SUBSEA CARRIER

Abstract

A subsea carrier (100) for transporting a fluid, e.g. CNG or crude, comprises a main body (101) for containing the fluid at a predetermined internal pressure, wherein the main body (101) preferably is made of concrete and designed to operate at a water depth where the external pressure substantially counteracts the internal pressure. The subsea carrier has a floating element (102) connected to the main body (101) by a stabilising cable (132), wherein the stabilising cable (132) comprises a first rope (321) for transmitting force and is attached to a first connector (134) that is movable with respect to the main body (101). A system wherein the subsea carrier is towed by a surface vessel (3) or is self-propelled and controlled remotely is also disclosed. The subsea carrier (100) reduces operational costs relative to subsea carriers with traditional control surfaces and ballasting systems at large cargo volumes, e.g. 150 000 m3 or more.

Description

BACKGROUND OF THE INVENTION

Field of the Invention

The present invention relates to a subsea carrier for a fluid, e.g. oil or compressed natural gas.

Prior and Related Art

In the following, hydrocarbons serves as an example of fluid cargo. However, embodiments of the carrier described herein can carry fresh water, inorganic gas or other fluid chemical compounds. Here, the term “hydrocarbons” means oil and gas, typically extracted from a subterranean reservoir. Natural gas, or NG, means hydrocarbons that are gaseous at normal conditions, i.e. 0° C. and 1 bar, such as methane, whereas the hydrocarbons also include components that are liquid at normal conditions, e.g. crude.

Starting from a reservoir and a pressure of several hundred bar, a mixture of hydrocarbons, water and other substances, e.g. H₂S, are brought to the surface under closely super-vised conditions. Temperature and pressure is controlled and chemicals may be added, e.g. to prevent or inhibit formation of hydrates and/or scaling. An initial processing may be performed on the field, e.g. to remove some or most of the water, sand and corrosive substances like H₂S. After initial processing, the hydrocarbons are transported downstream toward a final destination, e.g. a refinery for cracking the mixture into individual components, a chemical factory for further processing or a consumer for heating. In some cases, keeping the hydrocarbons, for example NG, at a pressure of a few hundred bar from production to a loading point and further during transport may save energy and costs compared to decompressing the gas for treatment and thereafter compressing it for transport. In the following, “transport” is defined as bringing the hydrocarbons from a loading point, where all pre-processing has been performed, to an unloading point where no post-processing has commenced.

Transport of hydrocarbons in this sense includes transport by pipeline and transport by carrier, each with known advantages and disadvantages. If carrier transport is preferred over pipeline transport, a comparison of alternative carriers remains.

A first example concerns transport of natural gas. LNG (liquefied natural gas) has about twice the density of CNG (compressed natural gas) at 250 bar and standard temperature. Thus, if speed, initial and final treatment of the NG is disregarded, a CNG-carrier would need to have a little less than half the investment and operational costs compared to an LNG-carrier with the same loading volume, e.g. 150 000 m³, in order to be cost competitive in transporting a given mass of NG. Of course, a real comparison must also account for the speed of transport, cost of plants for cooling and evaporating LNG vs. compressing and decompressing CNG, as well as the cost for cryogenic equipment for storing and transporting LNG vs. similar equipment for CNG.

A second example concerns carrier size. In surface transport of crude, different sizes are favourable for different applications. For example, a Panamax-carrier able to pass through the Panama channel may be preferred for one application, whereas a VLCC (very large crude carrier) may be preferred for another. Similarly, surface gas carriers have different sizes for different applications. Thus, it is expected that subsea carriers for hydrocarbons will have different sizes depending on the cargo, the distance and other factors similar to those determining the favourable size of a surface carrier.

CA2028273A1 describes an unmanned subsea carrier made of concrete, wherein the external pressure from ambient water substantially compensates for an internal pressure. The carrier is loaded at a subsea loading point and unloaded at an unloading point. The loading and unloading points are preferably placed on solid rock due to the weight of the carrier.

In the following, a non-limiting example concerns a subsea carrier with a loading volume of 150 000 m³. This size is selected partly because it corresponds to a typical surface LNG-carrier, and partly to illustrate that subsea equipment does not scale linearly. For example, handling 200-300 bar in a 10 litre scuba tank is hardly comparable to handling 150 000 m³ CNG at 25 MPa (250 bar), and propelling and steering a subsea carrier of such a size is not directly comparable to propelling and steering a submarine a fraction of the size.

There are several reasons why traditional submarine designs are expensive and/or impractical for a large subsea carrier operating at an external pressure of 250 bar, i.e. at a water depth of 2 500 m.

Firstly, cabins and other facilities for a crew at an acceptable safety level 2 500 m below sea level would be expensive, so an unmanned carrier controlled remotely, e.g. from a surface vessel, would be advantageous.

Secondly, the size and required speed may render a traditional submarine design too expensive for a subsea carrier competing with a surface crude carrier or LNG-carrier. For example, some of the largest submarines ever built, the Typhoon class (Russian “Akuna” or “Shark”), has a displacement of less than 48 000 tonnes fully submerged. In some of these submarines, the missile launchers are replaced with cargo holds for up to 15 000 tonnes of cargo. In comparison, a subsea carrier with cargo volume for 150 000 m³ of CNG at 24 MPa may have a displacement of 190 000 tonnes, i.e. about four times the size and mass of a Typhoon. As the required force and power for propulsion and trajectory control depends on momentum, a smaller velocity can compensate for control challenges resulting from a very large mass to some extent, preferably to a level where affordable, propulsion, ballast pumps and control surfaces can handle the momentum. However, an economical transport requires a certain minimum speed or forward velocity, e.g. about 2.5 m/s (˜3-5 knots).

More particularly, depth control, steering and changes in forward speed are changes in momentum, and hence equal to a sum of impulses, e.g. a as an integral of a varying force F(t) over a response time, or equivalently as a constant mean force F applied over the response time. Accordingly, the force required to alter momentum at a given velocity increases with increasing mass and decreasing response time. Other factors must also be considered.

Depth control serves as an example. In particular, imagine a submarine moving forward at 5 m/s at an upward angle of 2°. This submarine would raise about 50 m in five minutes. Thus, the submarine must be able to change its momentum, in particular the direction of the velocity more than 2°, in a response time of five minutes to maintain a depth range within a set depth±50 m. A subsea carrier with four times the mass and the same speed than those of the submarine, would have four times the momentum, and hence require four times the control force to perform the same change of velocity in the response time of five minutes. This can be done by increasing the area of fins and control surfaces, angle of attack; hydrofoil profiles etc. all of which increase the drag, and thereby cost of operation. Finally, a large carrier is likely to be used for long distance transport, e.g. 5 000-10 000 km. The extra work caused by a larger drag over a long distance requires fuel, so the added operating costs may become considerable.

