concretesubmarine.com/ FORUM

Large Floating Concrete LNG/LPG Offshore Platforms
By Dale Berner, Ph.D. and Ben C. Gerwick

A number of studies and projects for large floating concrete offshore platforms with storage capacity for liquefied
natural gas, LNG, and liquefied petroleum gas, LPG, have demonstrated the feasibility and significant advantages
of such structures. These platforms, which tend to be large in order to take advantage of economies of scale, can be
used for production, storage, off-loading, and re-gasification. With the recent increase in the price of gas, the
general consensus in the industry is that several such platforms (using either steel or concrete hulls) will be built
within the next ten years.
Cryogenic liquids such as LPG and LNG, are typically stored at temperatures ranging from approximately –40 ºC,
to –160 ºC, respectively, and are highly flammable. Prestressed concrete hulls have several advantages over steel
hulls for containing such cryogenic liquids including: excellent resistance to cryogenic temperatures and thermal
shock, and excellent marine performance. Indeed, one such floating concrete structure that Ben C. Gerwick, Inc.
has worked on, the Ardjuna Sakti LPG terminal, has been in service since 1975, with good service performance.
This paper discusses several key aspects of such platforms including:
1) Material selection and performance, including the use of lightweight concrete and cryogenic steels.
2) Thermal stress and strain considerations, including both global and local concerns.
3) Platform construction consideration, including construction in areas with shallow water.

INTRODUCTION
Liquefied Natural Gas, LNG, at approximately –160ºC,
and Liquefied Petroleum Gas, LPG, at approximately –
40ºC, occupy approximately 630 times, and 310 times,
less volume than their respective gas forms at stand
temperature and pressure. This volume reduction allows
these cryogenic products to be transported overseas.
Floating offshore LNG/LPG storage platforms can
serve at either end of the transport route, either as a
liquefaction, or as a re-gasification, facility. Such
floating facilities offer several advantages over
conventional land-based facilities for offshore fields,
such as: a) the elimination of both harbor facilities and
of long pipelines from the production platform to shore;
b) the ability to relocate the facilities from one field to
another (which encourages more rapid depletion of
small fields); c) faster field development time,
particularly for remote fields, with little or no site
development work; d) normal processing procedures
such as separation and dehydration can be performed
along with the LNG/LPG manufacturing; e) better
control of construction schedules and costs; f) the plant
can be commissioned in transit to the operation site,
thus reducing the time to start-up; and e) enhanced
safety by isolating the facilities away from populated
areas.
Ben C. Gerwick, Inc. has been involved with floating
LNG/LPG offshore platforms ever since before the
Ardjuna Sakti LPG floating concrete terminal (60,000
DWT, see Figure 1) was install in Indonesia, for
ARCO, in 1975.
While the Ardjuna Sakti has proven the viability of
such cryogenic storage platform technology for LPG,
only relatively recently have advances in technology
and changes in market conditions made a similarly
compelling case for the development of floating LNG
platforms. Currently, this case has only been made
convincingly for relatively benign environments such as
West Africa, Southeast Asia, and the Caribbean.
2
Fig. 1. Ardjuna Sakti LPG Floating Concrete Terminal
Numerous cryogenic storage platforms will soon be
required to: a) develop stranded offshore gas fields
identified over the past several decades but which are
too small to develop conventionally; b) improve the
economics of associated gas which can no longer be
flared, and which is an expense re-inject; and c) to
expedite, and improve the economics of large offshore
gas fields. Still further platforms will be required to
serve as floating re-gasification facilities near populated
markets. To further improve their economics, and also
to improve their technical viability, proposed
LPG/LNG liquefaction platforms tend to be large;
which allows the platform to support large topside
processing facilities, and which typically results in
improved wave response. Topside liquefaction
processing facilities are sensitive to wave induce
motions; however, recent processing advancement
allow processing to continue over a wider range of
motion.
