Proceedings of the Fifth (1995) International Offshore and Polar Engineering Conference

The Hague, The Netherlands, June 11-16,1995

Copyright © 1995 by The International Society of Offshore and Polar Engineers
ISBN J-lJ80653-16-8 (Set); ISBN 1-880653-17-6 (Vol I)

TLP Design Philosophy - Past, Present, Future

B.i. Natvig
Norwegian Contractors, Stabekk, Norway

H. Vogel
Aker Engineering, Oslo, Norway

Abstract Introduction
The paper starts with capturing the situation of inventive Development of offshore oil and gas fields were pioneered by
engineering caused by the transfer of mature offshore Americans in the Gulf of Mexico. This was considered 10 be
technology from the Gulf of Mexico to the North Sea during a mature technology at the end of the 60's, when
the early 70's. developments started in the North Sea. Unfortunately, the
transfer of offshore technology from the Gulf of Mexico to
This period coincided with the early stages of TLP the North Sea showed that this technology needed substantial
development and it is suggested that the TLP is a product of modifications in order to combat the rougher environment of
this rather unique situation. Some of the reasoning of tether the North Sea.
design of this early period is outlined.
With the long European tradition of sea transport in this rough
The 5 TLPs now having been build are being discussed environment Naval Architects and Civil Engineers of
especially in the context of the early reasoning on TLP design countries such as UK, France, Germany, Holland, Denmark
and with the emphasis on the tethers. It is indicated how a and Norway were soon attracted to solving the problems of
kind of design consensus has been evolved. North Sea platforms. Also, they were making a large number
of proposals for more economic structures which could better
Regarding future TLPs it is suggested that the present design withstand the North Sea environments.
consensus should be critically scrutinised to remove any
extravaganza which are not strictly required for safety or The Europeans, who had little or no past experience and thus
operational reasons. A few areas which could be scrutinised no inhibitions, went into the steep part of the learning curve.
are discussed. It is a characteristic of people in such situations that they can
be very creative. In fact, during the early 70's a large number
The paper is somewhat subjective in trying to explain why proposals for offshore structures were put forward. Many
certain critical decisions have been made. In some cases there proposals were quite unrealistic, but many ideas were brought
may be other reasons which the authors will not be aware of. into new structures:
The purpose of the paper, however, is to inspire fresh thinking • The Pentagone (NOU) semisubmersible with column
into developing future TLPs by trying to go through the footings, rather than pontoons, and with long
arguments of the past and the present. unsupported trusses.
• The first platform was made in concrete. This was the well
The views expressed in this paper are personal and only those known Ekofisk tank with the perforated break water. This
of the authors. was fo]]owed by a number of concrete platforms in UK
and Norway.
KEY WORDS: TLP, Design Philosophy.

64
At tlds time, Alter pioneered the following structures: relates to the manufacture, installation, operation and
•. Twin pontoon semisubmersibles with unsupported long installation. Also, seabed completions limit well interventions
trusses. Around 50 such units have been built altogether. which can reduce the ultimate oil and gas recovery. The
• Concrete platforms of the Condeep type of which 14 lifetime cost of such systems will therefore need to be
have been built to date. considered together with thier initial cost.
• The fully outfitted deck where the processing equipment
where integrated with the structural arrangement. These
decks are mated inshore onto the tower structure using Design Philosophy During the Early Days
barges. Most Condeeeps are built with some form of
integrated deck (Sandna, 1973). of TLP Developments
The rising interest in Offshore Engineering of the Europeans The early 1LP proposals featured 3 column triangular, 4
must have fuelled a fresh interest in developing novel column square, 4 main columns with 2 auxiliary column
structures in the US. rectangular configurations. 5 column and hexagonal
configurations were also proposed. Most proposals had
Developments of TI..Ps started in such an inventive period supporting vertical and horizontal trusses and circular
both in US and Europe. Whilst the idea is some 100 years old pontoons. Only a few had box shaped ring pontoons. The
it was fust picked up by the industry in the earJier part of the large variety of the proposals is typical of the early stages of
70's. Two prototype large scale tests, the Triton (MacDonald an emerging technology.