At zero forward speed, fins and control surfaces create no lift, so a traditional submarine design uses ballasting with water to control depth. The pumping rate must be sufficient to maintain the carrier within the predetermined range of depths, i.e. determining a response time as described above. The rate of water into or out of ballast tanks, and thus cost of tanks and pumps, increases with mass, as does the price for any thrusters intended for changing the momentum in any direction. Therefore, mainly due to the larger mass of a subsea carrier, the costs associated with control surfaces, ballast tanks with associated pumps and/or thrusters are relatively high, and may become prohibitive for a commercial subsea carrier that should have life-cycle cost, for example, less than half as much as a LNG-carrier of comparable cargo volume. Similar considerations apply to all spatial directions.

Thirdly, in a traditional submarine design, electrical power is used for propulsion and steering, i.e. to drive one or more thrusters, several actuators for external control surfaces and/or pumps to fill or empty ballast tanks. An electric generator powered by an engine within the submarine vessel, e.g. a combustion engine or a steam turbine, may provide the electrical power. Providing fuel and air for combustion in a submarine vessel requires space for fuel tanks and air tanks. Further, the cost of maintaining a combustion engine far below sea level is certainly higher than the cost of maintaining a similar engine at the surface. The costs associated with all of the above increase fast with the size of the carrier, and some costs increase with the depth of operation. In particular, operating a combustion engine at a depth of 2 500 m for years might not be considered for practical reasons, and might be found too expensive even if it was considered. While alternatives to a combustion engine are known for a military submarine, e.g. air independent propulsion systems such as nuclear power plants, chemical batteries, fuel cells or stirling engines, such alternative energy sources are generally not available, unsuitable and/or too expensive for a large commercial subsea carrier, and are not further described herein.

Thus, an objective of the present invention is to provide an improved subsea carrier solving at least one of the problems above while retaining the benefits of prior art. In particular, the invention should enable transporting hydrocarbons and other fluids at a lower cost per unit than what has been previously possible.

SUMMARY OF THE INVENTION

The above and other objectives are achieved by a subsea carrier for transporting a fluid according to claim 1, a system for transporting a fluid according to claim 10 and a method for operating the subsea carrier according to claim 16.

In a first aspect, the present invention relates to a subsea carrier for a fluid. The subsea carrier comprises a main body for containing the fluid at a predetermined internal pressure, wherein the main body is designed to operate at a water depth where the external pressure substantially counteracts the internal pressure. The subsea carrier further comprises a floating element connected to the main body by a stabilising cable, wherein the stabilising cable comprises a first rope for transmitting force, and is attached to a first connector that is movable with respect to the main body.

Preferably, the external pressure from ambient water exceeds the internal pressure such that a compressive force acts on the main body. Then, the main body can be made of a brittle and inexpensive material, e.g. concrete as known from prior art. However, embodiments with a somewhat positive internal differential pressure are also anticipated because handling a limited differential pressure acting on the hull structure, for example 10 bar rather than 250 bar, reduces the design requirements significantly. Thus, the term “substantially counteracts the internal pressure” includes embodiments wherein the pressure difference from the inside to the outside causes an economically viable design, for example up to 10% higher than the external pressure. In that case, the hull structure must be able to withstand the tensile stress resulting from an internal pressure greater than the external pressure.

The floating element can be located at any depth above the main body, and helps orienting the main body through the stabilising cable. More particularly, the stabilising cable exerts an upward force on an upper part of the main body to reduce roll, i.e. rotation about a longitudinal axis pointing in the direction of travel.

The first rope transmits the force acting upward on the main body, and can be a synthetic rope or a steel wire capable of transmitting the buoyancy force from the floating element. The first rope preferably has close-to-neutral buoyancy to avoid undesired vertical forces on rope and the floating element, and would thus be provided with buoyancy elements if a steel wire is selected as the rope.

The first connector is primarily movable in a longitudinal direction in order to adjust the pitch of the main body. The pitch should be adjusted such that the main body presents a minimum area perpendicular to the direction of travel at all times in order to minimise drag and thus operational costs. Equally important, e.g. a positive pitch causes the main body to move upward. This is undesired because the external pressure should exceed the internal pressure in order to compress the structure. The first connector may also be movable in a lateral direction in order to control roll, in particular at loading and unloading points.

In a preferred embodiment, the length of the stabilising cable exceeds the water depth. In this embodiment, the floating element floats on the surface, so that if the main body starts rising, the floating element rises out of the water. This reduces the buoyancy and thereby causes the main body to sink due to gravity. Opposite, if the main body starts sinking, an increased volume of the floating element is submerged. This increases the buoyancy and causes the main body to rise due to buoyancy transferred through the stabilising cable.

A preferred embodiment comprises a bridle for distributing a towing and lifting force. The bridle is preferably a length of steel wire running over two pulleys attached to the main body at two separate attachment points, e.g. one behind the other. Then, a connector attached to the steel wire can be moved along the main body by turning the pulleys in order to adjust the pitch. Similar pulleys can be employed to reduce roll of desired. Besides pulleys any other actuator may be employed to change the geometry of the bridle and to introduce forces to control pitch or roll of the main body.

Preferably, the main body comprises a ballast element connected under a tank element. The effect is to lower the centre of gravity and thus minimise roll. A suitable ballast material could be, for example, gravel of magnetite, i.e. iron ore, due to its relatively low cost and high density. In some applications, such ore could also be used in the walls of the main body.

Preferably, the main body comprises a first cargo compartment separated from a second cargo compartment by a movable, gas-tight sealing element. The main purpose of the gas-tight sealing element is to prevent gas in the first cargo compartment from dissolving in water within the second cargo compartment.

In a particularly preferred embodiment, the gas-tight sealing element is a flexible membrane. However, other sealing elements are anticipated. For example, one or more pipes could be placed within the tank volume, each pipe having one end open to the cargo compartment, the opposite end open to the ballast compartment and an axially movable piston. Several steel tanks known in the art, e.g. of a kind used in surface crude carriers, enclosed in a shell of concrete might still be less expensive than a surface carrier.

The subsea carrier preferably has at least one ballast tank for trimming and depth control. During a transport journey with a cargo fluid and during a return journey with a return fluid, the buoyancy of the main body is close to neutral, i.e. that a transport mass is close to the displacement of the main body. Ballast water is pumped into or out of the ballast tanks to adjust the transport mass to the actual displacement as known in the art. The ballast tanks can e.g. be made of relatively inexpensive steel pipes, and they can be oriented parallel to a length axis though the main body to help control roll. Separate ballast tanks at a front end and a rear end of the main body can be used to control pitch and depth.

In some embodiments, the main body comprises a control surface and/or a thruster for causing rotation of the main body about at least one of tree mutually perpendicular axes of rotation. Control surfaces and/or thrusters can replace or supplement the ballast tanks for trimming and depth control, and they need not be used actively for the entire journey. For example, the bottom must be down at the loading and unloading points, but roll does not matter during a journey. Also, ballast tanks can be used for depth control if the main body halts for some reason, whereas relatively small control surfaces can keep the main body within an allowed range of depths as long as the main body moves through the water. The actual choice of means for placing and orienting the main body within a body of water depends on the mass, response time etc. as discussed above.

