Você está na página 1de 12

OTC 14256 A Dry Tree FPDSO Unit for Brazilian Waters

L. Poldervaart, J. Pollack, Single Buoy Moorings, Inc.


Copyright 2002, Offshore Technology Conference This paper was prepared for presentation at the 2002 Offshore Technology Conference held in Houston, Texas U.S.A., 69 May 2002. This paper was selected for presentation by the OTC Program Committee following review of information contained in an abstract submitted by the author(s). Contents of the paper, as presented, have not been reviewed by the Offshore Technology Conference and are subject to correction by the author(s). The material, as presented, does not necessarily reflect any position of the Offshore Technology Conference or its officers. Electronic reproduction, distribution, or storage of any part of this paper for commercial purposes without the written consent of the Offshore Technology Conference is prohibited. Permission to reproduce in print is restricted to an abstract of not more than 300 words; illustrations may not be copied. The abstract must contain conspicuous acknowledgment of where and by whom the paper was presented.

Abstract In the last decade Brazilian offshore oil fields have become one of the most interesting areas containing prospects in water depths ranging from 500 - 3000 m. Many of these field developments require a stand-alone production and storage facility with tanker export. An FPSO is the natural solution and should ideally be installed during the drilling phase to obtain early recovery. To further improve the FPSO concept a dry production tree deck has been included in the moon pool of a spread moored FPDSO. This concept presents to the industry a cost effective solution where drilling, production and work over activities can be carried out from one single floater. The paper describes this new concept with its unique feature that the rigid production risers are suspended from, and tensioned by, a Tension Leg Deck (TLD) located above the water level in the middle of the FPDSO moon pool. The BOP stack and the surface completed production trees are located on the TLD and therefore easy accessible. The tension or uplift force of the deck is provided by weights. The weights are located in dedicated hull compartments where their vertical motions are guided by hinged lever arms. These lever arms are located transverse in the longitudinal direction of the FPDSO vessel. The FPDSO-TLD facility has the flexibility to support all required drilling and production equipment for a wide range of field development scenarios. The FPDSO-TLD concept presented in this paper was studied for a Brazilian deepwater field development scenario where an FPSO and a nearby Dry Completion Unit is the base case development scenario. The FPDSO-TLD concept has been model tested. Introduction The current drive to produce oil in deeper and deeper waters has pushed oil companies to review their field development

strategies. Deepwater fields generally differ from their shallow water counterparts in some or all the following aspects: Absence of or limitations in existing infrastructures, Reservoir size, Reservoir horizontal extent and thickness of oil pockets, Oil properties and well maintenance, Faster return on large capital expenditure. In the new and promising deep water regions (West of Africa and Brazil) infrastructures are scarcely available on the seabed to export the oil and gas production. This has paved the way for floating production systems which have flourished in recent years in various forms (FPSOs, TLPs etc..). The size of the proven reserves in these deepwater fields calls for large storage facilities that can only be provided by ship-shaped floating production systems (new-built barge or converted tanker). The new deepwater fields often differ from shallow water fields by their geology. While the latter consisted of horizontally localized oil pockets with a substantially thick oil layer, the deepwater fields come in shallow and thin oil pockets scattered over a large area. This new situation requires wellheads to be placed at large distances from each other. Optimum production rates and oil recovery often requires a comprehensive work-over and well maintenance program. Given the high day rates of drilling/work-over rigs and the limited fleet able to operate in deep water, availability of onboard work-over if not full drilling facilities brings about a definite economic advantage. Large capital expenditures are involved in these deepwater field developments. Ways of reducing the lead-time to first oil must be investigated. An earlier return on investment can be achieved by starting production from a few pre-drilled wells soon after FPSO hook-up and long before the drilling phase is over. The remaining wells are drilled directly from the FP(D)SO if full drilling capability is incorporated. In summary, the typical deepwater field in West of Africa or Brazil would call for a hub, preferably a large FPDSO, and satellite units located above secondary drilling centers within a few miles of the hub. Well maintenance issues would call for completion/work-over facilities to be included. Dry well heads as on SPARs and TLPs facilitate workover and well maintenance activities. Direct and continuous access to the wellhead is possible without ROV assistance. Furthermore, these concepts allow the use of casing strings for

