fully balanced in all positions when empty.

The counter weights being combined into one rotating assembly, this arrangement keeps the number of moving parts to a minimum and is particularly suitable where manual separation is considered desirable. The RCMA is a medium priced arm which meets the most stringent design specifications and is available with removable elbows at both the apex and trunion to enable seal changes to be undertaken without the use of heavy lifting equipment.

RCMA (Rotart Counter Marine Arm)

Manual quick connectors

They permit manual flanging operations in as little as 10% of ordinary flange connection line because the quick connectors are secured by a tightening a few clamps. A friction device holds the clamp in an open position while the arm and connector are moved into position. Each clamp is tightened individually to assure a leak proof seal even when shipboard flanges are nor perfectly uniform. Available in the clamps being rotated by handle bar by application of the helical cam pressure principle or being tightened by turning screws, complete tightening can be accomplished.

26

A Manual Quick Connector

Hydraulic Quick Connector

As an integral part of the loading arm, the hydraulic quick connector is easily positioned and hydraulically connected to the ships manifold in a matter of seconds. Independent reversible fluid motors drive a screw to tighten each clamp to assure positive connections even with irregular manifold flanges. No power is required to maintain the flange connection even with irregular manifold flanges. In case of power failure clamps may be closed or opened manually.

HYDRAULIC QUICK CONNECTORS

27

LNG containment systems

Many types of containment systems have been designed and developed for purpose of LNG containment. Material and structural conformation are the primary considerations in designing a system for use. Most of these systems have been designed for installation aboard ocean going LNG carriers. However these designs have been incorporated in FRSUs around the world.

Design philosophy of tank construction

LEAK BEFORE FAILURE

The features of the construction design of the tank are that the tank is designed utilizing breaking mechanics; secondary barriers are also considered to maintain the following 3 step safety: No fatigue cracks due to repeated loads shall be produced in the tank construction. The degree of cumulative fatigue damage to a ships cargo tank shall be 0.1 or less. The design will ensure that no cracks develop to such an extent that they penetrate the inside and outside surface of the tank shell, if any cracks are produced. Should a crack penetrate the tank shell and causes LNG to leak, the tank construction shall not collapse and the hull construction shall maintain conditions without thermal stress and enable the repair of the tank. SMALL LEAK PROTECTION SYSTEM (SLPS)

The cargo tank insulation system shall be designed to meet the requirement of the Small Leak Protection System which form the essential part of spherical LNG containment concept of moss Rosenberg Verft type. In case of crack propagation through tank material, it shall be ensured by the SLPS that the leakage of the LNG is detected in a very early stage of crack initiation. SLPS have following functions:

28

GAS DETECTION IN INSULATION SPACE Since the LNG leakage for upper or lower hemisphere produces some amount of cargo vapor due to the heat flux through the tank insulation, the leaked LNG cargo shall float towards the equatorial region or wedge space where a gas sampling is conducted.

INSULATION AS SPRAY SHIELD

The expanded foam material shall in itself have an effective mechanical resistance to cargo leakage. The aluminum foil of 0.25mm thick applied on a warm surface of the insulation shall act as an ultimate barrier for the cargo leakage.

DRAINING OF LEAKED CARGO LIQUID

The leaked cargo liquid shall be drained by gravity through the annular space between the tank and the insulation to the liquid collector (drip pan) through the drain pipe for upper hemisphere, to the South Pole for lower hemisphere and to liquid carrier.

Types of containment systems

All LNG terminals are constructed with double hulls. This construction method increases the integrity of the hull system, provides insulation for the LNG and provides protection for the cargo tanks in case of an accident Four basic tank designs have been developed for LNG ship containment and transport.

1. MEMBRANE: Non- self- supporting tanks with thin layer completely supported by insulation 2. INDEPENDENT: Self supporting tanks not forming part of the ships hull. The three categories of independent tanks are

29

3. SEMI-MEMBRANE: Non self- supporting tanks with a layer, flat parts of which are supported by insulation, whereas the rounded parts of the layer are designed to accommodate the thermal and other expansion or contraction. 4. INTEGRAL Tanks are integrated as a structural part of the ships hull. TYPE A : tanks designed by conventional rule and procedure TYPE B: Tanks designed by highly reliable tools to determine stress level, fatigue life and crack propagation. TYPE C: Tanks meeting pressure vessel criteria.

Most of the above mentioned tank types are constructed of metallic material. Besides these there are several types of non-metallic cargo containment systems.

The features of each cargo containment system are briefly described below.