In some embodiments, the main body comprises at least one thruster for causing translation of the main body along at least one of tree mutually perpendicular axes. Thereby, the main body does not need towing, but can be towing the floating element. Such thrusters can also lift the main body and/or shift the main body to one side with or without a rotation in any plane.

In a second aspect, the invention concerns a system for transporting a fluid comprising a subsea carrier as disclosed above, further comprising: a communication line for conveying control related signals between a controller at the surface and the main body and a power line supplying electrical power from an electric generator at the surface to the main body.

The communication line facilitates transmitting sensor data from the main body to a controller at the surface, and appropriate steering commands in the opposite direction. The communication line is preferably an electric conductor or a fibre-optic cable, as acoustic signalling through, for example, 2 500 m of water with varying salinity is expected to attenuate and/or distort the acoustic signal to a degree that the communication would be unreliable or inefficient due to low signal/noise ratios.

The power line enables an electric generator powered by a combustion engine at the surface, where air is readily available from the atmosphere, and one or more electric motors for driving steering devices on the main body, where the orientation of the main body is controlled most efficiently. The power line preferably comprises an electrically insulated metal wire, e.g. made of aluminium for low loss and low cost, provided with floating elements to make the power line neutrally buoyant. Such a power line may be similar to dynamic cables used in the subsea oil & gas industry and known in the art.

In a first embodiment, the system further comprises a surface vessel with the controller and the electric generator, and a control cable with the communication line and the power line. The control cable is preferably provided with a fairing in order to reduce drag. An embodiment without a towing cable would require a thruster for propulsion on the main body.

Preferably, the control cable in the first embodiment of the system also comprises a second rope for transmitting force, i.e. such that the surface vessel tows the subsea carrier. The second rope can be, for example, a synthetic rope or a steel wire like the first rope transferring forces between the main body and the floating element. However, in this embodiment, the second rope tows the main body in addition to the floating element, so the second rope must be able to sustain greater tension than the first rope. Furthermore, a steel wire might be preferred over a synthetic rope for the second cable because the synthetic rope is more elastic than the steel wire, and thus would apply a towing force on the main body varying over a greater range of elastic strain than a steel wire. Regardless of the selected material, the second rope is preferably included in a fairing together with the communication line and the power line.

In some embodiments, the control cable further comprises a fuel line from the main body to the surface vessel. Thus, hydrocarbons from the main body may feed a combustion engine driving the electric generator, and possibly the propulsion motor for the surface vessel.

In a second embodiment of the system, a thruster for propulsion on the main body is mandatory. In the second embodiment of the system the floating element comprises the controller and the electric generator, and the stabilising cable comprises the communication line and the power line. The stabilising cable is preferably equipped similar to the control cable of the first embodiment, in particular provided with a fairing and floating elements to make it neutrally buoyant.

As in the first embodiment of the system, the stabilising cable may comprise a fuel line from the main body to the floating element.

In a third aspect, the invention concerns a method for operating a subsea carrier with a first cargo compartment and a second cargo compartment. The method comprises the steps of: filling the first cargo compartment at a subsea loading point with a cargo fluid at the predetermined internal pressure; transporting the cargo fluid from the loading point to a subsea unloading point; expelling the cargo fluid from the first cargo compartment; filling the first cargo compartment at the unloading point with a return fluid at a second internal pressure; transporting the return fluid from the unloading point to the loading point. The method is distinguished in that expelling the cargo fluid involves allowing ambient water to flow into the second cargo compartment; and filling the first cargo compartment at the unloading point with a return fluid involves pumping water out from the second cargo compartment.

The cargo fluid can be any fluid defined in the introduction, e.g. CNG from a subterranean reservoir. The return fluid can be payload, e.g. N₂ or CO₂ intended for injection into a subterranean structure, e.g. for pressure support of the reservoir or for deposit in an aquifer or depleted reservoir. Thus, there is no clear distinction between the loading point and the unloading point, and the method works in either direction.

If the external pressure at the unloading point is greater than the predetermined internal pressure, expelling the cargo fluid is simply done by opening a throttle valve. For loading the first cargo compartment with the return fluid, the pressure difference between the second internal pressure and the external pressure determines the size and required performance of the water pump. As the second internal pressure is much greater than atmospheric pressure, the water pump can be relatively inexpensive.

To illustrate the method, assume that the cargo fluid is CNG at a predetermined internal pressure 240 MPa, the external pressure is 250 bar and the return fluid is dry air. During transport from the loading point to the unloading point, a certain volume of CNG, e.g. 150 000 m³, has a certain mass, in this example 33 900 tonnes. As CNG and oxygen in the air should not mix, the first cargo compartment should be completely void of CNG before air is loaded for the return journey. Hence, water at 250 bar flows into the second cargo compartments and expels the cargo fluid completely. The dry air for the return journey should also have a mass of 33 900 tonnes in order to achieve close to neutral buoyancy. This means the air must have a second internal pressure of 206 MPa. A pressure of 206 bar cannot expel water at 250 bar, so a pump handling a pressure difference of 44 bar is required. The pump(s) must also have a throughput that displaces 150 000 m³, of water in a reasonable time.

In some embodiments, the subsea loading point and/or the subsea unloading point comprises a subsea platform serving as a foundation for the main body during loading and/or unloading. The step of expelling the cargo fluid involves allowing the water to flow from a storage tank in the platform, and the step of filling the first cargo compartment with a return fluid involves pumping the water back into the storage tank.

Thereby, a main body arrives with near neutral buoyancy and settles on the platform. As the water flows from the platform to the main body or vice versa, the combined mass of the main body and platform does not change during loading or unloading. After loading or unloading, the main body leaves the platform with near neutral buoyancy. In other words, the weight working downward from the platform changes very little during loading and unloading. The platform can be provided with near neutral buoyancy when no main body is present, thereby further reducing the ground pressure. It can even hover above the seabed, e,g, over soft soil or rough surfaces. A further benefit of the closed loop for water is that contaminated water is not released into the environment at the loading and/or unloading point.

Further features and benefits appear in the independent claims and the following detailed description.

BRIEF DESCRIPTION OF THE DRAWINGS

The invention will be explained in greater detail by means of an exemplary embodiment and with reference to the accompanying drawings, in which:

FIG. 1 illustrates a first embodiment and some important forces acting on the system;

FIG. 2 illustrates a second embodiment;

FIG. 3 illustrates a third embodiment;

FIG. 4 illustrates a fairing with different components comprised in a cable;

FIG. 5 illustrates the fluid connections to a main body;

FIG. 6 illustrates schematically a segment of the main body;

FIGS. 7a-c illustrate a first method of ballasting;

FIGS. 8a-c illustrate a second method of ballasting;

FIG. 9 illustrates the main rotational axes of the main body and

FIG. 10 illustrates loading and unloading at a subsea platform,

FIG. 11a-c illustrates loading and unloading at a subsea platform in greater detail.