Currently, ship-shaped FPSO’s, are the most commonly
proposed platform configuration. Such platforms
provide the combined benefits of: a) a large topside
area; b) a large storage volume; c) the ability to
weathervane into on-coming waves; d) the feasibility of
conventional offloading schemes in benign areas, and e)
good economics. However, this configuration is not
suitable for deepwater developments, and has
undesirable motions in harsh wave climates. For
harsher wave climates, alternate configurations have
been proposed including: torus-shapes (See Figure 2),
spar platforms, and semi-submersible vessels. Another
variation of platform configuration, which is not
discussed in this paper, is the float-in Gravity Base
Structure, GBS, which has been seriously proposed for
several locations around the world, and which are
typically designed using concrete hulls.
Fig. 2. Torus-Shaped Floating Concrete LNG Terminal
STEEL HULLS VS. CONCRETE HULLS
Prestressed concrete has been used in the onshore
storage of liquefied gases since the 1950’s; however,
relatively few concrete containment structures were
constructed prior to the 1970’s. Since the 1970’s, the
proportion of onshore concrete containment tanks has
been gaining rapidly, compared to steel containment
tanks, largely due to prestressed concrete’s excellent
service record.
The use of steel hulls for floating platforms has also
generally been more common offshore than the use of
concrete hulls. However, concrete hulls have
performed well offshore, and they offer a unique
combination of advantages over steel hulls for
cryogenic containment; which should make them more
common for floating LNG/LPG applications than is
generally the case offshore. Table 1 provides a
comparison of advantages for both concrete, and steel,
hulls for LNG storage.
3
Table 1: Comparison of Advantages Between Concrete and Steel Hulls for LNG Storage
ADVANTAGES FOR CONCRETE HULLS ADVANTAGES FOR STEEL HULLS
Superior Cryogenic Behavior Fabrication in Existing Shipyards
Good Separation of Processing/Storage Potentially Lower First Cost for One Hull
Reduced Down-Time due to Inspection Traditional Engineering
Reduced Maintenance Costs Traditional Construction
Economies of Scale More Steel Fabricators are Available
Good Impact Resistance More Steel Designers are Available
Low Center of Gravity/Good Station Keeping
Behavior/Reduced Motions
Greater Flexibility Reduces Thermal Stresses
Excellent Fatigue Life Not Subject to Freeze-Thaw Damage
High Mass Moment of Inertia Prestressing Not Required
Slower Thermal Response/Better Insulation Impermeable to Gas and Liquids
Resistance to Fatigue and Crack Propagation Similar to Numerous LNG/LPG Ships
Resistance to Buckling Does not Require a Membrane Liner
Properties of Prestressing and Mild Reinforcing
Steel at Low Temperatures
As a composite material, the behavior of
reinforced/prestressed concrete is influenced as much
by the mild reinforcing, and high-strength prestressing,
steel as by the concrete. At low temperatures, such
steels typically gain from 4 to 30 percent in yield and/or
ultimate tensile strength, while they typically become
somewhat less ductile, depending on the steel
composition. For proper design against stress
concentrations, rapid loadings, and fatigue, as well as
for favorable post-ultimate behavior at both ambient
and low temperatures, the reinforcing steel must exhibit
a combination of both strength and ductility, jointly
termed toughness. Toughness is a property measured
independently of an element’s configuration and is a
measure of the amount of energy that the material can
absorb before failure.
Reinforcing steels can broadly be divided into ferritic
steels, with a body-centered cubic crystal lattice
structure, and austenitic steels with a face-centered
cubic lattice structure. Ferritic steels tend to be less
expensive but more brittle, and less tough, than
austenitic steels. Toughness can be promoted by
controlling the steel processing to induce a fine-grained
lattice structure containing austensite; however,
depending on the design requirements it may be more
economical to select a ferritic steel with a fine grain
structure with a minimum number of lattice defects.
Processes that promote tough, ductile steels, include
deoxidizing with silicon, or aluminum, heat treating,
and cold-working.