Membrane tank system

A membrane system is formed by installing thermal insulating material into the hull of the ship and covering the surface with a metallic membrane. The purpose of the membrane is to maintain liquid-tightness so as to prevent any leakage of the cargo liquid. The load of the cargo liquid acts directly from the thermal insulating material to the hull. The membrane system makes the deck flat and the smaller size of the ship compared with a moss system.

In the membrane tank type, the insulation material is installed on inner hull and its surface is covered with a metallic membrane sheet. This containment system aims at reducing the metallic material exposed to low temperatures. The membrane keeps

30

liquid-tightness to prevent cargo leakage, and has no strength against cargo load. The cargo liquid load acts directly on the hull via the insulation material, therefore the insulation structure must have not only insulation performance but also strength. There are two principal types of membrane system in use.

Gaz Transport system

Comprises of 0.5 mm thick INVAR (36% nickel steel) primary barrier attached to inner (cold) surface of 200mmthick pearlite filled boxes used as primary insulation. INVAR is chosen for the membranes because of its very low coefficient of thermal expansion, thus making expansion joints, or corrugation, in the barriers unnecessary.

Technigaz Membrane system

Comprises a primary barrier of 1.2mm thick stainless steel raised WAFFLES to allow for expansion or contraction. It adopts corrugated stainless steel (both longitudinally and transversely) as the membrane. The thermal expansion and contraction is absorbed by the corrugation.

Independent spherical tank system (Moss system)

With the independent tank type, the hull and tanks are independent structures and selfsupporting tanks are arranged inside the hull. Therefore, deformation due to thermal expansion and contraction is not directly conveyed to the hull. The liquid cargo load in the tanks acts on the self-supporting tanks, not directly on the insulation material, and all loads act on the tank supporting members. Therefore sufficient strength and insulation performance are required for the supporting structure. Secondary barriers are required to be installed from the viewpoint of hull protection against leakage of

31

LNG in an accident emergency. A major drawback of this model in the FRSU design was that it took up additional space on the deck.

32

Semi Prismatic Type B (SPB) Tanks

The tank is of stiffened plate structure of aluminum alloy or 9%Ni steel and covered by PUF insulation and supported by tank supports and chocks made of specially reinforced plywood. The tank is subdivided by a centerline liquid tight bulkhead and a swash bulkhead into 4 spaces. Because of this, natural frequency of the liquid inside tank is far from that of ships motion, eliminating any chance of resonance of the liquid cargo and ship two motions. Therefore no sloshing problem is expected and any level loading in tank is always possible. This enables partial loaded voyage, quick dispatch from the berth in emergency, and this makes SPB best suited to FPSO, FSRU etc. in which tanks are always half loaded. Because of the nature of stiffened plate structure, the tank has the same strength against inner and outer pressure SPB does not need differential pressure control between hold space and tank, while membrane and moss are weak against outer pressure and differential pressure control is essential to them. Moreover, the hold space is used as inspection space facilitating inspection and maintenance.

33

SPB has plenty of advantages as follows: 1. Less fuel expense Less horse power with compact and ideal hull form

2. More cargo delivery Less heel (pump well) Less BOG No spray in ballast voyage

3. Less down time

Less influence from bad weather Shorter dry-docking time (2 weeks / 5 years)

4.Less Maintenance Expense

Simple visual inspection (type-B) No vacuum global test No purging of insulation space No heating coil between tanks Easy hull painting Easy deck-fitting maintenance

5. Easy Operation No differential pressure control No temperature profile control No cargo level restriction No heating coil operation Easy navigation

34

Choice of tank
Thus considering the operational and structural advantages of the Membrane type tanks with the SPB design with minimum sloshing problem has been selected for installation on board the FRSU.

Design possibilities for the future

35

Pyramid Tanks
The key feature of the pyramid tank's unique shape reduces free surface area, thus reducing the high impact sloshing loads and resonance period in the tank. The free surface reduction resulting from the design is important because, according to hydrodynamic experts at ABS, the impact pressure due to sloshing motion from the cryogenic liquid cargo inside tanks is one of the most critical load factors when designing containment systems for LNG carriers.

Boil off from tanks and its disposal

Since LNG tankers rely on insulation rather than refrigeration to keep their cargo refrigerated, a small percentage will "boil off. Traditionally, LNG tankers have utilized this boil off as fuel in steam turbine based propulsion plant.

That's changing. One way or another, the diesel is coming on board. One approach is to reliquefy the boil off gas, so that it remains as valuable cargo.