1974) and DOT (Anon, 1975), were carried out in sheltered
sea conditions. Aker started their 1LP work in 1974 together A very attractive feature of TI..Ps was the ability to shift any
with Saga Petroleum. resonance outside the frequency region of active wave energy.
Hardly anything was known about non-linear wave effects.
Despite these two prototypes, the way ahead should be long It was believed that any resonances could be avoided by
and difficult. One, of many reasons for this, was that there providing sufficient tether stiffness such that the vertical, and
existed proven technology for bottom fixed structures with two rotation modes about horizontal axis degree of freedom
acceptable field development economy. Poor experience in (heave, roll and pitch) would be below, say 2-3sec and thus
the past had told the oil companies to be sceptical of new below the regime of any appreciable wave action. For the case
technology even if the economic aspects were promising. It of vertical parallel tethers the resonant periods for the two
seems that the more proven the existing technology becomes horizontal and the rotation about vertical axis (surge, sway
the more convincing the economic arguments for a novel and yaw) could easily be placed way above the 20-30sec
structure need to be. periods which would be above any appreciable wave energy.
The later appreciation of 2nd order wave non-linearities and
Another aspect of promoting a new design is that it seems to wind gust dynamics has made this feature less attractive.
be expected that the safety, such as its ability to withstand a Despite this, the ability to avoid resonances in the linear wave
progressive failure, should be much better than for existing frequency range is still an attractive property of TI..Ps.
designs. One might argue that some of the existing platform
designs could not have been promoted from scratch today. The major new and unproven components o{ TI..Ps were the
Jack-Up rigs are, perhaps, one such example. tethers and risers. Whilst tethers had never been used before
in the Offshore Industry, pretensioned laterally unsupported
With the natural resistan~ against the use of unproven risers had been extensively used from semisubmersible
tec~ology the 1LP is one of only a few methods of drilling platforms. There were, however, no experience from
developing deep water fields. Fixed bottom platform long term use of risers for production under extreme sea
technology has shown an impressive ability to adopt to conditions. Extensive work was therefore carried out on
continually greater water depths, ego Cognac and Bullwinkle tether design and production riser design.
steel jacket platforms in the Gulf of Mexico and Troll
Condeep Tower at 303m in the North Sea. However, there Early Work on Tethers
will clearly be a practical limitation to their use in deep water. There was a lot of experience in tension cables subjected to
The deep water capabilities of the 1LP is, perhaps, the main static and dynamic loading when used for pretensioned
reason why this concept broke the natural resistance barrier. concrete structures. Cables used for suspension bridges was a
similar application of tensioned cable technology which was
The main attraction of using a 1LP for the exploitation of found to be of relevance to 1LP design. Many of the early
deep water fields is its ability to complete the wells on the 1LP design proposals thus used parallel strand bridge cables
deck in much the same way as for bottom fIXed platforms. packed within a corrosion collar.
This means that at seabed level only simple emergency shut
down wells are requirded as for fIXed platforms. This is The advantages of the parallel strand cables were noted to be:
possible because of the low heave, roll and pitch motion of • Excellent fatigue properties since no welding
TI..Ps. Other floating platforms will have higher heave, roll • High breaking strength since the strength properties of,
and pitch motions and costly sea bed completion systems say 12mm, cable strands is superior to thicker steel
must thus be used. The heavy cost of these sea bed systems
6S
• A possible failure of one strand would not progress to the this is the ability to replace and possibly remove tethers for
other strands in a way such that' a fatigue crack could inspection. The documentation of tension behaviour under
progress through the whole section of a tubular or solid the condition of one removed tether is a consequence of this
tether. safety consideration. It should be noted that the tethers will be
designed to satisfy the same margin of safety against failure as
The problem areas of stranded cables were: any critical structural element of any permanent offshore
• TLP tethers required larger steel cross sedions, ie more strudure. To build into the strudure the possibility to replace
strands within a collar, than used in bridges a tether thus reflects an improved safety of the 11P compared
• Long term properties of the corrosion protection under to most other structures. Since a tether replacement strategy
heavy hydrostatic pressure was questioned has been adopted some people might interpret this as sign of
• Installing the long flexible cables in one unit was found lack of confidence in tether design which thus warrants these
to be a problem due to high weight in water replacement measures. This is typical of the difficulty in
• Inspedion of individual strands promoting new concepts. A clearly advantageous feature may
• Whilst end termination solutions for bridge cables were sometimes be viewed as a weakness of the proposal.
