WO1997001480A1 - Improvements in, or relating to, the control of buoyancy underwater - Google Patents

Abstract

A device to provide buoyancy comprising a flexible envelope which can be attached to the object to be manipulated, a supply of gas sufficient to keep the envelope full at all times and a means whereby excess gas can escape.

Description

IMPROVEMENTSIN,ORRELATINGTO,THECONTROLOFBUOYANCY

UNDERWATER

Introduction

A heavy load underwater is often lifted by attaching a bag made of stout fabric to it and inflating the bag with air. The air displaces the water in the bag, increasing the buoyancy, and eventually the object rises to the surface. This method is simple and cheap, but it is also potentially

dangerous since the compressibility ofthe gas makes it virtually uncontrollable. As the bag rises, the pressure falls, so the gas inside the bag expands; this increases the buoyancy so the rate of ascent increases. Conversely, if the bag descends for any reason, the increasing pressure compresses the gas and the bag accelerates downward. As a result, it is impractical to use air-lift bags for placing heavy objects on the sea bed ("Emplacement").

This invention overcomes the inherent instability of air-lift bags, making them capable of

being externally controlled, and suitable for both lifting and emplacing heavy loads underwater.

The invention has two aspects which will be discussed separately in the interests of clarity, although they are interdependent in practice. Although the buoyancy medium is referred to as "air" in these descriptions, this term includes other gases, such as nitrogen.

First Aspect - Variable Volume

Fig 1 shows a conventional "parachute" air-lift bag, completely full of air, attached to a

load. Since the bag is completely full, any additional air will spill out ofthe open bottom, and will not increase the buoyancy; similarly, if the bag rises, and the air expands, the effective volume, and therefore the buoyancy, is unchanged. The first aspect of the invention will be discussed with reference to a bag of this type, but, as noted below, this is by way of example only, and the invention is not limited to the specific bag and mechanisms specifically described.

Fig 2 shows a similar air-bag to Fig 1, but modified by the addition of a cord, rope or

similar tension member attached to the inside ofthe top ofthe bag, and passing down through the open bottom to a small winch mounted above the load. As before, the bag is completely filled with air. Drawing in the cord by means ofthe winch will pull down the crown ofthe bag, as shown, partially tuming the bag inside-out. This will reduce its effective volume, as shown by the dotted line, and hence its buoyancy. However, if the bag ascends, and the air within it expands, the excess will again spill out ofthe bottom, and the lifting force will be unchanged. Similarly, if the bag is kept full of air, either by a continuous supply or by a demand valve (vide infra), the lifting force will not decrease significantly (apart from the minor change due to the alteration ofthe density of the air) as the bag descends. The buoyancy ofa bag modified in this way, and provided with a sufficient supply of air, is therefore independent of pressure, although it can be controlled by the winch. In Fig 2, the winch must draw in the crown ofthe bag to decrease the buoyancy, and it will therefore be working against a significant fraction ofthe total lifting force. In lifting heavy objects from the sea-bed, it is often observed that the lifting force required to pull the object free ofthe so-called "Bottom Suction" exceeds the in- water dead weight by as much as 30 - 50%. In such cases, it is important to be able to reduce the buoyancy rapidly to prevent the subsequent ascent being too fast. A very powerful winch would be required to pull in the central cord rapidly against the considerable lifting force on the crown ofthe bag, but the force can be reduced by the system shown in Fig 3. The skirt ofthe balloon is attached by short strops to a rigid framework which is attached in turn to a second cord. This cord and the cord from the crown ofthe bag shown in Fig 2 are wound in opposite directions around two separate drums on a common shaft, so that the torque due to the tension in one cord opposes the torque due to the tension in the other, and the net torque which must be overcome by the motor is greatly reduced. The motor itself is mounted on the stirrup which links the shaft bearings to the load, and the weight ofthe latter provides a torque reference. In practice, the system is slightly modified. In the preferred arrangement, there is a short tube in the centre ofthe rigid framework with two attachment points welded onto its lower end. The cord to the crown ofthe bag passes through this tube; the cord to the skirt is replaced by two cords which are attached to the points on the bottom ofthe tube and wound onto two identical drums either side of that for the cord to the crown. All three cords - webbing straps are the preferred form - pass through a pair of guide rollers between the drums and the attachment to the balloon. This guide is necessary to define the geometry ofthe drums, which is discussed below.

