This way of lifting was intended to relieve the bending moment in the main cylindrical body of the OFJ. For conservatism, the OFJ with different boundary condition as shown in Fig. 30 was also investigated. This is in order to capture the likely maximum bending moment in the OFJ structure. Fig. 31 shows the stress contour extracted from ABAQUS for case I. The maximum stress of 213.5 MPa occurs at the armor pot due to the load concentration. However, it still remains in the elastic state. The stress at the cylindrical body shows a low level of stress. The utilization factor is given as 0.80 ( ¼ 213.5 MPa/265 MPa). Fig. 32 shows the stress contour for case II. The maximum stress (219.2 MPa) occurs at the cylindrical body near the lifting point (intermediate � ange). Again, yielding of the cylindrical body did not occur. Therefore we can ascertain that the allowable bending moment of the OFJ structure is at least 125 kNm which is harsher than the loading condition of the inline deployment. The utilization factor is 0.83 ( ¼ 219.2 MPa/265 MPa).


167

Fig. 29. The boundary and loading condition of analysis case I.

Fig. 30. The boundary and loading condition of analysis case II.

Fig. 31. Stress contour of OFJ (Fixed at 2 Main Flanges, Bending Moment: 125 kNm).

5. Step 6 – conduct the mechanical test followed by visual inspection and RWP test 5.1. Testing concept for simulated in-line deployment of OFJ

It was indicated from the OrcaFlex simulations that during the deployment operations, there are three critical stages which have high tension or bending moment as i llustrated in Fig. 33. The loads given in this section are extracted from Orca�ex simulations undertaken as part of the installation engineering. Before conducting the test, it was decided to compare the results obtained from Orca�ex against the analytical calculations of residual tension. The detailed hand calculations of the mechanical load during the in-line deployment of the OFJ are presented below. The calculations are based on Electra 171 (Electra 171). Cable and offshore � eld joint (OFJ) parameters Cable weight in air, W Air ¼ 70 kg/m Cable weight in water, W water ¼ 41 kg/m

OFJ weight in air, W OFJ, Air¼ 5500 kg (from bend restrictor to bend restrictor including the cable) OFJ weight in water, W OFJ, water ¼ 3517 kg (from bend restrictor to bend restrictor including the cable) Length of OFJ, LOFJ ¼ 10.55 m (from bend restrictor to bend restrictor) Inline deployment

For these calculations, refer to Fig. 18. Water depth, D ¼ 38.4 m (Max.) Vertical distance from water level to center of the joint body, d ¼ 9.0 m Vertical distance from seabed to center of the joint body, þ D d ¼ 38.4 þ 9.0 ¼ 47.4 m Unsupported catenary length, L c ¼ 74.8 m Departure angle, ф ¼ 36° Catenary factor, f c ¼ Lc/(D þ d) ¼ 74.8/47.4 ¼ 1.58 Residual tension factor, f R ¼ 0.2 (Based on Electra 171) During the in-line deployment, there are 3 representative stages which have high tension or

168


Fig. 32. Stress contour of OFJ (Fixed at 2 Intermediate Flanges, Bending Moment: 125 kNm).

bending of OFJ as shown in Fig. 33.

Table 10

Stage I: above the stern chute

Load applied during the on-land simulations.

Catenary length corresponding to air section, Lc,Air ¼ f c  d ¼ 1.58  9.0 m ¼ 14.22 m Catenary weight corresponding to air section, W c,Air ¼ W Air  LC,Air ¼ 70 kg/m  14.22 m ¼ 995.4 kg Catenary length corresponding to water section, Lc,water ¼ LC  Lc, ¼ Air 74.8 m  14.22 m ¼ 60.58 m Catenary weight corresponding to water section, W C,water¼ W water  Lc,water¼ 41 kg/m  60.58 m ¼ 2484.0 kg Total catenary weight, W C ¼ W C,Air þ W c,water ¼ 995.4 kgþ 2484.0 kg ¼ 3479.4kg Residual tension, T res ¼ f Rx( W Air  d þ W water  D) ¼ 0.2  (70 kg/ m  9 m þ 41 kg/m  38.4 m) ¼ 440.9 kg Dynamic factor, f D ¼ 1.6 Tension at left hand side of OFJ, T L ¼ f D  (WC þ T res) ¼ 1.6x (3479.4kgþ 440.9 kg) ¼ 6272.5 kg Tension at right hand side of OFJ, T R ¼ T L  sin ф ¼ 6272.5kg  sin 36° ¼ 3686.9kg Angle at left hand side of OFJ, θ L ¼ þ 45° (from Fig. 33) Angle at right hand side of OFJ, θ R ¼  45° (from Fig. 33)