welI proven, it was argued that a TLP application would
experience a higher level of axial and bending dynamics Much attention was given to tethers going slack. It was
for which there was little experience considered that the snap loads that would fo]]ow a
retensioning of a slack tether would be such that it would fail
Rather than adapting the bridge cable technology to TLP in much the same way as being overloaded in dired tension.
application work on this solution was discontinued after some The procedure for determining the level of tether pretension
efforts. Even today, little work is being done in this field. was thus to make sure that the safety level against overloading
was the same as that against zero tension. Since a utilization
In the land based and offshore oil industry there is factor against zero is meaningless one would often assign the
considerable experience with driII string technology. The same tension variation margin against zero tension as needed
drill string is assembled by screwing together steel segments. to satisfy the overloading criteria. One should note that an
The drill string is able to withstand the torsion due to drilling unnecessarily high pretension means less abiJity to carry deck
and high tension forces during lifting of the drill string. Ie. payload. The pretension level should therefore be selected
near relevant proven technology existed for connecting and with great care. Also, one should note that a retensioning
assembly of long metal strings. Also, there were in-service tether would not necessarily produce high loads and thus not
experience with this solution under similar static and dynamic lead to faileure. Ie. design against overloading and slack
loadings. For this reason the major development work waS should be viewed as different failure modes. '
soon shifted to drill string type tethers during the early 70's.
Early Work on Risers
The tether arrangements were subjected to much discussion Heavy efforts were put into advancing riser technology in this
during the early 70's. The Triton and also some of the other period:
proposals featured non-vertical tethers such that the • Major software projects for analysing riser dynamics
horizontal motion modes would be elastically restrained in were started at a number of institutions and engineering
addition to the vertical modes. This was found to be companies
unattractive for the folIowing reasons: • A large joint industry projed on model testing of a
• There are no major operational or design disadvantages number of proposals for grouping of production risers
associated with horizontal excursions with export risers to obtain applicable drag coefficients
• The compliancy of the horizontal motion modes means was undertaken
that the risers experience less relative velocity and thus • Full scale testing by recording the behaviour of an
lower drag forces than for a horizontal}y restrained tether instrumented segment of a drilling riser co]]ar was done
arrangement
• Wave loads and thus structural dimensions would In addition, there were heavy efforts on developments of the
increase for a horizontalIy restrained platform associated hardware needed for production through laterally
• Tether dimensions as well as foundation design would be unsupported pretensioned risers.
severely affected
The TLP would be meaningless in the absence of confident
For those reasons, vertical parallel tethers were selected. This production riser technology. This paper wiJ], however, limit
implies that the horizontal motion as well as the rotation about itself to a discussion of the adual structure with emphasis on
the vertical axis are compliant modes. The vertical motion the tethers. '
and the rotation about the two horizontal axis are thus the
restrained or stiff modes.
Present Day Design Philosophy
In a situation where a proposal for a new field development
concept is being promoted it is inevitable that one is Conoco pioneered TLPs by building the first production TLP.
demonstrating the robustness by including safety features The Hutton field is only 150m depth and thus stridly too
which other structures can not comply with. One example of
66
shallow for an economic TLP, but one motivation was to try the importance of slowly varying wave drift and wind gust
out this new method of field development. Having made the dynamics on the horizontal offset were well advanced.
first TLP it was easier for the followers. A table of these are
shown below. Coinciding with the Snorre preliminary engineering an
offshore wind gust spectrum was derived from anemometer
Year Depth TetherslNumber recordings at British Maritime Technology as part of a joint
Hutton 1984 150m SegmentedIRodI16 industry project. Wind gust is of importance for horizontal
offset analysis.