Various ways of handling boil off are:

Reliquification Use of LNG as a transportation fuel

Reliquification
While reliquefaction is widely used in gas handling on land, it has been used on board ship as well as on FRSU. Recently, the technology for reliquefying LNG has been matured and commercialised. The present analysis is based on the Moss Reliquefaction, sold worldwide by Hamworthy KSE. The patented system (Moss RS) for reliquefying boil-off gas,

36

establishes a solution for pumping LNG back to the tanks and selling more LNG to the buyers of gas. The three basic systems adopted are: Liquefaction system, cascade Liquefaction system, expander cycle Liquefaction system, mixed refrigerant cycle

Cascade Re-liquefaction System

A cascade liquefaction system uses a series of refrigeration loops, each having a single refrigerant, to cool the natural gas stream to successively lower temperatures. A cascade system generally works as follows. The first refrigeration loop utilizes propane as the refrigerant to cool the natural gas to about -40F, and to condense the refrigerant used in the second loop. The second refrigeration loop uses either ethane or ethylene as the refrigerant to cool the natural gas to about -120F, and to condense the refrigerant used in the third loop. The third loop uses methane or nitrogen as the refrigerant to further cool and condense the natural gas. The major pieces of equipment in a cascade system are the compressors (at least one compressor for each refrigerant), a cooling water system or ambient air coolers, refrigerant accumulators, expansion valves, and numerous heat exchangers. In the interest of energy efficiency, many of the heat exchangers and other parts of the refrigeration loops will be grouped inside one or more cold boxes.

37

Liquefication system, Expander cycle

An expander cycle liquefaction system makes use of a turbo-expander to chill the incoming natural gas stream, and liquefy a small portion of that stream. Passing high pressure natural gas through a turbo-expander causes a large decrease in pressure of the gas, resulting in a decrease in temperature of the gas. By passing the chilled natural gas through one or more heat exchangers, a portion of the gas stream can be liquefied. Generally, no more than about 10% of the incoming natural gas stream can be liquefied in an expander cycle.

Liquefication, mixed refrigeration cycle

38

The mixed refrigerant cycle uses a mixture of refrigerants (such as propane, ethane, methane, and nitrogen) within a single refrigeration loop. The various stages of refrigeration are accomplished by a series of pressure reduction steps. At each pressure reduction step, the liquid is partially flashed, which produces a colder liquid and vapor. The colder liquid is used as the refrigerant for the next stage of the refrigeration cycle. In this way, the mixed refrigerant cycle produces a series of refrigeration loops at different temperatures, much like the cascade cycle. One of the most common types of mixed refrigerant system uses a separate refrigeration loop, with propane as the refrigerant, to pre-cool the gas feed before it is introduced into the mixed refrigerant system.

39

Moss reliquification system

The boil-off gas reliquefaction concept is based on a closed nitrogen cycle extracting heat from the boil-off gas. Several novel features such as separation and removal of incondensable components have resulted in a compact system with low power consumption. The concept has the following technical merits: The nitrogen in the LNG boil-off gas (BOG) is not reliquefied; this results in reduced nitrogen in the tanks during the voyage, better control of tank pressure and lower power requirement for the RS system. The system uses only proven components with extensive references from airseparation and peak-shaving plants world-wide The system is prefabricated on skids for easy installation and hook-up. The system has automatic capacity control. The system can be stopped when the cargo pumps are in operation. This eliminates the need for extra generator capacity. During ballast voyage, the cargo tank temperature can be maintained by spraying reliquefied LNG back into the cargo tanks. The system must be installed with 100% redundancy. No extra personnel are required for operation and maintenance. The process can be described as follows: The LNG boil-off is compressed by the low duty (LD) compressor (BOG compressor), and sent directly to the so-called cold box. The cold box in which the boil-off is reliquefied is cooled by a closed refrigeration loop (Brayton cycle). Nitrogen is the working medium.

40

Standard Moss RS reliquefaction system

Nitrogen Cycle

Nitrogen compressor/expander

41

The cryogenic temperature inside the cold box is produced by means of a nitrogen compression-expansion cycle, shown in Fig. Nitrogen gas at a pressure of 13.5 bar is compressed to 57 bar in a 3-stage centrifugal compressor. The gas is cooled by water (seawater or indirect) after each stage. After the last cooler, the gas is led to the warm part of the cryogenic heat ex- hanger where it is pre-cooled to about -110C and then expanded to a pressure of 14.5 bar in the expander. The gas leaves the expander at about -163C and is then introduced into the cold part of the cryogenic heat exchanger where it cools and reliquefies the boil-off gas to LNG.

The nitrogen then continues through the warm part of the cryogenic heat exchanger before it is returned to the suction side of the 3-stage compressor. The N2compressor/expander unit is three-stage integrated gear centrifugal compressor with one expander stage. The unit has a gear with 4 pinions where each of the 4 wheels is coupled to a separate pinion. The result is that the expander work goes directly into the gearbox and relieves the electric motor.