DETAILED DESCRIPTION

The drawings are schematic and intended to illustrate the principles of the invention. Hence, the drawings are not necessarily to scale, and numerous details are omitted for clarity.

An important aim of the invention is to carry a cargo of fluid, e.g. hydrocarbons, from a loading point to an unloading point along a predefined path as inexpensively as possible. In the present context, this implies keeping an elongated main body 101 aligned with a predetermined path in three dimensions fixed to the Earth, e.g. longitude, latitude and depth, by controlling local coordinates fixed to the main body 101, e.g. roll, pitch and yaw as illustrated with reference to FIG. 9.

FIG. 1 illustrates a first embodiment of a system according to the invention, comprising a subsea carrier 100 partly submerged in a body of water 1 having a surface 2. A towing vessel 3 on the surface 2 tows the subsea carrier 100 by means of a towing cable 32, in the claims expressed as a control cable with a second rope. The subsea carrier 100 comprises a main body 101 connected to a floating element 102 by a stabilising cable 132. The floating element 102 is at the surface 2 for passive depth control of the main body 101. More particularly, if the main body 101 starts to ascend, the floating element 102 starts to come out of the water 1, which immediately decreases a buoyancy B acting on the floating element 102, and hence decreases the upward pull from cable 132 on the main body 101. The reduced pull upward on the main body 101 causes a net vertical force acting downward, such that the main body 101 sinks. Reversely, if the main body 101 starts to sink, an increasing part of the floating element 102 comes below the surface 2, which immediately increases the buoyancy B such that the main body 101 moves toward the surface 2 of the body of water 1.

The non-limiting example mentioned in the introduction illustrates the order of magnitude. That is, a main body 101 could have a loading volume of 150 000 m³ and an internal pressure suitable for CNG, e.g. in the order of 250 bar or 25 MPa. The external pressure acting on the main body 101 should be somewhat larger in order to ensure a limited, compressive force on the walls of the main body 101. 250 bar corresponds to a depth of water of 2 500 m., such that a suitable towing depth would be 2 500 m±50 m if the differential pressure should be kept within limits 10 bar (˜100 m of water depth) apart. In this example, the stabilising cable 132 might be 2.5-3 km long and the towing cable 32 somewhat longer.

Some important forces working on the subsea carrier 100 are shown on FIG. 1. These are a towing force T, gravity or weight G, a resistive drag R acting on the main body 101, a drag L acting on the stabilising cable 132 and the buoyancy B. For simplicity, a towing force acting on the main body from the stabilising cable 132 and a drag acting on the towing cable 32 are not shown. The towing force T acts along the towing cable 32, and may be decomposed into a vertical component T_v and a horizontal component T_h. The vector sum of forces in the vertical direction may be varied around zero to provide controllable depth adjustments. Furthermore, the cables 32, 132 should preferably have approximately neutral buoyancy to minimize their effect on the vertical force balance.

A dynamic depth control using, for example, thrusters, ballast tanks and/or fins with control surfaces is anticipated. However, use of active elements should be kept at a minimum to minimise operational costs. The embodiment in FIG. 1 provides an inexpensive, passive depth control. It also provides a fail-safe behaviour in the case of power failure, preventing the main body from uncontrolled sinking or rising without electrical controls.

The force L acting on the cables 32, 132 generally have different directions and magnitudes at different depths due to their own motion during towing and different currents in the body of water 1.

In the horizontal direction, the horizontal component T_h of the towing force must overcome the resistance or drag R on the main body and a horizontal drag component of the force L acting on the cables 32 and 132. As known in the art, the drag R on the main body 101 can be modelled as a sum of pressure drag and friction drag, both of which increases with the towing speed. At an assumed towing speed around 5 knots or 2.5 m/s and a substantially cylindrical form, the friction drag is likely to dominate.

The drag component of L acting on the towing cable 32 and the stabilizing cable 132 can be estimated in a similar manner. Due to the length of the cables, their combined cross section and surface areas can be significant compared to the corresponding areas of the main body. Fairings and other techniques known in the art can reduce the drag from the cables.

The towing cable 32 is attached at an attachment point 34. The attachment point 34 for a towing cable 32 is preferably close to the centre of gravity of the main body 101, as a long arm from the centre of gravity to the attachment point 34 might provide an undesired permanent pitch or trim. For example, if the attachment point 34 was at the front of main body 101, the vertical towing component T_v could easily lift the forward end of main body 101, and thus provide a larger cross sectional area during towing. This will also introduce induced lift and induced drag similar to an aircraft wing. In addition, it will create interference drag from vortices forming behind the misaligned body. To compensate, extra trimming ballast in front and/or extra trimming buoyancy aft would be required to keep the main body 101 substantially parallel to the direction of travel.

The towing cable 32 is preferably attached to a bridle 340. The bridle 340 comprises a length of rope, e.g. steel wire, and adjusts itself to variations in the actual towing force T. The bridle 340 may also be dynamically adjusted to move the attachment point 34 back and forth along the main body 101. Thus, the bridle 340 contributes to avoid a permanent and unwanted pitch as discussed above. The bridle 340 can be attached at four points on the main body 101, and thereby decouple the towing force from rotations around the x-axis in addition to the rotation about the y-axis, i.e. to facilitate control of roll in addition to pitch.

A permanent, uncorrected pitch would move the main body 101 up or down in the water, which is potentially more harmful than added drag. Thus, it is understood that additional means, e.g. passive fins and/or dynamic control surfaces, should be provided to avoid undesired pitch. Similarly, an uncorrected yaw would cause the main body to move consistently to one side, thereby causing added drag and less accurate steering. Again, fins and/or dynamic control surfaces might be used to stabilize the main body and provide damping to control oscillations.

In FIG. 1, the floating element 102 is a vertically elongated buoy at the surface 2, preferably with a height of a few tens of meters. While not shown in the drawings, two or more physical floating elements may be provided. One purpose of the floating element is to provide a stabilising force through the stabilising cable 132 and the attachment point 134.

A floating element 102 on the surface 2 is exposed to oscillating forces from surface waves. In essence, the shorter wave components rise the sea surface with respect to the floating element 102, which is connected to the heavy main body 101. Thus, the shorter waves are transformed to long wave components with smaller amplitudes and longer periods than the original waves. The long wave components are more easily handled by a dynamic control system if required.

The stabilising cable 132 is connected to the main body 101 through a connector 134. The connector 134 is similar to the connector 34 for the towing cable. Likewise it is movable on the main body, e.g. to compensate for vertical component of towing force induced when the floating element 102 and cable 132 is towed through the water 1. The connector 134 may also be connected through a bridle (not shown).