L. POLDERVAART, J. POLLACK

OTC 14256

risers with potential cost savings as compared to flexible riser solutions. The above reasons account for the popularity of dry trees. The subject of this paper is to present a dry tree FPDSO unit that could be used in deepwater field developments offshore Brazil. The dry tree design is based on the Tension Leg Deck (TLD) concept developed at SBM. This concept, unlike the spar and TLP systems, is rather insensitive to payload and does not rely on buoyancy to pretension the risers. Instead, the riser deck is pre-tensioned by weights connected to the deck by means of cables going over sheaves. These weights are either located in a dedicated hull compartment or underwater well below the sea surface to avoid wave action. This tensioning system allows for dynamic vessel motions and draft variations associated with storage. The concept was first designed as a Dry Completion Unit ( DCU) for W. A. waters. Therefore the paper will follow the path from the DCU to the FPDSO-TLD for Brazilian waters. Modeltest results will be presented from modeltest performed at MARIN the Netherlands, during the months of February and March on this concept. The design basis of the DCU is discussed in Section 2. Section 3 covers model test experiments performed at MARINTEK and the verification of the numerical tools used for detailed analysis and engineering. Applicability of the concept to a Brazil environment is discussed in Section 4. DCU design premises In this section, the design premise for the DCU is outlined. As the weight-based tensioning system is new, it is worthwhile to discuss first the physical principle of the TLD TLD physics. The TLD concept illustrated in Figure 1 is a novel means of tensioning hard risers. This concept has the following main components (see Figure 1): a tendon stabilized deck to which production risers are attached. This deck with trees is located above the mean sea level; cables running over a vessel fixed sheave system hold the deck in place; weights hanging from the cables well below the mean sea level keep the required pretension in the risers and tendons. This concept can be adapted to a large variety of surface floaters provided the motion characteristics are suitable. Figure 1 shows the TLD concept in the moon-pool of a barge or tanker. The barge shown in Figure 1 is merely supporting the sheaves. When the barge heaves up by one meter, the deck remains in place while the suspended weights move up by two meters. Likewise, when the barge heaves down by one meter, the suspended weights move down by two meters with the deck remaining in its position. The TLD concept effectively de-couples the wave frequency motions of the floater from those of the deck thus allowing the barge with its TLD to ride a storm with all risers/tethers connected.

In Figure 1, the weights used to pretension the risers are located well below the mean sea level. In Figure 2 an alternative design is illustrated whereby the weights are located inside the hull and are guided in their vertical motion by hinged lever arms. More details about the TLD can be found in references [1] and [2]. Design Basis for DCU. The basis of design for the satellite DCU is as follows: West of Africa (Angola) environment. Small overall horizontal dimensions for flexibility in choice of yard, Small draught for easy quay side outfitting, Limited storage capacity, 12 risers (9 5/8 in OD casing strings and wall thickness 0.395 in), 4 tendons to support the Tension Leg Deck (again 9 5/8 in OD), 4m spacing between the trees on the deck, Large enough moon-pool to insure deck/hull clearance, Environment. The environment is typical of Block 17 in Angola. Two wave spectra have been assumed corresponding respectively to the long south Atlantic swells and to shorter seas: JONSWAP spectrum =2, Hs=4.2m, Tp=15.4s, JONSWAP spectrum =2, Hs=3.2m, Tp=13.0s, A third spectrum has been assumed to explore larger environments: JONSWAP spectrum =2, Hs=5.5m, Tp=15.4s, The water depth was originally set to 1300m. This water depth could not however be modelled at a reasonable scale in a 10m deep wave basin. A smaller depth of 680m was chosen instead to also allow current generation in the model basin. Preliminary design of tensioning system. Deck size and weight: The TLD deck dimensions are dictated by the number of risers and by the 4m riser spacing. Assuming a circular pattern, the radius is about 7.6m. Space must be allotted for the tendon attachments at the four corners of the deck. A square deck of 18m by 18m is adequate. A deck weight of 100 tons has been estimated. Riser/tendon weight: The weight in water of 16 (12 risers and 4 tendons) 9 5/8 in risers with wall thickness 0.395 in is approximately: 5440 kN i.e. about 555 tons. Pretension weights: When the barge heaves the suspended weights also heave by twice as much. These accelerations induce fluctuations in the tension member connecting the weights to the deck. A good measure of these tension fluctuations is the so-called Dynamic Amplification Factor (DAF) defined as the ratio of the tension standard deviation to the mean tension. An approximate formula for this DAF is simply DAF = a Z 1 / 3 / g where a Z 1 / 3 refers to the significant single amplitude total heave acceleration and g to the gravitational acceleration. With this approximate formula,