High-strength, prestressing, steels that retain a good
degree of toughness at low temperatures can be divided
into three groups: a) carbon-manganese cold-drawn
wire and strands; b) high-tensile strength steel alloy
bars, and c) stainless steels.
Cold-drawn wires and strands are most commonly used
for prestressing at both ambient and low temperatures
due to their combination of relatively low cost, high
strength and good ductility. The usual process for
manufacturing prestressing wire and strands entails a
controlled cooling procedure from the austenite
temperature range to give the ferritic steel a desirable
microstructure for cold-drawing. The cold working
improves the strength and cryogenic properties of the
steel, but cold-working also induces undesirable
stresses that are typically removed by further heat
treatment. These cold-drawn steels are not weldable
due to their high carbon contents.
Both the steel alloy bars and stainless steels are more
costly than cold-drawn wires and strands; but their
typically greater ductility may make them appropriate
for selected applications. High-tensile strength alloy
steels include silicon chromium wires and bars, 9-
percent nickel steel bars, and micro-alloyed bars. These
alloy steels may contain combinations of ferritic and
austenitic matrix structures. High-strength stainless
steels contain large quantities of chromium (15 to
20%), and nickel (8 to 10 percent); which induce an
austenitic lattice structure. These stainless steels have
excellent cryogenic properties, they are highly
corrosion resistant, and are readily welded; however,
they are also relatively expensive.
4
Common grades of mild reinforcing steel, with from 0.1
to 0.2 percent carbon, generally retain sufficient
toughness above 0ºF (-18ºC) to be used where lower
temperatures are not expected. Below this temperature,
higher quality control standards need to be established
for the steel with regard to chemical composition, heat
treatment, and cold working. Due consideration also
needs to be made for the potential dynamic loading and
for the desired level of safety.
Cryogenic grades of passive reinforcement can be
provided by replacing the mild reinforcing steel with
passive prestressing steel; however, this can add
unnecessary expense, and may not fully satisfy ductility
requirements, strain capacity requirements, or
steel/concrete bond requirements. Consequently, some
special grades of deformed rebars have been developed
for cryogenic use by maintaining high quality control in
processing.
Cryogenic grades of mild reinforcing steels can be
roughly divided into three groups: a) carbon-manganese
steels; b) austenitic steels; and c) austenitic-ferritic
steels. Carbon-manganese steels, are the most costeffective
of these three groups, and have a fine-grained
ferritic matrix structure, frequently modified by the
additions of such elements as niobium, and
molybdenum. The austenitic steels may have limited
applications due to their high cost and typically have
lower yield and ultimate tensile strengths than the
carbon-manganese steels. The austenitic-ferritic steels
incorporate nickel to produce a fine nickel-rich ferrite
grain structure bounded by small quantities of reformed
austenite to provide ductility.
Prestressed Concrete Material Performance
Due to conservative design and the inherent capacity of
load-carrying membrane structures, failures related
purely to design have been rare. Several recent failures;
however, have indicated that deficiencies in materials
and/or construction methods, especially when combined
with factors such a thermal stress, fatigue degradation,
creep, and freeze-thaw damage, can be important.
Selected desirable material behavior of prestressed
concrete for hulls:
1. At low temperatures, the strength of concrete
increases, while neither the prestressing steel nor
the reinforced concrete become brittle at low
temperatures.
2. When loaded to failure, prestressed concrete is not
subject to sudden progressive collapse. Also,
cracks from temporary overload tend to close upon
removal of the overload.
3. Proper prestressing will keep the concrete
watertight and free from any major through
cracking. Additionally, concrete, that is
continually moist, will continue to hydrate, which
can “heal” any minor static cracks that have
occurred. Furthermore, moist lightweight concrete
has extremely low permeability at cryogenic
temperatures.
Selected undesirable material behavior of prestressed
concrete for hulls:
1. If moisture in the form of water vapor is present, it
will permeate the concrete and migrate to the cold
face, where it will freeze, thus tending to debond
the membrane (thus the nitrogen gas, that is
typically circulated between the hull and the
primary containment tanks, should be dessicated to
prevent this).