As shown in FIG. 2, the floating element 102 can be disposed anywhere between the main body 101 and the surface 2. The main advantage of submersing the floating element 102 is a shorter stabilising cable 132, and hence less and more predictable drag L on the stabilising cable 132. The advantage of better control of orientation, in particular pitch, provided by a movable point of attachment and/or a bridle is retained in this embodiment. On the other hand, sensors and steering means on the main body 101 for depth control are required in this embodiment.

As noted in the introduction, several sizes of carriers are expected. Thus, the embodiment on FIG. 2 does not necessarily have the same dimensions as the one shown in FIG. 1.

In the embodiment on FIG. 2, the main body 101 is propelled in its longitudinal direction by a thruster 103. The cable 32 is assumed to provide power to the thruster 103, but not to provide a significant towing force. If the cable 32 does not impose any significant force on the main body 101, its attachment point 34 can be placed anywhere on the main body 101, including in the front as shown, without adverse consequences.

FIG. 3 shows yet another embodiment, wherein the subsea carrier 100 is self-propelled and controlled by means of a radio link 105. In particular, a floating element 102 near the surface 2 contains a conventional generator 4 comprising a combustion motor coupled to an electric generator. The fuel required in the combustion motor, e.g. natural gas, can be supplied from the main body 101 through the stabilising cable 132. The oxygen required for combustion can be supplied as ambient air through a snorkel 104. Thruster 103 receives electrical power through the stabilizing cable 132, and propels the main body 101 in a forward direction, i.e. provides a force F acting on the main body 101 in the longitudinal direction termed x in FIG. 9. Again, the embodiment on FIG. 3 may be suitable for the same or different sized carriers than the embodiments on FIG. 1 or 2.

In the embodiment in FIG. 3, sensors recording position, direction and orientation of the main body 101 at certain points in time, transmit the relevant data by means of the radio link 105 to a remote controller (not shown). The controller (not shown) compares the data to a planned course and depth, and returns appropriate steering commands though the radio link 105. These steering commands are used to adjust the speed, course and/or depth of the main body 101 using suitable steering means. An autonomous control without a radio link, based on a pre-programmed trajectory, is also possible for some or all parts of the operation.

It is understood that a similar control loop would be practical or required also in the embodiments shown on FIGS. 1 and 2. In these embodiments, the radio link 105 would be replaced with a communication line 322 (FIG. 3) in the cable 32, and the controller would be aboard the surface vessel 3.

Summarising the above, the bridle 134 limits a misalignment of the main body 101 that would increase the drag and the risk for an undesired ascent. The reduced changes of momentum, i.e. directional deviation from a desired trajectory can be handled by relatively small dynamic control surfaces, further limiting drag and providing other benefits such as an ability to dampen control oscillations. The need for an active control system can be further reduced by using a passive system for depth control, e.g. a buoy floating on the surface.

A control system is required to limit the changes of momentum, and thereby the force and time required to amend the deviations. In particular, accurate and inexpensive sensors, e.g. three-dimensional MEMS accelerometers, are readily available, and so are accurate and inexpensive controllers, e.g. FPGA or microprocessor based embedded controllers. Thus, the actuators, e.g. motors for the bridle 134 and dynamic surfaces, the controlling surfaces themselves and any ballast pumps, are important components of investment and operational costs.

FIG. 4 schematically illustrates a cable with all the functions of the embodiments in FIGS. 1-3. It is understood that a practical cable 32 or 132 may include some or all of the components 320-324 shown in FIG. 4.

In particular, reference numeral 320 refers to a fairing 320 to minimise the effects of drag force L (FIG. 1) on the cable. Contrary to the main body 101, there is no need to maximize the volume of fairing 320, which accordingly is shown with a streamlined cross section.

The tension rope 321 transfers a towing force T (FIG. 1). Cable 32 on FIG. 2 is not used for towing, and therefore lacks the tension cable 321. All instances of the stabilising cable 132 need a rope 321, e.g. a synthetic fibre rope or a steel wire, for transmitting a towing or lifting force. A downsized version of the rope 321 may also be practical for stress relief in cables not used for towing, such as the control cable 32 in FIG. 2. The towing cable 321 may comprise any rope, e-g—a steel wire or a synthetic rope. As discussed briefly above, a steel wire stretches less under load than a typical synthetic rope, but may require additional floating elements. Also as noted above, a steel wire requires additional buoyancy elements to neutralise its weight. Such buoyancy elements can conveniently be disposed within the fairing 320.

A communication line 322 may transmit sensor data and responses or steering commands between the main body 101 and a controller on a surface vessel 3.

A fuel line 323 may convey fuel from the main body 101 to an engine-generator unit 4 as discussed with reference to FIG. 3. A fuel line 323 may also be present in the cable 32 in FIG. 1 or 2 if it is desirable to fuel the propulsion system of the vessel 3 with hydrocarbons from the main body 101.

An electrically conductive wire 324 can transmit electrical power from the vessel 3 FIGS. 1 and 2) or the generator 4 (FIG. 3). Electrical power supplied through the wire 324 is the preferred way of powering steering means such as thruster 103 on FIGS. 2 and 3, pumps for ballasting tanks and actuators for control surfaces, e.g. a rudder.

FIG. 5 illustrates fluid connections for a preferred embodiment of the main body 101. The main body 101 comprises a cylindrical tank 140 with an inner surface 141 and a cradle 150. A flexible membrane 115 divides the inner volume of the tank 140 into a first cargo compartment 110 and a second cargo compartment 120. The flexible membrane 115 is optional, and an example of a general movable sealing element. One important function of the membrane 115 is to ensure that the first cargo compartment 110 can be completely emptied by filling the second cargo compartment 120 with water. Another important function is to prevent gas from dissolving in water as further explained below. Hence, the flexible membrane 115 must be gas-tight. Any sealing element, e.g. a design using pistons, may replace the flexible membrane 115 provided that the sealing element is gas-tight and capable of emptying the first cargo compartment 110 by filling the second cargo compartment 120 with water.

In the context of the present invention, a “gas-tight” membrane or sealing element should be construed as having sufficiently low permeability to prevent undesired dissolving of gas in water. In addition to the examples above, a fluid or granular material floating on the water might provide a barrier between water and gas.

Permanent ballast 151 provides close to neutral buoyancy. That is, the permanent ballast 151 has a mass to make the transport mass, i.e. the mass of the main body 101 when fully loaded with cargo fluid, approximately equal to the displacement, i.e. the mass of water displaced by the main body 101. Ore is relatively inexpensive and has a relatively high density, and may thus be a suitable ballast material. The ballast tanks 152 are adapted to contain more or less water to equalize variations in buoyancy, e.g. caused by local variations in density of the ambient water.

A channel 121 through the tank wall into the first cargo compartment 110 is shown schematically at the top of the tank 140. A similar channel 122 through the tank wall into the second cargo compartment 120 is shown schematically at the bottom of the tank 140. Elements 510-541 belong to an external network, and are not part of the main body 101. Similar valves, pipes, connections etc. are required to seal off the cargo compartments 110 and 120 during transport. These valves etc. within the main body 101 are not shown for clarity.