OTC 14256

A DRY TREE FPDSO UNIT FOR BRAZILIAN WATERS

DAFs are shown to be independent of the weights and depend only upon the first order acceleration RAOs of the floater and the wave environment DCU dimensions. The DCU is a square barge with a square moon-pool in the middle. Its main dimensions are summarized in Table 1. The moon-pool walls are vertical. The moon-pool dimensions are dictated by riser clearance at keel level (barge horizontal offset and roll/pitch motion). A clearance of 3.5m is provided all around the deck in calm water. TLD system layout. The TLD system is illustrated in Figure 3. Eight weights each with a 300-ton mass are used to pretension tendons and risers. Each weight is connected to the deck via a cable running over two sheaves. The position of first/outer sheave is such that weights do not cause interference with the barge hull or risers. The position of the inner sheave allows the connection to the deck located inside the moon-pool. For simplicity the 12 risers and 4 tendons were lumped into 4 equivalent risers (in mass and stiffness). Mooring layout. The DCU is held on station by an eightline mooring system (4 bundles of 2 lines). The line composition consists of a 150m-long bottom chain segment, a 900m long wire segment and a 50m top chain segment. Model tests and numerical calibration Model test. General description. The model test campaign was performed in the ocean basin at MARINTEK in July 1998. The scale used is 1:90 [3]. The scaled model is shown in Figure 3. A comprehensive series of tests was performed including load excursion verifications, decay tests in surge, heave and pitch in calm water and finally irregular wave tests. These tests were performed each time with and without the TLD with a view to assessing its influence on the DCU motion response. Table 2 summarizes the irregular wave tests performed. The first three tests were performed with the design wave height. Test 5230 was performed to extrapolate the behaviour to more severe environments. As the first order motions were expected to be larger in their test, the elevation of the sheave axis above main deck was increased by 2.5m to prevent any contact between the deck and the sheave. Test 5250 was performed to investigate the influence of the pretension weight on DAF. Result highlights. Natural periods. Decay tests with and without TLD system gave the natural periods in surge, heave and pitch as shown in table 3. Note that the surge natural period is decreased by 25% when the TLD system is connected. When the barge moves the tendons and risers incline from the vertical thus providing an additional restoring force proportional to the pretension weights. The decrease in period is therefore primarily due to a stiffness increase. Changes in heave and pitch natural periods are small/marginal owing to the mass and added mass ratios involved.

First order motions. Standard deviations of surge, heave and pitch motions are shown in Table 4 for all irregular wave tests. The following comments are in order: With the TLD system mounted and a 4.2m significant wave height, the most probable maximum single amplitude heave and pitch motions assuming a Rayleigh distribution are respectively 3.9m and 7.1; Increasing the wave height from 4.2m to 5.5m causes a proportionally smaller increase in the heave and pitch responses. This suggests that viscous effects play an important role; Changing the pretension weights from 300 tons (5240) to 200 tons (5250) has only a minor effect on the standard deviations of the surge, heave and pitch motions. This confirms an earlier conclusion that the motion of the barge is decoupled to a large extent from the suspended weights. Tensions in TLD system. Tensions in the TLD cable systems have been measured at 6 locations referred to as T1, T2, TLD1, TLD2, TLD3 and TLD4 as indicated in Figure 4. Tension T1 and T2 are top tensions for the bow and starboard risers. TLD1 to TLD4 are tensions measured along the cables connecting suspended weights to deck. Tensions TLD2 and TLD4 are measured just above the suspended weights. Sheaves were modelled with roller bearing thus implying 1% to 2% friction. A DAF is calculated as the ratio of the standard deviation of TLD2 and TLD4 tensions to the weight in water used for pretension. These DAF values are listed in Table 5. The following remarks can be made: DAFs are higher for TLD4 which is located up-wave (see Figure 4), The largest DAF for Hs=4.2m (all tests except 5230) is 8.2%, Increasing the wave height from 4.2m to 5.5m induces roughly a 25% increase in DAF, The DAF values for Hs=4.2m and test # 5240 and 5250 are unaffected by the decrease of the pretension weights from 300 tons to 200 tons. Most Probable Minimum top riser tensions (T1 and T2) are listed in Table 6. The following can be pointed out: These MPM values must be greater than the riser weight in water. In the model tests the risers were flooded and their weight in water was 1360 kN. The above table shows that the risers remain in tension throughout the water column. It is expected that the minimum tension be larger in the presence of current (test # 5210) and lower with the larger wave height (Test # 5230) or the smaller weights (Test 5250). Numerical calibration against model tests. General. A numerical model of the DCU is built with the AQWA suite of programs. This numerical model is tuned to