Jolliet 1989 536m One Pieceffubularl12
Snorre 1992 310m Segmentedlfubular/16
High frequency springing excitation was observed during the
Auger 1993 872m Se~ented!fubular/12
Snorre model testing. This was of sufficient magnitude to be
Heidrun 1995 345m One Piece(fubular/16
of concern for tether fatigue. Based on the model test results
and some simplified assumptions the first engineering tether
TLPs built to date fatigue analysis with the combined effect of springing and
wave frequency effects were pedormed. Major software
Apart from breaking the first o/barrier, the Hutton platform is developments in UK, France and US to compute the second
important in setting ways to do things which the other order TLP hull loading coefficients were undertaken. A
platforms more or less have followed: major joint industry project on vertical resonant motion of
• All platforms have four tether groups with 4 or 3 tethers TLP was also run.
within each group.
• Hutton has 4 main columns and 2 auxiliary columns. The Ringing was first observed on the Hutton TIP model test.
other have 4 columns in a square. All platforms have a The magnitude was sufficient to arise considerable discussion
ring pontoon arrangement with no vertical or horizontal at the time, but the effect on tether design was small. The
supporting diagonal truss work. Heidrun model test revealed dramatic ringing. The magnitude
• Screw segmented partly buoyant tethers was nearly of ringing were comparable to the wave frequency response.
copied on . Snorre. A variation with near neutrally The fact that ringing became an item on Norwegian 1V News
buoyant segments was used on Auger. Jolliet and is a measure of its importance to the design of the Heidrun
Heidrun choose close to neutrally buoyant tethers towed tethers. Despite a recent completion of a major Norwegian
out and installed in one piece. An important common joint industry project on ringing there is still no reliable
feature of. all these tethers were that they were either method for computing the hydrodynamic loading causing
made up mechanical connected segments or by tubes ringing response. Extensive model testing using long wave
welded together. Fatigue of any such discontinuity will realisations is thus the only means at present to design for
be a major concern. ringing.
• Motion compensating systems or some ratcheting (a top
connector slips downwards by grips when force In summary, the original attraction of a resonance free
upwards) were used in the transition from the free structure is thus by no means longer valid. The major efforts
floating to tethered TLP condition. For all TLPs the in TIP design is, at present, directed towards understanding
tether attachment is complex and the cost of the operation and formulation of the various non-linear effects that drive all
is a significant part of the total cost of the hull. the 6 rigid body degrees of freedom.
• The pretension was set such that there would be a similar
margin against slack as against maximum tether loads. 4 TLPs are built in steel and the last, Heidrun, in concrete. The
Heidrun is an exception to this since short period zero choice of building material has been a major discussion within
tension will be acceptable during the very extreme the industry for some years.
conditions.
• Top and bottom flex elements are used to prevent tether
bending strain at the tether ends.
• Expensive measures are taken to reduce mispositioning
Future TLP Design Philosophy
• Continuos tension monitoring
Future TIPs will be designed for 1000m water depth and
• Means for tension adjustments throughout the platform
operation are available deeper. The platform itself will be little affected by increasing
• A replacement procedure for tethers and the ability to water depth. The importance of an efficient tether system,
survive in extreme weather with one tether missing however, increases dramatically with increasing water depth.
Focus on design of future TIPs should therefore be on the
• There is enough deck air gap to prevent any major wave
impacts from occuring aspects of the platform geometry that affects tether loading
and on the tether system itself.
• The decks are mated inshore and carried to the field with
the hull structure
Experience with 4 column TLPs has shown that the
Through the model testing and building of these structures indeterminate tether system implies some very heavy cost
items:
there was a growing appreciation of the importance of second
order forcing effects. By the time of the Hutton programme
67
• Accurate placement of foundations to limit terminations are not fully developed and bonding of fibres
mispositioning tether loads into heavy duty tethers have not yet been attempted.
• Systems for monitoring comer tether loads and facilities Considerable progress is made in civilian application of
for providing equal mean comer tension carbon fibre and it is likely that they wiJ] be appJied to deep
• Complicated mechanical systems for tether installation to water TLPs in some not too distant future.