2. Global thermal shortening in un-insulated tank
walls can potentially produce through-thickness
cracks unless offset by prestressing (which can
readily be provided).
3. If no membrane is used, the cryogenic liquid will
permeate the outer layer of concrete over time. If
the tank is then taken out of service and warms up
rapidly, gas will be generated that may cause local
spalling (therefore membrane liners are required).
Most concrete offshore structures have been built with
standard weight concrete, including the Ardjuna Sakti
and the N'KOSSA floating terminals. However, the
more sophisticated concrete platforms including Troll,
Hibernia, and Draugen have used modified density
concrete, and some of the most relevant platforms
including Super CIDS, Tarsiut, and Heidrun have used
high-strength lightweight concrete. Therefore, a brief
discussion of the key differences in these materials is
provided.
Standard weight concrete is more widely available, is
less expensive, and generally has both higher
compressive, and higher shear, strength than either
modified density, or lightweight, concrete. However,
as cited above, although most offshore
platforms/terminals use standard weight concrete, the
5
more sophisticated, and more relevant, platforms have
used either specified density, or lightweight, concrete
for reasons including: 1) reduced deformation (thermal,
shrinkage, etc.) loads; 2) reduced draft, and 3) better
durability. By using modified density concrete little, or
no, reduction in design compressive strength or design
shear strength need to be taken into account, while
lightweight concrete normally requires a reduction in
both of these design values. However, the use of
lightweight concrete has several advantages at
cryogenic temperatures as compared to standard weight
concrete including:
Plain lightweight concrete has approximately twice the
tensile strain at cracking as plain standard weight
concrete, and thus can sustain twice the thermal
deformations before cracking. Furthermore, the
modulus of elasticity of lightweight concrete can be
about half that of standard weight concrete, thus
prestressing of the lightweight concrete will result in
proportionately greater resistance to thermally induced
cracking than standard weight concrete.
Air entrained lightweight concrete has been 200 and
1,000 times lower permeability than standard weight
concrete, at both room and cryogenic temperatures.
This raises the possibility of eliminating the vapor
barrier if high strength air entrained lightweight
concrete is used.
The coefficient of thermal expansion/contraction of
lightweight concrete is approximately 7 x 10-6/ºC as
compared to 10 x 10-6/ºC for standard weight concrete.
This results in lower thermal deformations and lower
thermally induced stresses/strains for lightweight
concrete.
The coefficient of thermal conductivity is
approximately 30% lower for high strength lightweight
concrete as compared to standard weight concrete. This
would result in lower boil-off of LNG/LPG for a
lightweight concrete hull.
Thermal Strains/Stresses
Thermal contraction and expansion of members of an
LNG/LPG storage vessel, during both scheduled, and
accidental, cooling and warming cycles, induce thermal
stresses that depend on: a) internal restraint for
members subject to thermal gradients, b) external
restraint of members subject to an absolute temperature
change, and c) differential restraint for members
containing materials with different coefficients of
thermal expansion/contraction. For simple cases,
restraint factors can be used together with the
coefficients of thermal expansion/contraction and the
temperature change in order to calculate the induced
thermal stresses. For more complicated cases, finite
element analysis are required (see Figure 3).
Figure 3. F.E.M. Model for Thermal Analysis of
Floating Concrete LNG Terminal
Determination of thermal strains requires an
understanding of the coefficient of thermal
contraction/expansion for the composite member
section; however, this composite behavior is dominated
by the concrete response, which in turn is primarily
influenced by the contraction/expansion of the
aggregate. Use of aggregates such as expanded shale
not minimizes thermal expansion and contractions, but
they also reduce the induced thermal stresses by
reducing the stiffness of the structure, thus allowing it
to comply with the expansions/contractions, which
occur.
Virtually all offshore LNG containment vessel designs
utilize a primary containment system that insulates the
concrete hull from the cryogenic liquid, because the
global thermal stresses induced in the hull during
cooling and warming cycles would be excessive.