The advantages of this solution are: More compact design Reduced cost Improved control of the refrigeration Reduced power consumption.

Small containers to discharge it as transportation fuel

One of the barriers to the use of LNG as a transportation fuel is the lack of costeffective refueling infrastructure that interfaces between a LNG distribution system and fleets of vehicles Fortunately, this barrier can be eliminated by the introduction of a distributed, fleet-size refueling system. This system takes LNG from an LNG pipeline network, purifies and enriches the gas as necessary, liquefies the gas, stores the liquid and delivers LCNG or LNG to vehicular fuel tanks under automated control. This refueling system is capable of operating in several different modes with various feed gas pressures and fleet refueling needs. LCNG is produced by a flow through the LNG pump/compressor from the storage tank, in counter flow through

42

vaporizer/heat exchanger and into the LCNG buffer tank/dispenser. Both LCNG and LNG could be dispensed simultaneously as desired by the fleet operator. LNG can be transferred as quickly as gasoline or diesel. On-site liquefaction requires smaller, less expensive storage tanks compared to refueling systems supplied by off-site liquefaction. Reliquefaction of boil-off gaseous NG coming from the storage tank, pressure adjustments during transfer and cool-down of transfer lines can be readily fed back to the liquefier in CFS systems. Weathering is no longer a problem because the boil-off vapors are returned to the storage tank. Pipeline NG has a range of compositions and pressures depending on location; this system is largely immune to these variations because the NG is cooled in a process stream separate from the liquefier internal working fluid. High purity methane (98 + %) can be provided as well as a more consistent grade of LCNG due to mixing in the LNG storage tank and drawing from the bottom of the tank. The existence of the NG collection and distribution pipeline network is an essential element of being competitive with gasoline or diesel fuel. Gaseous fuel infrastructure will be attractive to fleet owners/operators who already have pipelines bringing NG to nearby operations for space heating, industrial processing or electricity generation. While LNG could use the same centralized liquefier plus tanker distribution of gasoline/diesel fuel, the existence or creation of a pipeline network offers the possibility of distributed refueling systems with the advantages of more autonomous fuel supply.

SHORE DISCHARGE PIPIELINE

Introduction
Increased activity in the field of liquefied natural gas (LNG) has brought to the minds of many gas-industry representatives the possibility of pipelining natural gas in its liquid, rather than gaseous, form. The advantage of liquefying the gas for transportation is that more standard cubic feet of natural gas can be carried through a given cross-sectional area of pipe in comparison with pressurized transmission systems atambient temperature. Any advantage of increased throughput must, however, beweighed against the necessary costs to initially liquefy the gas and to keep it cold through insulation and reliquefaction (LNG pipelines will, most likely, be operated over a temperature range of -259 to -140 F). Also, increased costs may be

43

expected in order to circumvent design problems resulting from the cryogenic temperature.

In general, three broad areas in which the potential for transmitting LNG would appear to exist are: (1) Short imperative transfers, such as in-plant piping and lines from storage to loading or unloading facilities. (2) Optional transport methods, where a choice between trucking, barging, or pipelining LNG might exist (e. g., transporting LNG from a central liquefaction plant to satellite storage depots or between storage depots) (3) Base-load operations where transmission of LNG might be considered to markets over relatively long distances.

Pipe Materials
Fracture-toughness properties are important from two viewpoints: propagation resistance and initiation resistance. If a fracture should start, propagation resistance should keep it as short as possible. Because the decompression velocity of LNG is expected to be high (perhaps higher than any possible fracture speed), all fractures might be expected to be short, whether brittle or ductile. If this is true, it allows consideration of materials which have transition temperatures below the service temperature. Furthermore, because the LNG pipeline will have a pronounced temperature profile, it may be possible to telescope fracture-toughness properties, using higher transition temperature materials at the warmer temperatures. Therefore, the overriding criteria for material acceptance from a fracture-toughness viewpoint might be its initiation resistance. Austenitic stainless steels meet many of the requirements for low-temperature service Aluminum alloys are considered good candidates for use in an LNG pipeline. It appears that good strength can be achieved and resistance to fracture propagation appears satisfactory, at least for several alloys. Fracture-initiation resistance, however, is uncertain and will have to be determined.

44

Both Inconel and Monel have excellent properties at low temperatures and would be candidates for an LNG pipeline if they were not so expensive. Unless overall costs are significantly reduced by the fact that expansion bellows or loops are not needed, their use does not appear to be justified from economic considerations. Some of the present quenched and tempered steels such as 5Ni-Cr-Mo-V and HY-80 may be suitable for LNG pipeline service if it can be demonstrated that (1) Fracture speeds are sufficiently below the decompression velocity of LNG (2) Resistance to fracture initiation is sufficient. In order to be suitable for the proposed service, a material must meet the following engineering requirements: (1) Exhibit adequate yield strength (2) Be ductile enough to withstand fabrication into pipe and field bending (3) Possess adequate weld ability for mill welding (if not seamless) field welding (4) Possess suitable fracture-toughness properties at operating temperatures.