provide equal comer tether loads in order to reduce the
cross-diagonal tension ballance error Parallel strand bridge cables have some of the excellent
mechanical properties of carbon fibre. It is suggested that a
A TLP with 3 groups of tethers, which will be staticalJy second look into their use for tethers could result in practical
determinate, will not require such devices and the foundations tether applications. Having experienced the high costs of
can be placed with larger tolerances without affecting tether manufacture and the complications of installation of the
behaviour. The main importance of a 3 corner UP, however, present drill pipe and neutrally buoyant tethers it is possible
is that all tethers share approximately the same loads despite that the problems of parallel strand tethers, found in the early
weather directions. A 4 comer TLP is ineffective in this 70's, would not be so serious after all. Regarding installation
respect since all tethers have an equal load sharing only for one might find practical methods by the use of available
beam and head seas. For diagonal seas, in fact for any sea off offshore vessels and equipment. Ie, viewed in retrospect and
the beam and head directions, the greater part of the loads are with today's technology one might find that this is an
taken by the leading and trailing tether groups whilst off- attractive tether solution after all.
diagonal tethers take less loads. With the near equal load
sharing of the 3 comer TLP the maximum load level in one The purpose of a mean tether pretension is to provide
group is less thus requiring less tether cross sectional material horizontal restoring through the geometric pendulum stiffness
than for a corresponding 4 comer version. Current studies and to prevent the tethers from going slack. Pretension in
indicate that 12 tethers are feasible for a 3 corner TLP whilst excess of what is strictly required is a Joss of a potentional
16 would be required for the 4 comer equivalent. This is thus deck pay load. To improve the economic feasibility of future
an important area for savings since tethers are important cost deep water TLPs one should investigate how far the mean
items. pretension can be reduced without any reduction in
functionality or safety. Less pretension causes higher mean
To mount a deck on 3 columns might seem complicated and and dynamic offsets. This is probably acceptable or one
costly. However, the experience from, ego the Beryl A might be able to design against this. The other problem is
Condeep showed that this was not necessarily so. In fact, the slack tethers. However, temporary zero tension tethers, ie. for
additional cantilevering of decli girders could result in a short periods in extreme situations, has been shown not to
weight effective design. introduce snapping type loads or any motion instabilities. One
should have in mind that, due to lateral hydrodynamic and
A deep water TLP, to be economic, would probably have inertia loads, any reduction in pretension will be taken up by
relatively flexible tethers, meaning h~gher heave, roll and sideways tether deflection and some positive tension will thus
pitch resonant periods than for the present UPs. The more probably be maintained in any situation. During retension the
flexible tethers might mean that the accurate tension tethers will straighten gradually since the lateral
adjustments within one comer would not be required. Tethers hydrodynamics will make the tethers act as a hardening
of premeasured lengths with fixed end terminations, spring. Snap loads on retensioning is thus not likely to occur.
accounting for any foundation heeling, could therefore be This has been shown analytically (Brekke.1986) and by some
employed. Any differential elongation, caused by measuring model tests. It is believed that more work in this field will be
errors, leading to a variation in mean tension within the done on future TLPs.
tethers of a corner would probably be acceptable. Any tether
length errors across the 3 comers would imply an For triangular UPs, it is possible to reduce the horizontal
insignificant static heeling angle. excursions considerably by introducing non-vertical tethers,
ie. incline the tethers outwards towards the foundations. This
There is clearly a water depth limitation to the use of the increases the maximum tether forces a little, but it also seems
present dril1 pipe and neutrally buoyant tethers. The weight of to increase the margin against zero tension. Ie. the effect of
the drill pipe will be to heavy and the hydrostatic pressure on increased horizontal excursions, caused by low mean
the neutrally buoyant tethers will reduce their efficiency. One pretension, can be counteracted by outwardly inclining the
solution to the hydrostatic colapse problem is to make the tethers. The fact that this wiJ] work is a feature of 3 comer
upper part neutrally buoyant and make the lower part thick TLPs since inclined tethers on 4 comer TLPs would cause
wall, heavier than water. The difficulty of installing such very high mispositioning forces.
long tethers is another reason for the depth limitation.