L. POLDERVAART, J. POLLACK

OTC 14256

match the experimental results. A similar exercise was performed with a TLD system mounted on a 300,000 dwt tanker [4] also tested at MARINTEK in July 1998. A view of the AQWA model is shown in Figure 5. Result highlights. Figure 6 shows a comparison of the measured and calculated load-excursion curves. A good agreement is found. The details of the calibration of decay tests and irregular wave tests will be reported elsewhere Conclusion. Model tests have been performed that confirmed the validity of the TLD concept for West of Africa. Dynamic Amplification Factors of about 8.5% have been measured. Minimum tensions at mud-line are such that the risers are well tensioned throughout the water depth. Successful calibration of numerical tools against experimental results has already been reported for a TLD system on a 300,000 dwt tanker [4]. Our numerical tools are well suited for the detailed analysis and engineering of the TLD systems. New FPDSOTLD layout/design for Brazil Environment The TLD concept has been experimentally validated for a 65m x 65m square DCU in West African environment. When the wave height was increased to Hs=5.5m, the measured Dynamic Amplification Factors were about 10%. For a typical Brazil environment, this figure would be even higher thus making the small DCU floater unsuitable for this oil region. Large spread moored FPDSO-TLD designs for West of Africa have been designed for relatively small beam waves provided the floater bow is oriented towards SSW [1]. The associated roll motion is small enough to insure riser/hull clearance. Should this floater be spread moored bow to the South or to SSW in Brazil, the roll response in beam waves would surely be prohibitive from both the riser/hull clearance and tension fluctuations points of view. A new hull shape must therefore be developed to be spread moored offshore Brazil. This section will highlight the preliminary stage of the design process of this new hull shape and of the TLD system mounted on it. Design requirement. Brazilian environment, Storage capacity: 1.5MBBLs, Wells: 12 productions/water injection wells (outer casing 13 3/8 x 68 lb/ft, tubing string 7 5/8 x 29.7 lb/ft, power control coil 2 3/8 x 5 lb/ft), Topside capacity: 100,000 Bopd. Design environment. The water depth is assumed to be 1200m. The following 100-year Brazilian design waves are listed in table 7 for the sectors from East to South-West. Tidal variations are as follows: Tidal Astron. Max. +Tidal Atm. Max = 1.5m,

Tidal Astron. Min. +Tidal Atm. Min = -0.4m, The total tidal range is thus 1.9m.

Hull design. The stated requirements include a 1.5 MBBL storage capacity and a 100,000 bopd process capacity. The design philosophy adapted for this hull design is as follows: Rectangular barge shape, Presence of horizontal skirts to increase the roll and heave natural periods, Limitation of the overall beam (including skirts) to around 80m for flexibility in the choice of the yard. Several hull shapes have been considered as shown in Table 8. The models with a 16m draft result from a detailed investigation regarding the ballast and storage capacity required achieving a near zero-draft variation during operation. These exercises lead to the 16m draft, instead of the earlier assumed 22m draft. Figure 7 illustrates the Tex 11 barge mesh. The radii of gyration are respectively 0.35*Bpp and 0.25*Lpp for roll and pitch. Table 9 summarizes the natural periods and radiation damping (expressed as percentages of critical damping) for the heave, roll and pitch dominated coupled modes. The heave natural period is higher with wider skirts. Decreasing the height of the skirts from 5m to 2.5m results in a small reduction of the heave and roll periods. Having skirts at the bow and stern increases the heave, roll and pitch natural periods by 4, 7 and 15% respectively. Again, the wider the skirts are, the larger the roll natural period is. Radiation damping is negligible in roll. First order response. First order heave motion RAOs for beam waves are shown for all configurations in Figure 8. One observes that the implementation of the moon-pool appears to be beneficial in terms of maximum response at the peak. Note that the above motion responses assume no other sources of damping other than radiation. Viscous damping is to be expected due to the sharp edges of the skirt The barge is spread moored with the bow pointing to SSW. The environments of Table 10 have been assumed. The waves are respectively from the port beam, bow quartering and head directions. Table 11 summarizes the largest significant single amplitude heave motion and acceleration response at COG and the environment responsible for these values. These results have been obtained by a frequency domain approach with radiation damping only for heave. Table 12 summarized the largest roll and pitch angles and the environment these are associated with. These results are obtained with a frequency domain approach (LF ignored). A 2% of critical has been added in roll to account for viscous effects at the keel. No additional damping has been taken in pitch. We observe that for the 16m draft hulls (Tex5, Tex7, Tex8 and Tex11), the wave frequency roll motion is quite small while the wave frequency pitch motion is larger. For the above hulls, the roll natural periods are much longer than the wave periods associated with the spectra of Table 10 whilst, in contrast, the pitch natural periods are close to the wave