Pumping- and Cooling-Station Spacing

The three types considered are: (1) Short imperative transfers (less than 1 mile), such as in-plant piping and lines from storage to loading or unloading facilities. (2) Optional transport methods (1 to 10 miles) where a choice between trucking, barging, or pipelining LNG might exist. (3) Base-load operations where long-distance LNG transmission might be considered.

Case A B C D

Pressure drop, psi 120 120 20 20

Terminal temperature, F -249 -258 -248 -258

Recommended Pump discharge press, psig 135 135 30 30

Figures 5 through 8 present the pumping and cooling station spacing for Cases A through D

45

46

Case A indicates pipe insulated with 2 inches of urethane, a pumping station would be required every 25 miles, and a cooling station would be required every 100 miles. For Case B (Figure 6) the pumping-station spacing is the same as for Case A, because the flow and pipe diameter are identical. However, the distance between cooling stations

47

has increased to 375 miles. For transmission lines under 100 miles it appears that Case A is the more economical system, because intermediate cooling stations are not required for either Case A or Case B, but the construction and material costs would be much greater for a high-vacuum-jacketed line than for a line mechanically insulated.

Figure 7 indicates that for Case C the pumping and cooling station spacing is 130 and 175 miles, respectively. LNG enters the first cooling station at -140 F and 895 psi, and is discharged at -259 F and 895 psi. The pressure drop from 895 to 700 psi locates the third pumping station 83 miles from the first cooling station. The distance between cooling stations can be increased by increasing the thickness of urethane insulation from 2 to 4 inches. Because of the high-vacuum multilayer insulation used for Case D (Figure 8), a cooling station is required every 1950 miles. For this case, the major heat sources are the fluid friction along the pipe and the heat inputs at the various pumping stations. A pumping station is required every 130 miles, the same as for Case C. For long-distance transmission of LNG, the power requirements were broken down into three categories: (1) Pump power necessary to raise the pressure of LNG from 0 to 1000 psig at the initial pumping station (2) Pump power required to overcome the 300 psig pressure drop between stations (3) Cooling power necessary to prevent LNG from flashing because of the heat influx and the heat generated due to internal friction

Design considerations
Piping systems for transporting liquefied natural gas are not basically different from those for the more familiar fluids. But, some differences must be considered by the designer because of the temperature range encountered - ambient to -259 F. These differences are due to differences in the properties of materials at low temperatures, the emphasis placed on the selection of an economic insulating system, and 48

the necessity of providing flexibility in the system due to thermal contraction of the pipeline. Insulations required for piping systems for liquefied gases range from no insulation to the best vacuum-type insulation. The type of insulation used for a particular application is determined by several considerations: allowable heat leak, material and installation costs, operating and maintenance costs, ruggedness, length, and frequency of use. The insulations covered in this report fall into two classes: (1) non-vacuum or mechanical- type insulation and (2) vacuum insulation. With the former, the insulation is fabricated by shaping and fitting pieces of insulation to cover the pipe; the seams between the pieces can be sealed by a suitable mastic, epoxy, etc., or by an external covering. The foamed-in-place plastics which also belong to the first class of insulations may not require a vapor barrier if they themselves have closed cells, do not have fissures, and bond to the surfaces that they insulate. However, this cannot be relied upon and vapor barriers of some sort appear to be advisable for the foamed plastics. Vacuum insulations are in effect two concentric pipes with a vacuum in the annulus. In such insulations the heat transport due to gas conduction and convection is reduced (to a negligible amount in most cases) by removing most of the gas from the insulation space, i.e. , by evacuation. Long-term permeation through the pipe or jacket and out gassing (i e., release of gas from foreign materials within the vacuum space or from the pipe or jacket metal) would probably require periodic vacuum pumping to maintain heat losses below those for mechanical insulation.

Fully Restrained line

The use of Invar pipe without expansion joints shows several practical advantages over the other materials. This pipe has very low thermal stresses from expansion and cool down. For this reason Invar pipe construction is similar to conventional transmission-line practice - no expansion joints or intermediate anchors are required and the pipe can follow the earth surface contours to a reasonable degree. Invar has been used satisfactorily in cryogenic applications although further work needs to be done to determine realistic design stresses for this material.