It was noticed that the hand calculations yield conservative results compared to those from OrcaFlex. During the simulation, the tensions from the hand calculations are used to perform the testing. The three critical stages can be summarized as follows:

Catenary length corresponding to water section, Lc,water W water,OFJ ¼ 60.58 m Catenary weight corresponding to water section, W c,water¼ W water  (Lc,water  LOFJ ) ¼ 41 kg/m x 50.03 m ¼ 2051.2 kg Residual tension, T res ¼ f Rx( W Air  d þ Wwater×D) ¼ 0.2x(70 kg/ mx9 m þ 41 kg/mx38.4 m) ¼ 440.9 kg Tension at left hand side of OFJ, T L ¼ f Dx(W C,water þ T res) ¼ 1.6x (2051.2 kg þ 440.9 kg) ¼ 3987.4 kg Tension at right hand side of OFJ, T R ¼ f Dx(W C,waterþ W OFJ,water þ T res) ¼ 1.6  (2051.2 kg þ 3517 kg þ 440.9 kg) ¼ 9614.6 kg Angle at left hand side of OFJ, θ L ¼ 0° (from Fig. 33) Angle at right hand side of OFJ, θ R ¼ 0 ° (from Fig. 33)

Table 10 presents the values obtained from hand calculations. These values shall be employed during the on-land simulations.

STAGE II: below the seawater level

STAGE III: touchdown

Tension at left hand side of OFJ, T L ¼ F dxT res ¼ 1.6  440.9 kg ¼ 705.4 kg Tension at right hand side of OFJ, T R ¼ f D x (W OFJ,waterþ T res) ¼ 1.6x(3517 kg þ 440.9 kg) ¼ 6332.64 kg Angle at left hand side of OFJ, θ L ¼  45° (from Fig. 33) Angle at right hand side of OFJ, θ R ¼ þ 45° (from Fig. 33)

Sta ge

1 2 3

Tensi on (Tonnes)

Bending angle ( Degrees)

Left Side

Right Side

Left Side

Right Side

6.4 3.9 0.7

3.8 9 6.3

þ 45

 45

0  45

0

þ 45

1. Maximum bending: in this stage, the maximum bending takes place roughly at 45 degrees and with maximum tensile stress of 6.4 t exerted on the joint. This is when the joint is lifted at one side towards the end of the vessel ’s deck as shown in Fig. 33. 2. Maximum tensile force: The bending is negligible but the tensile force encountered by the joint is maximized at 9.6t. 3. Opposite maximum bending: When the joint reaches the seabed, the cable will be bent in the opposite direction with a bending angle of about  45 degrees but without tensile force. The upper side of the in-line joint also experiences a bending in the opposite direction with a tensile force of 6.3 t.

5.2. In-line test items & acceptance criteria

Table 11 highlights the tests undertaken during the simulation and after the completion of the on-land deployment simulation. The same table lists the acceptance criteria associated with each test. 5.3. Testing installation set-up

This section describes the step-up adopted during the simulation to mimic the situation during the offshore installation. Noting that the materials used in the simulation such as cable, housing,


169

Fig. 33. In-line deployment stages.

Table 11

Summary of in-line test items & associated acceptance criteria. Speci�c In-Line Test Item

Acceptance Criteria

1. Fiber attenuation test

Increase of attenuation per loop using power meter. During the test: maximum 0.1 dB After the test: maximum 0.05 dB Change of torsion angle of the cable at bend restrictor is less than 5 degrees