OTC 14256

A DRY TREE FPDSO UNIT FOR BRAZILIAN WATERS

spectrum peak periods. The largest roll motions occur for the beam waves coming from the ESE and the largest pitch motions for the head waves from the SSW. It is shown that a low frequency (LF) roll motion also exists and must be considered in the design. Heave responses are important at sheave locations. These will govern the riser dynamic tensions. Responses have therefore also been computed for Tex11 at each corner of the moon-pool: bow port (x=20m, y=12m,z=42m), bow starboard (x=20m, y=-12m,z=42m), stern port (x=-20m, y=12m,z=42m) and stern starboard (x=-20m, y=-12m,z=42m). Table 13 summarizes the heave motion and acceleration responses at these locations for hull Tex11. This hull has been chosen for the design, as it satisfies the construction criteria of having a beam less than 80m. The SSE environment is therefore the designing one for heave response. It is found that the heave response at the sheave locations can be larger (up to 10%) than at COG owing to the contributions of pitch and roll. This increase is relatively small due to the short distances to the principal axes and would be higher if the moon-pool was located away from amidships. The maximum heave acceleration is found to be 0.40 m/sec. The associated DAF is therefore 0.40m/sec/g = 4.1%. For the Tex11 barge, a maximum double amplitude heave motion of 8.5m is found at the stern corner of the moon-pool on the starboard side for the bow-quartering SSE swells. Low frequency roll motion. LF roll motions exist due to the hulls large roll natural periods (roll period for Tex11 is 21.5s i.e. above the wave period range) and to the non-trivial second order roll drift moment. This has recently been confirmed during a model test campaign performed in October 2000 in MARIN new offshore basin [4],[5]. Second order roll moments, even if much smaller than first order moments, can generate a large roll response at the natural period depending on the amount of damping. Assuming an approach similar to that used to evaluate the standard deviation of the low frequency surge motion of a moored structure based on surge force spectral density, mooring stiffness and damping, we can estimate the low frequency roll motion standard deviation once the second order drift moment in roll is computed. The roll stiffness is primarily hydrostatic and the roll damping is conservatively assumed to be 2%. With these assumptions, the significant single amplitude of the LF roll motion is found to be 1.1 for the beam waves with Hs=6.7m and 2.8 for the bow quartering waves with Hs=7.0m. Assuming a Rayleigh distribution, the 3-hour maximum LF roll motion is then 5.2. As the phasing between the LF and wave frequency (WF) is not clear, we adopt the method often used for the LF and WF surge motion i.e. we combine the LF significant with the WF maximum and vice versa. The most severe combination yields a 6.3 maximum single amplitude low+wave frequency roll motion. This value will be used in the TLD weight and clearance calculations.