49

Figure 14 shows a proposed design of Invar pipe with foamed-in-place insulation. For a 40-foot length of pipe, 36 feet of the pipe are insulated with factory installed foamed-in-place insulation complete with either a mastic or a light-gauge-metal vapor barrier. After the girth weld is made in the field, foamed-in-place insulation and a vapor barrier is applied to the ends of the pipe which were not factory insulated to complete the installation.

A schematic diagram of an Invar V-J line is presented in Figure 15. This line is concentrically placed in the jacket and supported by a low heat-conducting suspension system such as Teflon, glass, ceramic, stainless steel pins or balls, or impregnated fiber glass, thus reducing solid conduction. The openings between the inner line and jacket at the ends of the jacket are sealed by a thin stainless steel seal having a long heat path. Various methods are used to create a long-heat-path seal at these ends, thus reducing heat loss by conduction.

50

Expansion loops
In conventional piping systems, sufficient flexibility is obtained by the use of expansion loops. The expansion-loop spacing required for uniform-thermal-cool down piping flexibility will be examined next.

Expansion-loop design for an underground line requires the construction of chamber to allow free movement of the loop piping.

The loop layout used for the LNG line is shown below:

51

The loop spacing U is limited by longitudinal stresses in the main line pipe outside the loops. The major longitudinal stress in this pipe is a friction stress between the insulation and the pipe. This stress was computed conservatively by assuming that the pipe was backfilled to a depth of 6 feet and that the insulation did not support any of the earth load.

Expansion Joints
In-line expansion joints of stainless steel bellow construction are available for cryogenic service. Because the requirements of the cryogenic industry up to the present time have been for low-pressure transfer lines, design and development of bellows for high-pressure applications at liquid-methane temperatures have not been extensive. Several manufacturers have indicated interest in the design and manufacture of bellows for a pressure of 1000 psi; however, they also indicated that the present state of the art would limit the axial movement of such a bellows to approximately 1 inch for either a lo-inch or a 20-inch-diameter pipe. The longitudinal contraction for a temperature drop from 60 F to -259 F would be 107 inches/mile for 9 percent nickel, 154 inches/mile for stainless steel, and 202 inches/mile for aluminum. Because of the large number of bellows required to compensate for the axial movement of the LNG transfer line, the use of bellows does not appear to be economically sound at the present time for a line pressure of 1000 psi. An LNG line using expansion joints only would have the bellows located in the back-to-back as shown schematically below:

Where x = intermediate anchor = expansion joint. Figure 16 shows details of a possible expansion joint, anchor, and concrete enclosure.

52

Insulation materials
There are many materials available which can be used for low-temperature insulation. The majority of these materials depend on finely divided air spaces to retard the transfer of heat. The insulating materials can be categorized as belonging generally to the following four classes:

(1) Fibrous (2) Cellular (3) Granular (4) Reflective. The fibrous insulation such as hair felt, asbestos, and mineral wool offer little resistance to the passage of water vapor through the air spaces, have a relatively high thermal conductivity, and are not recommended for LNG pipeline usage. The granular and reflective materials will be discussed in the sections dealing with powder insulations and multilayer insulations.

Cellular Insulation
The cellular insulations such as foamed glass, plastic foams, cork, and balsa wood are discussed presently. The following properties of foams are desirable for

53

monolithic insulation systems exposed to LNG temperatures: (1) Low tensile and compressive moduli (2) Low thermal- expansion coefficient (3) Low Poissons ratio (4) Low thermal conductivity (5) Low density. It is desired that the foamed insulations provide good resistance to heat transfer under the imposed temperature boundaries, withstand the maximum internal stress generated by mechanical and thermal forces, and resist deformations in a radial direction due to earth loads.

Vapour barrier
The primary problem associated with the use of conventional insulations stems from the requirement of a vapor barrier. Effective low- temperature insulation involves effective moisture control because most insulating materials will absorb water or transmit water vapor. (This, of course, ceases to be a problem when the temperature of the warmest part of the insulation drops below 32 F. But, since the pipeline may be operated intermittently it must be considered). Water vapor which gets into the insulation will condense when cooled to its dew point. The condensate can solidify and produce a layer with the relatively high thermal conductivity of the solid. Also, some of the water will freeze inside the insulation and cause local fissures, thus allowing even more water vapor to enter. These cumulative effects will soon damage the insulation system within the affected section. The vapor barrier may be any type of material that will retard the absorption of water vapor by the insulation. Metals, foils, plastics, and various types of coating may be used, although the most practical and most widely used vapor barriers are a combination of mastics and membranes.