The basic TLP problem stems from the fact that all 6 rigid
Carbon fibre, or tethers of similar materials, are the most body degrees of freedom resonances will be excited by either
obvious choice for very deep water TLPs. They are light slow or fast varying forces. As mentioned these excitation
weight, have high stiffness, high breaking strength and have forces are generally low. Since the damping is also low the
good fatigue properties. Unfortunately, the cost is still very dynamic amplification will thus be high. The behaviour of
high, present world production is not sufficient, end low damped systems can be significantly improved by just a

68
small increase in damping whilst the effect tapers off as • Easy to make water tight divisions to satisfy flooding
damping increases. Rather than to design to satisfy highly regulations
amplified resonances, which is the present practise, it could be • Easy connection to steel deck
profitable to introduce damping mechanisms which will
reduce the greater part of the amplification. Such damping
devices can be passive or active. The space industry, aircraft Concluding Remarks
industry, tall rise buildings and even the car industry uses
damping devices extensively. People having driven cars with It is the views of the authors that TLP design in its present
a worn out suspension system will appreciate their importance form has reached a fairly mature technical status. It is
in driving comfort and safety. Present efforts show that to suggested, that along the path of developments a number of
identify damping devices which meet approval in the critical critical decisions has been made, or in many cases naturally
offshore environment and with long trouble free service life is evolved through some industry consensus. At present some
not easy. of these decisions appear as hard rules hindering more
efficient TLPs to be developed.
Air gap, or the deck elevation relative to the still water level is
also a critical dimension. It affects stability during tow-out By attempting to scrutinise the background for some of these
and thus the amount of weight to be carried to the field. It rules it is hoped that they can be judged in the perspective of
also affects the centre of gravity during operation and thus future needs for safe and economically efficient deep water
tether behaviour. Further, the increase in column height TLPs.
affects the cost of the structure and there is an effect on wind
force coefficients both due to the increase in projected area It is the views of the authors that TLPs have the potential for
and due to the wind velocity increase with height. A considerable refinements. However, this will require the oil
reduction of column diameters reduce wave upwelling and industry, engineering companies and regulatory bodies to
thus some of the need for the present high deck elevations. have open minds.
Also, it is suggested that the present practice of not allowing
waves to hit the underside of the deck might be too restrictive. It is the views of the authors that future cost efficient TLPs
Wave slams with associated noise and vibrations as well as will be based on the fo1lowing principles:
some water over the deck structure is acceptable to mariners.
With decks properly designed to withstand such loads this • Tethers grouped in 3 corners. The pontoons could be
should be acceptable to offshore personnel as well. There is triangular os circular ring pontoons
one note of warning since tether loads could be severely e Tethers of fixed length with premounted end
affected. The buoyancy load of a wave submerging a part of terminations. Thus no complicated top connectors
the deck will need to be restrained by the tethers. A wave • No facilities for adjustment of tether tension
hitting the horizontal or vertical part of a deck girder or a deck • Simpler field instal1ation not needing a heavy marine
house will set up ringing type tether loads. spread
• No requirement for a missing tether
The present platform/tether installation methods with the
• Concrete pontoon and lower columns. Steel columns
associated complicated mechanical equipments, some of from the, ego water line and upwards.
which is constructed for the installation only, accounts for a
fair proportion of the cost of a TLP hu1l. Since the cost of the
existing proven methods are known other methods could be
developed to the stage where they could be ranked against the References
present methods.
Anon (1974) New Platform begins Offshore Tests The Oil
Since there is now considerable experience in coJllstructing and Gas Journal, May 1975
TLPs in both steel and concrete it seems natural to think of a Brekke J N and Gardner T N (1986) Analysis of Brief tension
hybrid TLP where the two materials can be use to the best Loss in TLP Tethers OMAE, Tokyo, April 1986
benifit to the industry. Concrete seems the obvious choice for MacDonald, R D (1974) The Design and Field testing of
pontoons for the following reasons: the Triton Tension Leg Platform and it Future
• Cheep construction within a graving dock Applications for Petroleum Production and Processing in
• Resistant to hydrostatic pressure Deep Water Proceedings OTC'74
• Ease in making smooth transitions in complicated NOU (1981) The Alexander L Kielland Accident (in
pontoon/column joints Norwergian), Norges Offentlige Utredninger NOU
• No maintenance thus minimizing the need for internal 1981:11
access. (Maintance of gangways, ladders, manholes etc. is Sandnres P C, Lindeberg PG, Salvesen 0, Natvig B J
costly for platform with Ufe expectency of say 50 years) Norwegian Patents NP 14544,1973 and NP 140279. Also
patents in UK, USA and Sweeden
Steel would be the obvious choice for the upper part of the
columns for the folJowing reasons:
• Low weight

69