Having most of the roll response at low frequencies means that the roll acceleration will come primarily from the small wave frequency component. This is beneficial in terms of total heave acceleration, as only wave frequency motions will contribute. Figure 9 shows the roll motion and acceleration response for the Tex11 hull at 90. Estimation of riser weight. The requirements call for 12 production/water injection risers. The risers are assumed to have the following characteristics: 13 3/8 OD outer shell, 68 lb/ft dry weight, 7 5/8 OD tubing string, 29.7 lb/ft dry weight, 2 3/8 OD power control, 5 lb/ft dry weight, The annulus space is assumed filled with seawater and the tubing string with 8lb/gal oil. The tendons are 10 3/4 OD with 15.1mm wall thickness and are filled with air. The weights in water and axial stiffness are listed below: Tendon: 44 tons, 2.1 MN/m, Production/water injection riser: 161 tons, 2.2 MN/m, The axial stiffness above is that of the outer shell as the tubing strings and power control are assumed fitted with slip joints at mud-line. TLD system. One rectangular 32m-long and 12m-wide deck will support the 12 risers. This deck will be sheltered from wave action in the 40m long and 24 m wide moon-pool. The total suspended weight (including a 400 ton deck) is 2500 tons. With a conservative 5% DAF, eight suspended weights each of 385 tons are required. The weights keeping the risers under tension are located in dedicated hull compartment and are each guided by a hinged arm (see Figures 10 and 11). The length of the arm is dictated by: Range of barge vertical motions, Maximum sheave fleet angle allowed. The range of vertical motions is broken down according to Table 14. The wave frequency total heave corresponds to the contribution of all three vertical motions at sheave locations and comes form Table 13 (2.3 * 1.86 * 2 = 8.5m). The low frequency roll motion must also be accounted for as it will induce a heave motion at the sheave location. The maximum single amplitude LF roll motion has been estimated to be 5.2deg. A distance of the sheave axis to barge longitudinal axis to of 11m is assumed. This results in the LF motion range of 2m = 2 * 11m * tan(5.2). The total range of heave motion is 17.3m. It is important to point out that this range results from simple superposition of various contributions without accounting for the probability of occurrence of joint events. This approach is therefore quite conservative. A typical arm length able to cope with the above range of vertical motion and admissible fleet angle from the sheaves is 15 m.

L. POLDERVAART, J. POLLACK

OTC 14256

Riser/hull clearance. When the barge rolls, the moon-pool walls near the keel come closer to the risers and tendons. A reliable estimate of the maximum single amplitude roll motion (low and wave frequency) is therefore important for clearance analysis. The value estimated is 6.3. Owing to the barge horizontal excursion (limited to 7% of the water depth), the risers will have a static inclination from the vertical of atan (0.07) = 4.0. In total, the relative roll angle is 6.3+4 = 10.3. Note that this value is conservative since the largest roll motions occur for bow-quartering waves that are unlikely to cause a 7% excursion in the beam direction. Based on the above assumptions, the lateral displacement of a riser/tendon from its initial position in calm water is h*tan(10.3) where h is the height of the sheave axis above keel. Assuming a height h = 42m, the lateral displacement is found to be 7.6m. In calm water, the distance of a production riser from the moon-pool wall is w/2-e = 24/2 -3.2 = 8.8m where w is the moon-pool width and e the riser distance to the deck longitudinal axis. In this configuration, a clearance of 1.2m is provided with the vertical moon-pool wall. Additional clearance can be provided if the moon-pool walls are slightly flared near the keel. A similar calculation made for a tendon shows it could be 4.4m from the longitudinal axis before interference occurs. Should the tendon be further out, the moon-pool would be flared out. Conclusions. A large number of hull shapes with various sizes of horizontal skirts near the keel, with and without moonpools have been investigated for a motion and acceleration responses using a frequency domain analysis. A 320m long, 64m beam, rectangular barge fitted all around with 7.5m-wide skirts has been shown capable of supporting dry trees in a Brazilian environment. A 40m-long, 24m-wide moon-pool located amidships is provided for the Tension Leg Deck supporting 12 steel risers in a 1,200m water depth. Preliminary design calculations call for eight 385-ton weights to keep the risers under tension at all time. These weights are located in dedicated hull compartments and guided in their vertical travel by 15m long arms pivotally fixed to reinforced hull compartment walls. Clearances between risers/tendons and the moon-pool walls near the keel are adequate. General conclusions A small, spread moored, Dry Completion Unit (DCU) supporting dry trees on a TLD has been validated experimentally as being well suited for West African field developments where multiple drilling centers could be required. In a proposed field layout, these DCUs are connected to an FPSO hub. This floater could also feature dry trees and a TLD system to support steel production/water injection risers [1]. As the small DCU is not suitable for a Brazilian environment, a new hull shape has been designed which is able to cope with