The external network comprises a cargo line 510 with associated cargo valve 511 for supplying or receiving cargo fluid to the first cargo compartment 110 through the opening 121. Water inlet 520 with associated valve 521 is configured to let ambient water flow into the second cargo compartment 120 through opening 121 in the tank wall. Water outlet 530 is provided with a valve 531 and a pump 532 configured to pump water out of the second cargo compartment 120. Finally, a separate return cargo line 540 with associated valve 541 illustrates equipment to supply or receive a return fluid.

Assume first that the main body 101 has just arrived at a loading point. The first cargo compartment 110 still contains dry air from a return trip, and should be loaded with CNG. The air contains oxygen, and should not be mixed with CNG for safety reasons. Hence, all air should be expelled from the first cargo compartment 110 before it is loaded with CNG. Normally, the ambient pressure is greater than the internal pressure within the first cargo compartment 110, so the air is conveniently expelled by opening valve 521 until the entire tank volume is filled with water, i.e. such that the membrane 115 engages the upper tank wall and all air is expelled through the outlet 540 for return fluid at the top of tank 140.

As the membrane 115 closes the opening 121, water from compartment 120 cannot expel gas from line 540. Hence, in some instances it may be desirable to flush a supply line, e.g. with N₂, before using the line for another fluid.

When the cargo volume contains water rather than cargo fluid, the mass of the main body is greater than the transport mass defined above. For a large cargo volume, the added mass may be considerable. This is further discussed with reference to FIG. 10.

The next task is to fill the first cargo compartment 110 with CNG. As noted above, the external pressure is preferably greater than the internal pressure to provide a compressive force on the tank 140. Continuing the numerical example above, the task is to fill CNG at e.g. 240 bar into a volume filled with water at ambient pressure, e.g. 250 bar. This is done by pumping water out of the second cargo compartment 120 by means of pump 532 while the valve 511 is open and cargo supply line 510 is connected to a source for cargo fluid, in this example CNG.

When the first cargo compartment 110 occupies the entire tank volume and contains a cargo fluid at a predetermined internal pressure, here CNG at 240 bar, the main body 101 has a transport mass close to its displacement as explained above. In this state, the cargo fluid is transported, preferably at depths providing a compressive force on the main body, to an unloading point similar to the generic loading and unloading point shown in FIG. 5.

At the unloading point, water with ambient pressure, e.g. at 250 bar, enters compartment 120 through valve 521, and cargo fluid is expelled through line 510 to the surface.

The next task is to fill the first cargo compartment 110 with dry air for the return journey. As air has a different density than CNG, the air pressure must be 206 bar to achieve the transport mass for the return journey in this example. Similar to the situation at the loading point, water is pumped out of the second cargo compartment 120 by means of pump 232 while the return fluid, here dry air, is supplied through line 510 or 540 from the surface.

As the amount of gas dissolved in water is proportional to the ambient pressure, a pressure drop of 44 MPa as in the present example may release a considerable amount of gas. The released gas could cause problems, so the flexible membrane 115, and in general any sealing element 115, is gas-tight to prevent gas from dissolving in the water. A similar flexible membrane 115 may also be useful for liquid cargo fluids and/or liquid return fluids.

From the above, it should be clear that lines 510 and 540 can be different lines or combined in a variety of ways. For example, line 510 can receive a cargo fluid, e.g. CNG, in one interval and supply return cargo, e.g. dry air, N₂ or CO₂ in a second interval. Similarly, line 540 can be a simple outlet for return fluid such as dry air or a line to the surface. Furthermore, the transport can be symmetrical in the sense that the return fluid in one loop is the cargo fluid in an opposite loop using the same two loading and unloading points. For example, a loading point at a platform producing CNG can be the unloading point for CO₂, e.g. for pressure support of the CNG-producing geologic formation.

FIG. 6 is a perspective view of a segment of the main body 101. The main body 101 comprises several such segments mounted end to end, and is closed at each end by a rounded end segment as shown in FIGS. 1-3. In a preferred embodiment, the tank 140 is made of precast rings with a wall thickness ΔD, and the walls and bottom 153 of the cradle 150 is made of precast concrete elements, e.g. such as the precast elements used for floors in a building.

If friction drag contributes more to the total drag R (FIG. 1) than the pressure drag related to the cross sectional area as indicated above, a small length-to-diameter ratio of the main body would be preferable to minimise skin surface and the total drag. The smallest length-to-diameter ratio of 1 would be achieved by a sphere. However, a cylinder is easier to construct and has the advantage that several rings can be cast in parallel and assembled to a tank. In general, the ratio of length-to-diameter depends on the application, and must be decided by the skilled person knowing the application at hand.

The invention does not exclude a double skin or sandwich structure, e.g. two concentric shells of steel with reinforcing, radial ribs and/or a concrete fill between them. However, a tank segment made of concrete has substantially lower manufacturing costs, and concrete is therefore used as much as possible in a preferred embodiment.

If the cylinder is made of concrete, the wall thickness ΔD has a minimum value required for strength that depends on the particular type of concrete in the wall, for example about 1 m for fibre reinforced ultra-high performance concrete (UHPC). In order to provide permanent ballast, the walls may intentionally be made thicker than this minimum value. Concrete walls are brittle in the sense that they break more easily when exposed to shear or tensile stress than when exposed to compressive stress. Thus, a main body made of concrete would preferably be operated at a depth where the external water pressure is greater than the internal pressure, e.g. from CNG. Furthermore, the brittleness can be counteracted by known means. For example, the entire main body 101 can be made of reinforced concrete on an assembly site. A structure in the order of 25 m in diameter and 350 m in length is well within the limits of conventional techniques. If a somewhat greater flexibility is desired, several segments such as the one shown in FIG. 6 can be joined with elastomeric gaskets in between or sealed permanently with grouted joints, and held together using conventional post tensioning techniques, e.g. with tensioned steel wire in tubing, which may later be filled with mortar or concrete for corrosion protection. In either case, the concrete can be reinforced by a conventional steel rebar, or by fibres of steel, polypropylene or any other material as known in the art.

The cradle 150 with ballasting elements 151 comprises largely permanent ballast with density larger than the ambient water. A high density material, e.g. magnetite or another ore, may be preferred to limit the ballast volume, cross section and skin surface of the main body and hence the drag R according to equation (1). This permanent ballast is provided at the lower part of the segment or main body 101 to lower the centre of gravity, and hence facilitate orientation of the main body 101, in particular at the loading and unloading points where it is important that the cradle 150 is below the tank 140 in order to support its weight.

Ballast tanks 152, e.g. in the form of commercially available steel pipes, extend along the outside of the tank 140. The ballast tanks 152 work in a conventional manner, and are essentially used to control buoyancy by pumping water in or out. For example, the pitch may be controlled using ballast tanks in the front and rear ends of the main body 101.