Cryogenic Pumping Equipment

The fundamental concepts and methods which are adequate for the fluid-mechanical design of pumps for common fluids are applicable to liquefied gases The performance characteristics will differ somewhat because of such things as greater leakage due to

54

lower viscosity The efficiency may be lower due to the fact that a less-dense fluid requires less total power input while the so-called fixed losses (such as those occurring in bearings) will not decrease appreciably. For specified flow rates, Figure 79 indicates the maximum discharge pressures which may be realized from currently available pumping equipment(44). Flows from 200 to 500 MMCFD would require liquid flow rates approximately from 1700 to 4200 gallons per minute and thus necessitate the use of centrifugal pumps. This is because positive-displacement pumps are generally used for low-capacity high-pressure applications. Most of the design information such as performance curves is considered to be proprietary information and released only to potential customers. Several manufacturers indicated that for the flows and pressures considered in this report, centrifugal pumps for methane could be built with technical knowledge and capabilities presently available. Although specific efficiencies were not given, the general consensus of opinion is that the liquid-methane pumps would have efficiencies in the neighborhood of 75 percent. This would need to be confirmed.

55

Valves
The problems arising in connection with valves for liquefied gases stem from heat leak, distortion due to temperature gradients, icing together of arts which must have relative motion, freezing of packing, and seat tightness. Beard (45) discusses several types of cryogenic valves which include shutoff, ball, butterfly, solenoid, gate, and safety. The article also contains a valve material compatibility guide, The performance of a cryogenic butterfly valve incorporating a spherical sealing surface is discussed in Reference 46. Although most butterfly valves leaked severely at cryogenic temperatures several years ago, butterfly valves are produced today with a guaranteed maximum leakage of 2.5 seem of helium per inch of diameter at -320 F .

Temperature
Thermocouples have been used widely in cryogenic applications. Their main assets are adequate accuracy, good response, and moderate cost. Copper-constantan junctions have been used satisfactorily from room temperature down to -423 F. Vapor-pressure thermometers are extremely accurate and simple to operate, but have limited ranges. Gas-filled thermometers are difficult to use and are undesirable for application in an LNG piping system. Resistance thermometers are accurate, stable, and linear down to -423 F, butthey are very expensive. Semiconductor thermometers depend on the rapid change of resistance with temperature of the metallic-oxide sensing elements for their accuracy. These thermometers are also relatively simple to operate, but have the disadvantages of nonlinear calibration curves which are nearly exponential. In certain instances, the semiconductor thermometers have been found to be unstable.

56

Fire Safety
Followings are the constrains to be followed for protection against Fire according to ABS Pipes & Materials In the case that of a fire on the facility, a fixed water fire fighting systems should be provided. The piping for the fire system should be arranged so that the water sources come from at least two different locations on the vessel. The piping system should be maintained against corrosion. If the engineers decide to put drains in the facility, then they must be placed at the lowest points. If the heat damages or renders some material, then that material is not to be used to in the fire piping systems. The valves on the system must pass fire test acceptances according to ABS standards. The plastics on the system are required to also meet guidelines written by ABS. Fire hoses that are located on the production deck should be constructed of materials resistant to oils, chemical deteriorations, mildew, rot, and offshore environmental exposures. They have to be comprised of a non-collapsible material with a maximum length of 100 ft. The hoses will be mounted on reels.

The water spray systems material is design from and must comply with a list of ABS standards found in (ABS 3-8/5.1.4).

Pumps
The primary connections and the standby pumps must be as far from each other as possible. The fire system must have at least two self-priming fire pumps that are independent of each other. The pumps are also independent from the entire system, having their own source of power, fuel supply, electric

57

cables, etc. Their placement on the vessel should be such that if a fire occurs, then it would not affect both pumps. Both the primary and standby fire pumps should be able to sustain the maximum probable water demand, which is described as the total water requirement for protection of the largest single fire area plus two jets of fire pressure at pressure of at least (50 psi). Pump drivers can be operated by diesel engines, natural engines, or electric motors which must comply with ABS and API standards, specifically (ABS3-5.2) and API RP 14G. The fuel systems should be able to operate for a minimum of 18 hours. ABS (3-8/7a) provides three floating installation fire pump arrangement scenarios as guidelines for a fire system design. The operability and control for this system is also carefully addressed by ABS.

Control Station
The fire stations need to be located on the perimeter of the process area. In case of an emergency such as a fire, at least two emergency control stations should be provided, both of which must have an efficient means of communication and process system shutdown, etc.

If the facility is shut down, then the following services must still be operable: 1. Emergency lighting 2. General Alarm 3. Blowout preventing control system 4. Public address system 5. Distress and radio communications

The helipad station also has a fire fighting requirement that it must follow. It must be constructed of steel or any other material that has the same fire integrity properties. The ABS manual refers to the Steel Vessel Rules for these requirements.