the waves a spread-moored floater would be subjected to. A preliminary design of the TLD system shows that there are no feasibility hurdles that cannot be overcome for a spreadmoored FPDSO-TLD offshore Brazil. In this paper, various floater shapes have been investigated for supporting risers using the TLD system. It has been shown how the performance of the TLD system (riser tension fluctuations) is closely linked to the motion and acceleration characteristics of the floater. This illustrates that an integrated hull/TLD design can properly support hard riser systems in the more severe environments offshore Brazil. References
1. An FPSO with dry well heads located on a novel Tension Leg Deck as Stand Alone Deepwater Field Development by L. Poldervaart, J. Pollack. Proceedings of Deep Offshore Technology Conference, 19-21 October 1999. Stavanger, Norway. A Surface Tree Riser Tensioning Systems for FPSOs by J. Pollack, L. Poldervaart and M. Naciri, Proceedings of 2000 Offshore Technology Conference. Paper OTC 11902. Tension Leg Deck (TLD) SBM test campaign. Main report, MARINTEK, November 1998. The Tension Leg Deck, From Drawing Board to Numerical Design Tools by J. Pollack, M. Naciri and L. Poldervaart. Proceeding of the 2000 Offshore Mechanics and Artic Engineering Conference. Paper OMAE-00-4001 - February 1417. 2000, New Orleans - Louisiana. Model tests on a rectangular barge with different skirt configurations MARIN report. December 2000. Non-linear low-frequency roll excitation of a rectangular barge by M. Naciri and N. Lledo. Proceeding of the 2001 Offshore Mechanics and Artic Engineering Conference. Paper OMAE-01-1247 - June 4th 8th, 2001. Rio de Janeiro Brazil

2. 3. 4.

5. 6.

OTC 14256

A DRY TREE FPDSO UNIT FOR BRAZILIAN WATERS

Length [m] Beam [m] Depth [m] Length moon-pool [m] Beam moon-pool [m] Draught [m] Displacement [m3]

65 65 19 25 25 7 25200

Table 1 DCU main particulars


Test # Hs (m) Tp (s) Vc (m/s) 5011 4.2 15.4 5110 4.2 15.4 5210 4.2 15.4 1.1 5230 (*) 5.5 15.4 5240 (*) 4.2 15.4 5250 (*) 4.2 15.4 (*) These tests are performed with sheaves lifted up by 2.5m. Mooring Horizontal Catenary Catenary Catenary Catenary Catenary TLD weight N/A N/A 8 x 300 tons 8 x 300 tons 8 x 300 tons 8 x 200 tons

Table 2 Irregular wave tests performed on DCU


Degree of freedom Surge Heave Pitch Test # 5011 5110 5210 5230 (*) 5240 (*) 5250 (*) Nat. period w/o TLD [s] 240 10.2 9.0 Nat. period with TLD [s] 180 10.5 9.0 pitch [] 1.85 1.70 1.50 2.21 1.89 1.83

Table 3 Natural periods with and without TLD ( 8 x 300 tons)


surge [m] 3.45 2.40 2.66 3.04 2.12 2.05 heave [m] 1.02 0.96 0.90 1.19 1.00 0.98

Table 4 Standard deviation of barge motions


Test # 5210 5230 5240 5250 TLD2 6.9 9.1 7.1 7.1 TLD4 7.1 10.1 8.2 8.1 Test # 5210 5230 5240 5250 T1 (kN) 3550 2848 3317 2120 T2 (kN) 3689 3298 3480 2270

Table 5 DAF values (%)


Sector E SE S SW Hmax [m] 8.7 12.4 13.0 14.6 THmax [s] 11.7 12.1 12.1 12.2 Hs [m] 4.7 6.7 7.0 7.8 Tp [s] 9.2 11.4 14.7 15.4 Tz [s] 6.9 8.5 11.1 11.5 2.19 1.58 1.62 1.70

Table 6 Most Probable Minimum top riser tension

Table 7 100 year wave conditions in Campos Basin


Hull Id Tex1 Tex2 Tex3 Tex4 Tex5 Tex7 Tex8 Tex11 Lpp [m] 320 320 320 320 320 320 320 320 Bpp [m] 60 60 70 70 64 64 64 64 D [m] 32 32 32 32 32 32 32 32 Draft [m] 22 22 22 22 16 16 16 16 Skirts w & h [m] 12x5 12x5 5x5 5x2.5 10x3 10x3 10x3 7.5x5 Skirt Location All around On the side On the side On the side All around All around All around All around Moonpool Dim. [m] N/A N/A N/A N/A N/A 80x24 40x24 40x24

Table 8 Hull shapes considered for diffraction/radiation

8 Hull Id Tex1 Tex2 Tex3 Tex4 Tex5 Tex7 Tex8 Tex11 Skirts width & height [m] 12x5 12x5 5x5 5x2.5 10x3 10x3 10x3 7.5x5 Skirt location All around On the side On the side On the side All around All around All around All around