Fiber attenuation change during testing 2. Torsion Test

Torsion of cable at end of bend restrictor 3. Radial Water Penetration Test

No water was to be found in the joint

Radial water penetration test on one pre-molded joint including plumbing areas, 24 hours water pressure test No visible crack. Visual check of plumbing area between cable sheath and copper housing ( three joints ) No Hole in plumbing area. No visible gap between plumbing and lead sheath and copper housing. Information for further analysis & usage. The accuracy if the measurements 5. Measurement of internal displacement of cable Check the measurements in axial and angular in the 3 dimensions was less than 5 mm. 6. Visual Checks No visible crack or deformation Visual check of armor pot and bend restrictor 7. Dimensional/material check on pre-molded and OFJ used This was done in accordance with the applicable manufacturing plans 4. Visual Check of plumbing area

tools for installation, were identical to those deployed for the installation on the vessel offshore. 1. The OFJ was installed in a straight arrangement. 2. During installation and assembly of the test arrangement. The inner cores of the joint were marked to determine any signs of axial movement, torsion or other displacement of the cable cores that may occur during the simulation. The position measurements were in the three dimensions. 3. A visible straight marking was applied on the outside of the cable, outer yarn at the armoring pot, outer yarn at the end of bend restrictor, bend restrictor as well as OFJ housing. Similarly, this line was used to determine any signs of axial movement, torsion or other displacement of the cable cores that may occur during the simulation. During the simulations, all the torsional variations in the entire testing assembly were monitored and listed. 4. The total length of the arrangement was 34 m. This is in part to perform the straight tensile test to mimic the in-line laying. Both armoring at both ends of the cable were terminated by pulling heads and a 1.5 m long �ber optics cable. This will be fed out of the pulling head to monitor the �ber optics readings and ensure that the changes in the light power are within acceptable limits. 5. Cable ends were �xed during the simulations. Thecable ends didnot rotate to ensure that any torsion in the cable due to bending or other remained in the test sample and did not leave the cable at the ends.

5.4. Stage-1 tensile bending test

In this test, the following steps were performed to mimic stage1 of the deployment procedure. 1. Before embarking on the tests, the initial attenuation of the �ber optics was measured and recorded. 2. The torsion angle of the cable in the front of the bending restrictor was measured as shown in Figs. 34 and 35. Moreover, the torsional variation along the assembly length was measured and recorded. 3. The pulling force was increased slowly up the values indicated in Table 10. The force was held for 15 min. 4. During the test, optical light power was continuously monitored to check for any cable damage. 5. Upon completion of this test, the torsion angle in front of the bending restrictor as well as the torsional variation were measured. 5.5. Stage-2 tensile test

This test was undertaken to simulate stage-2 of the deployment procedure as shown in Fig. 36. The steps followed during the test are the same as the steps adopted in stage-1.

170


Fig. 34. Tensile bending test (45 degrees) for in-line joint simulation.

5.6. Stage-3 opposite bending test

This is the stage-3 test where the opposite bending test was exerted on the OFJ as illustrated in Figs. 37 and 38. The following steps were implemented during the simulation. 1. Upon the completion of the tensile test, the OFJ housing was turned to a  45 degrees direction and fastened to the ground to keep its position during the tensile bending test. 2. Before embarking on the test, the initial attenuation of the � ber optics was measured and recorded. 3. The torsion angle of the cable in the front of the bending restrictor was measured as shown in Fig. 37. Moreover, the torsional variation along the assembly length was measured and recorded. 4. The pulling force was increased slowly up the values indicated in Table 10. The force was held for 15 min. 5. During the test, optical light power was continuously monitored to check for signs of any cable damage. 6. Upon completion of the test, the torsion angle in front of the bending restrictor as well as the torsional variation were measured. 5.7. Visual inspection check

After the completion of the three stages of the on-land deployment simulation, the OFJ was dismantled and inspected. The pre-molded joint was released from the compound �lling without introducing any additional stresses. This was followed by the following tests which were highlighted in Section 5.2. Test item # 5: the axial and angular displacements of the cable cores were measured. Test items # 4 & 6: the plumbing area between the cable lead sheath and copper housing were examined visually for any sign of cracks or deformation. Test item # 3: for one pre-molded joint a radial water penetration test was undertaken in order to check the tightness of the pre-molded joint after the installation simulation test. The test was performed in accordance with CIGRÉ TB490 ( CIGRÉ TB490), chapter 8.7.4 with the exception of the heat cycle test for 24 hours under water. The end of the cable was sealed by plumbed metal covers. The joint was then placed under pressurized seawater in a pressure vessel. After 24 hours, the joint was released from the water and checked for water ingress and damage. Item test # 7: a complete dimensional check was carried out on the pre-molded joint and OFJ to ensure the tested object was made fully according to the manufacturing plans. The tested objects used in this simulation included offshore �eld joints consisting of three pre-molded joints and one �ber

Fig. 35. Testing set-up, þ 45 degrees pulling test.

optics joint and were found to pass the visual inspection and the test plan which comprised of the serious of mechanical tests, � ber attenuation measurements and a water leakage test.