However, the appropriate use of prestressing, double6
hull configuration, and material selection can allow
LPG floating containment vessel to use the inner hull as
a primary containment system subjected to cryogenic
temperatures on a regular basis, because the local and
global thermal stresses are more manageable. Indeed,
concrete offshore platforms deployed in the Arctic
regularly sustain temperatures colder than those
associated with containing LPG, and have performed
extremely well for decades (see Figure 4). This
consideration can result in significant savings in
containment costs for a concrete LPG vessel as
compared to a steel LPG vessel.
Figure 4. Arctic Platform at Cryogenic Temperatures
CONSTRUCTION CONSIDERATIONS
Not only to concrete floating platforms typical have
deeper drafts than comparable steel floating platforms,
but also concrete platforms typically demonstrate
improving economics with increasing size, relative to
steel platforms. Thus construction considerations for
such large-floating concrete LNG/LPG offshore
platforms merit careful evaluation due to the size and
draft relative to existing conventional facilities.
Small barge shaped concrete floating platforms such as
the Ardjuna Sakti LPG terminal, of the N’KOSSA oil
terminal, can be built in large conventional shipyards,
or existing graving docks. Similarly, square torusshaped
floating platforms such as that shown in Figure
2 have been designed to be formed by joining afloat
four smaller barge shaped concrete vessels that were
planned to be fabricated in large conventional
shipyards.
Deeper draft concrete semi-submersible platforms, such
as the Troll Oil platform in the North Sea, can either be
built by beginning construction in a graving dock, and
then float-out to deeper water for completion afloat, or
if deeper water is not available near shore, they can be
designed to be completed while floating on their
pontoons.
Still deeper draft concrete floating platforms such as
spars and semi-spars can either be built in a manner
similar to a conventional concrete Condeep GBS if
protected deep water is available for slip-forming the
concrete shaft(s) while floating in the protected deep
water (see Figure 5), or if such protected deep water is
not available then single shaft concrete spars can be
constructed horizontally and then floated out to deep
water for up-ending in a manner similar to steel spar
platforms.
Figure 5. Slip-Forming a Condeep Platform
RECOMMENDATIONS
In addition to the storage plans considered in this paper,
it is recommended that investigations be made to
determine the feasibility of storing and transporting
pressurized LNG, using floating prestressed concrete
structures. As noted previously, LNG at atmospheric
pressure is stable at approximately –162 ºC (-260 ºF),
while LNG stored at the pressure at 170 m (560 ft)
water depth is stable at approximately –109 ºC (-164
ºF). This could significantly reduce the CAPEX, and
OPEX of the liquefaction facilities. Furthermore,
demands on the membrane liners, and the time of cooldown/
warm-up, would be reduced.
Figures 6 and 7 show a conceptual prestressed concrete
storage/transport vessel suitable for containing
pressurized LNG. Prestressed concrete is economical
to use for the construction of such a large pressure
vessel, costing less than half that for steel construction.
7
Figure 6. Cross-Section of Conceptual Pressurized
LNG Storage/Transport Vessel
Figure 7. Longitudinal Cross-Section of Pressurized
LNG Storage/Transport Vessel
ACKNOWLEDGMENTS
Some of the information presented in this paper is
based on the findings of a study conducted for Mobil
Technology Company, to determine the structural
feasibility of storing LNG in concrete structures afloat
in the open ocean.
REFERENCES
1. Berner, D. E., Gerwick, B. C., and Polivka, M.,
“Static and Cyclic Behavior of Structural
Lightweight Concrete at Cryogenic Temperatures,”
ASTM, Special Technical Testing Publication 858,
1985.
2. Berner, D. E., Behavior of Prestressed Concrete
Subjected to Low Temperatures and Cyclic
Loading, Dissertation, U. C. Berkeley, 1984.

http://www.nmri.go.jp/main/cooperation/ujnr/24ujnr_paper_us/Offshore_Structures_and_Systems/OSS_Berner_Gerwick.pdf