58

Fire Detectors & Structural protection

The facility should also be provided with fire detectors, gas detectors, smoke detectors, an alarm panel, fire and gas detection wiring, and a general alarm. Combustible gas detectors must be in accordance with API RP 14c and API RP 14F standards. Structural fire protection requirements address the need for protection boundaries which separate spaces onboard the installation from the process facility equipment. Table 3a and Table 3b on 3-8/9.3 describe fire integrity of bulkheads separating adjacent spaces, and fire integrity of decks separating adjacent spaces. It includes accommodation spaces, stairways, open decks, corridors, and other types of places on the facility which have open areas.

Firewalls should be designed from uncontrollable flare font wellheads. Its shut in pressure is required to be a minimum of 600psi. Firewalls are also used to protect from fire hazard to the vessels.

On the terminal, it is required to have marshalling areas for personnel before entering the lifeboats. Steel and Fiber Reinforced Plastic are to be used to construct these marshalling areas, as well as the lifesaving embankment areas. The material chosen must be in accordance with the Flag Administration in Appendix 3 of the ABS manual.

For the process equipment, a fixed water spray is installed to maintain a cool environment for the equipment. The other purpose is to reduce the risk of an escalated fire.

Personal Safety Requirements

Two escape routes have to be considered in the design of the terminal with markings and adequate lighting. There must also be escape route plans, which are to be displayed in and around the facility. Lifesaving equipment such as lifeboats, life rafts, like buoys, life jacks, work vest, and breathing apparatus must all be available for the personnel. Each of those items listed must follow the rules in section (ABS 3-8/15.5). Personal

59

safety equipment and safety measures are a very important issue, which are required to comply with ABS standards. iremans outfits and breathing apparatus are stored in an appropriate container together. Its material should be water resistant and radiate heat from fires. Surfaces that are exposed should not exceed temperatures of 71C. Surfaces that exceed 482C need to be protected from combustible gas, as well as weather, mechanical wear, and physical damage.

Others
The water spray systems should be provided with an automatic start. The minimum flow of the monitors is 500 gallons per minute at 100 psig. The nozzles on the fire stations must have diameters of at least 0.5 inches.

60

Summary of Conclusions and Recommendations

Three types of floating facilities were considered but it was the ship-shape barge design which was selected. This decision was based on the fact that it is a well established, proven design and it allows flexibility when selecting the mooring lines. The overall ship dimensions are 340 m in length between perpendiculars, 360 m in overall length, 65 m in width, 33 m in height, a calculated draft of 11.58m, and a displacement of approximately 265,000 tonnes. It has five semi-prismatic LNG tanks and ten J-shaped ballast tanks, five on the port side and five on the starboard. The vessel is oriented with the bow facing in the southwest direction. The mooring system has been optimized using the software programs, and the terminal uses a spread-moored system with twelve lines. Each line is 86.9 m (285 feet long), consists of 114.3 mm (4 -inch) chain, and has a Vryhof Stevpris Mk5 15 t anchor. The terminal satisfies all of the stability criteria for intact, single-tank damage, and double-tank damaged stability. The bending moment calculations for global loading have been completed, and a minimum outer plate thickness has been calculated to be 0.032 m (1.25 inches) based on the ships moment of inertia in crosssection. The heave, pitch and roll natural uncoupled periods are smaller than the natural period of the wave and peak wave for loaded and unloaded conditions, and thus no harmonic resonance on the vessel presented itself. The In-Air Flexible offloading arms can sustain offloading process in a one-year return period as required by the design constraints. In the event of an unavailability of this type of arm the FMC mechanical offloading arms can be substituted in their place without necessitating a redesign or reanalysis of the hydrodynamic motions. The team decided to have the terminal built in Spain, and its total cost including transportation is roughly US $563 million. Since this was a front-end concept design, several assumptions were addressed in the more detailed final engineering design. A more accurate estimation of the wind loads on the topside equipment, accommodations module, and supply cranes were made once the specific structural elements within each was defined. Similarly, the total mass of the vessel was accurately determined only after the structural elements of the hull (keel beam, longitudinal and transverse supports, scantlings, etc) were selected 61

and optimized. As a result, the 15 percent confidence margin added to the mass in this report was reduced since the masses was more accurately defined, consequently altering the terminals hydrodynamic motions and reducing the vessels total mass, draft, and final cost. In addition, completion of the structural engineering design will allow for a more accurate measure of the vessels moment of inertia in cross-section and the resultant changes in maximum bending moment, bending stress, and natural period. Design of the LNG intake manifold and the piping between the manifold and the five SPB tanks is paramount in that an efficient design will minimize the required offloading time of any carriers that use the terminal.

62