6. Conclusion

Currently, there are huge demands on the installation of subsea cables around the globe. Often the installation takes place between two distant locations where long cable lengths are required for which subsea installation joints are unavoidable. The installation of long cables requires repair joints. Based on that, it is essential that the repair joint and installation joint will be designed to withstand all the expected different loads during the design life the same as for the cable. This is to increase the reliability and availability of the subsea system and reduce the high cost involved in the repair in the future.


171

Fig. 36. Tensile test for in-line joint simulation.

Fig. 37. Tensile bending test (  45 degrees) for in-line joint simulation.

often dif �cult before the mobilization of the installation vessels or engagement of the installation contractor. This paper presents a set of standardised onshore testing regimes which improves both the reliability and affordability of these jointing's which would be advantageous. The onshore testing regime is suitable only for inline joints. The paper presents the design process which involves a series of steps. These design steps can be followed to determine the loads which can be applied during the onshore testing to mimic the installation conditions for in-line rigid joints. Also, an analytical method is presented to calculate the loads experienced by the in-line joint during the deployment. The new testing arrangements employs the design loads, determined from the engineering simulations, to test the offshore rigid joint in air. Once the load tests are completed, the joint shall be subjected to water penetration test. In other words, the new testing arrangement can be simpli�ed to load testing the joint in dry air and subsequently performing the radial water penetration test. This is rather than conducting a load test of the joint while the joint is under external pressure. The new testing arrangement offers an alternative to subsea immersion testing for subsea cable joints and offshore deployment simulations. Acknowledgment

The authors would like to thank Qatar Petroleum for their permission to publish this paper. Fig. 38.

 45 degrees pulling test.

Failure of submarine power cables have been shown to be attributed to cable � eld joints on some cables. This could provide indications that the current acceptance testing of these joints is inadequate. Subsea cable failure could result in signi �cant �nancial losses especially when power is transmitted via cables to offshore production platforms. Design guidance for cable �eld joints suggests sea trails to determine if the proposed �eld jointing technique is acceptable. Sea trails however are often prohibitively expensive, time-consuming, require large preparation and access to a vessel which is

References CIGRÉ (2010). NorNed, – World's Longest Power Cable, CIGRÉ, France, CIGRÉ publication B1_106_2010. CIGRÉ TB490. Recommendations for testing of long AC submarine cables with extruded insulation for system voltage above 30 (36 ) to 500 (550) kV. DNV-RP-J301. Subsea power cables in shallow water renewable energy applications, February 2014. DNV RP-C205. Environmental Conditions and Environmental Loads. Electra 189. Recommendations for tests of power transmission DC cables for a rated voltage up to 800 kV. Electra 171. Recommendations for Mechanical Tests on Sub-marine cables. International Cable Protection Committee (2009). Damage to submarine cables caused by anchors. Loss Prevention Bulletin.

172


Karlsdóttir, Svandís Hlín (2013). Experience in transporting energy through subsea power cables: the case of Iceland. Manual, OrcaFlex, 2014. Version 9.8 a.. Orcina, Ltd, Cumbria . Manual, ABAQUS Benchmarks, 2012. "Version 6.12.". Dassault Systémes Simulia Corp., Providence, RI. NDGL 0001-0. General Guidance for Marine Operations. Update of Service Experience of HV Underground and Submarine Cable Systems, to be Presented by CIGRÉ Working Group B1-10 2009.

Worzyk, T., 2009. Accessories, Submarine power cables: design, installation, repair, environmental aspects. Springer-Verlag Berlin Heidelberg, Berlin . Woo, Jinho, Kim, Dongha, Na, Won-Bae, 2015. Safety analysis of rock berms that protect submarine power cables in the event of an anchor collision. Ocean Eng. 107, 204 –211. Yoon, Han-Sam, Na, Won-Bae, 2013. Safety assessment of submarine power cable protectors by anchor dragging � eld tests. Ocean Eng. 65, 1 –9.

1-s2.0-S0029801815006733-main

Recommend Documents