The ballast tanks 152 can also control roll. However, roll hardly matters during the transport and return journey, so no energy should be spent on pumping water for controlling roll, except perhaps at the loading and unloading points. A similar argument applies to thrusters (not shown) for controlling roll.

FIGS. 7a-c illustrates transport of a fluid with a main body 101 and equipment 521-541 similar to those on FIG. 5. In particular, FIGS. 7a-c illustrate the method referred to in connection with FIG. 5. As in FIG. 5, valves and other equipment within the main body 101 are not shown. Hence, no valves appear on FIGS. 7a and 7c, which illustrate transport, and the valves in FIG. 7b correspond to the external valves shown in FIG. 5. Seawater has typically a density about 1 025 kg/m³. For simplicity, a water density 1 000 kg/m³ is used in some of the examples. Other inaccuracies in the numbers are likely to be bigger than those caused by this approximation.

In particular, FIG. 7a illustrates transport of the main body 101 fully loaded with a cargo fluid 701 at a predetermined internal pressure. Thus, the first cargo compartment 110 occupies the entire tank volume. The numerical example with 150 000 m³ CNG with an approximate density of 226 kg/m³ at 0° C. and 240 bar at a depth of 2 500 m is as before. The mass of the CNG is then 33 900 tonnes, and a differential pressure of Δp=250−240=10 bar exerts a compressive force on the tank 140.

The arrows G_b illustrate the weight of permanent ballast and the tank walls, which do not change. The arrow G illustrates the weight of cargo fluid, here CNG. The upward arrow B represent buoyancy, and has a norm equal to the mass of water, e.g. seawater, displaced by the main body 101. As described above, the transport mass should be close to the displacement during transport.

FIG. 7b illustrates unloading CNG. This is done by opening valve 521 at the water inlet and valve 541 to a CNG line to the surface. As the external water pressure is 250 bar, the water 702 will displace the CNG. The longer black arrow indicates increased weight as water 702 has replaced cargo fluid 701. When the cargo fluid 701 is completely replaced with seawater 702 in the numerical example above, the mass resting at the unloading point is about 120 000 tonnes larger than the transport mass. The foundation at the unloading point must of course handle or support the additional weight.

FIG. 7c illustrates the ballasted return journey. The first cargo compartment 110 now contains return fluid 703, and the complete mass of the main body 101 is close to the transport mass. In the numerical example, the mass of return fluid 703 should be approximately 33 900 tonnes, i.e. the mass of CNG during transport. However, dry air at 0° C. has a higher density than the CNG, so a second internal pressure of 206 bar yields the desired mass for the return journey. Thus, the net differential pressure acting on the outer wall of the tank 140 during the return journey has increased to 44 bar. Of course, the walls of the main body 101 must withstand this pressure.

FIG. 8 a-c illustrates an alternative method wherein the compressive force on the main body 101 is equal during transport (FIG. 8a) and return (FIG. 8c). As above, the cargo fluid 701 is CNG at a predetermined internal pressure 240 bar. The return fluid is dry air as above, but the second internal pressure is at 240 bar rather than 206 bar as above. As in FIG. 7, the main body 101 and external valves 521-541 correspond to the example in FIG. 5. The arrows G_b, G and B correspond to those in FIGS. 7a-7c., and are not explained again.

FIG. 8a illustrates a transport where the first cargo compartment 110 contains cargo fluid 701 at the predetermined internal pressure, e.g. CNG at 240 bar, and the second cargo compartment 120 contains water 703. 150 000 m³ of dry air at 240 bar has a mass of 46 400 tonnes. Thus, in FIG. 8a the first cargo compartment 110 can contain 30 000 tonnes of CNG at 240 bar, and the second cargo compartment 120 can contain 16 400 tonnes of water. The external water pressure is 250 bar as before, yielding a net differential pressure at 10 bar working on the outer surface area of the tank during transport.

In FIG. 8b, the cargo fluid 701, e.g. CNG, is expelled through valve 541 as water 702 enters through valve 511 as explained in connection with FIG. 7b.

In FIG. 8c, the tank is fully loaded with return fluid 701. That is, the first cargo compartment 110 occupies the entire tank volume. The second internal pressure in FIG. 8c is by presupposition equal to the predetermined internal pressure in FIG. 8a. Thus, in the numerical example, the predetermined internal pressure and the second internal pressure are 240 bar, and the cargo has a mass 46 400 tonnes in both directions.

The method illustrated in FIGS. 8a-c limits the compressive force acting on the tank. However, about 11% of the tank volume in FIG. 8a is occupied by water 702, and thus unavailable for cargo fluid 701. Furthermore, a slight change in pitch might cause the water 702 to flow toward one end of the tank volume. Thus, the second cargo compartment 120 would require internal bulkheads or the like to maintain the distribution of water 702. In contrast, the method in FIGS. 7a-c only involves water in FIG. 7b, where the main body 101 rests on a horizontal fundament and no significant amount of water can collect in either end.

FIG. 9 shows a main body 101 with the origin of an imaginary Cartesian coordinate system at its centre of gravity. The x, y and z-axes are fixed with respect to the main body 101, and the terms roll, pitch and yaw are used in their usual meaning, i.e. as rotations about the x, y and z-axis, respectively. The main body can also be translated along these axes, e.g. by means of thrusters. The thruster 103 in FIGS. 2 and 3 provides an example of translation along the x-axis. The benefits of active controls, e.g. dynamic control surfaces and ballast tanks, are discussed above. Further details regarding submarine control and towing are known in the respective arts.

FIG. 10 illustrates a loading and unloading site with a main body 101 resting on a subsea platform 200 comprising buoyancy tanks and/or storage tanks. The platform 200 provides buoyancy and a large surface. In FIG. 10, the platform floats over an uneven seabed 4. Alternatively, the platform 200 can be placed on the seabed, and the large horizontal area of the platform 200 can reduce the ground pressure from the main body 101, i.e. distribute the weight of the main body 101 over a large area to enable a loading or unloading point on soft ground. Thus, the platform 200 can reduce or eliminate the need for a solid seabed and/or major soil improvement work on the seabed 4 to prepare a loading or unloading point.

Preferably, the subsea platform 200 is comprised of a plurality of rectangular tank elements 201, each having an inner cylinder similar to the main body 101.

The platform 200 in FIG. 10 has passive buoyancy control in the form of a plurality of heavy chains 210 hanging from the platform and resting partly on the seafloor 4. If the platform 200 starts rising, chains are lifted from the seafloor and adds weight. If the platform starts sinking, a larger part of the chains comes to rest on the seafloor 4, thereby reducing weight. In both cases, the length of the chains that rest on the seafloor 4 counteract undesired vertical motion.