L. POLDERVAART, J. POLLACK Moonpool Dim. [m] N/A N/A N/A N/A N/A 80x24 40x24 40x24 Heave 17.2s/5.7% 16.6s/5.5% 14.9s/9.8% 14.7s/10.8% 15.3s/8.2% 15.2s/8.3% 15.3s/8.3% 14.8s/8.5% Roll 35.1s/0% 32.8s/0% 19.9s/0.1% 19.4s/0.2% 23.4s/0% 22.3s/0% 23.1s/0% 21.5s/0% Pitch 17.4s/2.3% 15.2s/3.1% 13.5s/6.8% 13.5s/7.6% 15.2s/4.8% 15.0s/4.7% 15.3s/4.8% 14.6s/5.1%

OTC 14256

Table 9 Periods of natural modes and radiation damping


Sector ESE SSE SSW Hs [m] 6.7 7.0 7.8 Tp [s] 11.4 14.7 15.4 Tz [s] 8.5 11.1 11.5 1.6 1.6 1.7 Dir of propagation [] -90 (beam) -135 (quartering) -180 (head)

Table 10 100 year wave conditions for response calculations


Hull Id Tex1 Tex2 Tex3 Tex4 Tex5 Tex7 Tex8 Tex11 Skirts width height [m] 12x5 12x5 5x5 5x2.5 10x3 10x3 10x3 7.5x5 & Skirt Location All around On the side On the side On the side All around All around All around All around Motion (m) 2.1 2.0 1.9 1.9 2.1 2.1 2.1 2.1 Env. SSE SSE ESE ESE SSE SSE SSE SSE Acceleration (m/s2) 0.28 0.27 0.39 0.41 0.33 0.33 0.33 0.37 Env SSE SSE ESE ESE SSE SSE SSE ESE

Table 11 Wave frequency significant single-amplitude COG heave motion and acceleration
Hull Id Tex1 Tex2 Tex3 Tex4 Tex5 Tex7 Tex8 Tex11 Skirts dim. [m] 12x5 12x5 5x5 5x2.5 10x3 10x3 10x3 7.5x5 Skirt Location All around On the side On the side On the side All around All around All around All around Roll [] 1.3 1.2 1.1 1.2 1.3 1.3 1.2 1.2 Env. ESE ESE ESE SSE ESE ESE ESE ESE Pitch [] 2.6 2.2 1.8 1.8 2.3 2.2 2.3 2.3 Env. SSW SSE SSE SSE SSW SSW SSW SSW

Table 12 Significant single-amplitude wave frequency roll and pitch motions


Location CoG Bow/P Bow/SB Stern/P Stern/SB Motion (m) 2.1 2.2 2.2 2.3 2.3 Env. SSE SSE SSE SSE SSE Acceleration (m/s2) 0.37 0.37 0.40 0.38 0.40 Env SSE SSE SSE SSE SSE

Table 13 Significant single-amplitude total heave motion and acceleration at various locations for hull Tex11
Origin Draft variations (1) [m] Set down (2) [m] Tide variations [m] Wave frequency total heave at sheave location [m] LF roll induced heave at sheave location [m] Total [m] (1) To allow for some draft variations. (2) Assuming a 7% excursion in the damage case. Associated range 2.0 2.9 1.9 8.5 2.0 17.3

Table 14- Range of heave motions

OTC 14256

A DRY TREE FPDSO UNIT FOR BRAZILIAN WATERS Figure 1 Schematic of TLD system with weights in water

Figure 2 Schematic of TLD system with weights in a dedicated hull compartment.

10

L. POLDERVAART, J. POLLACK Figure 3 DCU in MARINTEK ocean basin

OTC 14256

Figure 4 Layout of cable and sheave system and location of load cells

Wave TLD TLD 18m


T1 T

25m

65m

TLD TLD

OTC 14256

A DRY TREE FPDSO UNIT FOR BRAZILIAN WATERS Figure 5a AQWA model of DCU top view Figure 5b AQWA model of DCU isometric view

11

Figure 6 Verification of load-excursion curve

Figure 7 Mesh of Tex11 hull (1/4 shown only)

12

L. POLDERVAART, J. POLLACK Figure 8 Heave motion RAOs in beam waves

OTC 14256

Figure 9 Roll motion and acceleration RAOs at 90 for hull Tex11

Figure 10 Isometric view of tensioning system

Figure 11 Isometric view of tensioning system and support superstructure

Você também pode gostar