The platform 200 is connected to the surface through a line 220. This line 220 represents the lines 510 and 540 described with reference to FIG. 5. Similarly, a connection unit 230 represents the valves, pumps and lines required for the loading and unloading explained with reference to FIGS. 5, 7 and 8.

As noted in connection with FIGS. 5, 7 and 8, the main body 101 has a near neutral buoyancy during transport, and an increased weight when filled with water to expel cargo fluid or return fluid. However, the water can be supplied from storage tanks 201 in the platform 200 and be pumped back into the storage tanks 201 in a closed loop. Thereby, the combined weight of the platform 200 and main body 101 remains unchanged during loading or unloading. Thus, when the cargo volume within the main body is filled with water, the platform 200 supports the extra weight, but provides no extra downward force, e.g. on the ground. As mentioned above, this extra weight may be considerable for a large cargo volume, e.g. 120 000 tonnes as provided in an example above. Handling ballast water in a closed loop has the added benefit that the buoyancy and/or flow of water into or out of the main body 101 does not have to be closely monitored during loading or unloading. A further benefit is that any contamination of the water, e.g. due to a rupture in the membrane 115 described above, is kept within the closed loop, and not released into the environment at the loading and unloading points.

For this, some or all of the tank elements 201 can contain cargo fluid, return fluid and/or water. The network 221 connects the tank elements 201 with the main body 101 and surface line 220 through a connection unit 230 comprising valves, pumps etc. explained above.

FIGS. 11a-c show an alternative embodiment of the main body 101 and the platform 202. In particular, the main body 101 is made of a concrete with higher density than the previous embodiment. The higher density can be achieved by adding a suitable material to the cement, e.g. iron ore. The increased density reduces or eliminates the need for external ballast 151, and hence the nose drag on the main body 101.

In the example on FIGS. 11a-c, the platform 200 has storage tanks 201 and means for landing a main body on the platform 200, e.g. a winch 30 and wire 31. The main body 101 is prevented from moving on the deck by guiding comes and the landing means 30, 31.

Under, the platform 200 is provided with skirts to enhance the contact with a “soft” seabed 4, e.g. loose seafloor sediments. After deployment, the platform 200 is allowed to settle in seafloor sediments on the seabed 4, and a cement or grout 40 is injected between the seafloor sediments and the bottom of the platform 200 to fill any empty spaces and provide a stable fundament. The purpose of the preferred method for loading and unloading is to keep the load 41 on the seabed 4 within predefined limits.

FIG. 11a shows a main body 101 loaded with cargo fluid 701, e.g. CNG. Storage tanks 20 in the platform 200 are filled with seawater 702. The tanks 20 are understood to be part of the tank elements 201 referred to above.

FIG. 11b illustrates pumping seawater 702 from the storage tanks 20 to the main body 101. This displaces cargo fluid 701, which is piped to a region above the water 702 in storage the tanks 20. Preferably, a membrane or other seal separate the cargo fluid 701 from the water 702 for reasons explained above. The pumping of water 702 continues until the water 702 fills the main body 101 and the cargo fluid 701 fills the storage tanks 20. There is no change in total mass of the main body 101, platform 200, cargo fluid 701 and water 702 during this exchange of cargo fluid 701 and water 702. Hence, the load 41 on the seabed 4 does not change due to this step.

FIG. 11c illustrates a second step involving pumping water 702 back into the storage tanks 20. This time, however, a return fluid 703, e.g. air, replaces the water within the main body 101. Using the water as a piston driven by suitable pumps, the cargo fluid 701 is expelled from the storage tanks 20 and conveyed to a reception facility on the surface.

Loading the main body 101 with a cargo fluid on another terminal may be performed in a similar manner, i.e. using the water 702 as a piston. To avoid potential hazardous mixtures of air 703 and cargo fluid 701, the main body 101 and storage tanks 20 should each be completely filled with water 702 before a new fluid 701, 703 is let into the space above the water 702.

While the invention has been explained by way of example, the scope of the invention is defined by the accompanying claims.

Claims

1-17. (canceled)

18. A subsea carrier for transporting a fluid comprising:

a main body for containing the fluid at a predetermined internal pressure, wherein the main body is designed to operate at a water depth where the external pressure substantially counteracts the internal pressure;

a stabilising cable; and

a floating element connected to the main body by the stabilising cable, wherein the stabilising cable comprises a first rope for transmitting force and is attached to a first connector that is movable with respect to the main body.

19. The subsea carrier according to claim 18, wherein the length of the stabilising cable exceeds the water depth.

20. The subsea carrier according to claim 18, further comprising a bridle for distributing a towing and lifting force.

21. The subsea carrier according to any claim 18, wherein the main body comprises a ballast element connected under a tank element.

22. The subsea carrier according to claim 18, wherein the main body comprises a first cargo compartment separated from a second cargo compartment by a movable gas tight sealing element.

23. The subsea carrier according to claim 22, wherein the gas tight sealing element is a flexible membrane.

24. The subsea carrier according to claim 18, further comprising at least one ballast tank for trimming and depth control.

25. The subsea carrier according to claim 18, wherein the main body comprises a control surface and/or a thruster for causing rotation of the main body about at least one of three mutually perpendicular axes of rotation.

26. The subsea carrier according to claim 18, wherein the main body comprises at least one thruster for causing translation of the main body along at least one of tree mutually perpendicular axes.

27. A system for transporting a fluid comprising a subsea carrier according to claim 24, further comprising:

a communication line for conveying control related signals between a controller at the surface and the main body; and

a power line supplying electrical power from an electric generator at the surface to the main body.

28. The system according to claim 27, further comprising a surface vessel with the controller and the electric generator, and a control cable with the communication line and the power line.

29. The system according to claim 28, wherein the control cable further comprises a second rope for transmitting force.

30. The system according to claim 28, wherein the control cable further comprises a fuel line from the main body to the surface vessel.

31. The system according to claim 28, wherein the floating element comprises the controller and the electric generator, and the stabilising cable comprises the communication line and the power line.

32. The system according to claim 28, wherein the stabilising cable further comprises a fuel line from the main body to the floating element.

33. A method for operating a subsea carrier according to claim 22, comprising the steps of:

filling the first cargo compartment with a cargo fluid at the predetermined internal pressure;

transporting the cargo fluid from a subsea loading point to a subsea unloading point;

expelling the cargo fluid from the first cargo compartment;

filling the first cargo compartment with a return fluid at a second internal pressure;

transporting the return fluid from the unloading point to the loading point, wherein expelling the cargo fluid involves allowing water to flow into the second cargo compartment, and filling the first cargo compartment with a return fluid involves pumping water out from the second cargo compartment.

34. The method according to claim 33, wherein the subsea loading point and/or the subsea unloading point comprises a subsea platform serving as a foundation for the main body during loading and/or unloading, wherein expelling the cargo fluid involves allowing the water to flow from a storage tank in the platform, and filling the first cargo compartment with a return fluid involves pumping the water back into the storage tank.