The torques due to the tensions in the cords are proportional to the tensions themselves and to the effective radius ofthe respective drums. The tensions in the cords vary with the distance the crown is pulled down, so it is necessary to change the relative radii ofthe drums to achieve a balance at every point. The variations in the two tensions depend on the shape ofthe bag and the way it folds as the crown is pulled down, so it is not easy to calculate. However, the forces can be readily measured by suspending the bag from a suitable load-cell, filling it with water, and then raising the crown via a second load-cell. Given this information and the guides referred to above to define the line of action ofthe tension forces, the shapes ofthe two drums required to balance the torques can be calculated by standard methods. Preferably, the drums are large enough to cover the complete control range in slightly less than a single turn. Since rotation ofthe drums releases one cord and draws in the other, and both of these actions have the same effect on the buoyancy, the drums do not need to be too large. A multi-tuπi mechanism would require a difficult and costly machining operation to produce drums with spiral grooves of the required variable depths.

The torque-balance system described above greatly reduces the power required to adjust the buoyancy ofthe bag. The residual control torque can be provided in several ways. An electric motor supplied from the surface is feasible but the losses in the long cables are a disadvantage. This could be overcome by powering the motor by means of on-board batteries so that only low power command signals need be transmitted from the surface. The most preferred option is to provide the motive power from the compressed air supplied to maintain the bag volume and control the position electrically. Acoustic control can be used, but since an umbilical connection is essential to carry the air, it is usually simplest and cheapest to transmit control signals through cables, which can also be used to carry information back to the surface.

The cords to the crown and skirt ofthe bag work essentially in opposition:- releasing the cord to the crown increases the buoyancy, but releasing the cord to the skirt decreases it. If the two cords are longer than necessary, and the excess wound around two separate winch drums controlled by brakes, the latter can provide full buoyancy control without any motive power whatsoever. Preferably, the brakes are operated by a common lever so arranged that an upward pressure releases the central cord while holding the skirt fixed, while a downward pressure has the opposite effect. In this form, the unit will be very easily operated by a free-swimming diver, since it will effectively follow him as he ascends or descends. Such a device will not be capable of complex manoeuvres, since the number of changes of buoyancy will be limited by the length ofthe cords, but, providing the bag itself is not too large it will be able to carry out simple lifts and emplacements. Adding a tank of compressed air and a simple control valve will make the unit completely independent ofthe surface, and should allow a diver to manipulate objects of at least 200kg without any external assistance. The forces on the crown and skirt of an open-bottom air lift bag, and the way they vary

as the crown is drawn down, will depend upon the size and shape ofthe bag. Conventional air-lift bags are designed to minimise both the area of fabric and the stresses within it, and the mechanisms discussed above will deal with the forces in such bags. However, the basic principles of the first aspect of the invention, namely keeping the bag full of air and varying its volume mechanically, can be applied very widely. For example, Fig 4 shows an air lift bag made up ofa bellows section at the bottom surmounted by a hemispherical top. Suitable strops are attached to the envelope at the junction of these two sections to support the load:- as shown, they can be inside the bag or outside, in which case a rigid annular spreader is needed. In this design, virtually all the load is supported by the hemispherical upper part; raising and lowering the skirt to change the volume changes the pressure on this upper part. The fabric in the lower part ofthe bag does not support the load directly, so the only stress upon it results from the difference between internal and external pressures. This stress is largely tangential so relatively small forces are needed to change the volume ofthe bag. On the other hand, such a bag is considerably more difficult to make than the conventional type. There are clearly many possible variations within the general principles

ofthe invention.

Second Aspect - Constant Volume

The variable volume air lift bag discussed above relies on keeping the bag full of air at all times. The air can be supplied continuously but this is very wasteful. The second aspect of the invention addresses this problem and extends the invention to constant volume air bags.

In a parachute-type air bag, the internal pressure is equal to the external pressure at the

air/water interface, which will be at the mouth ofthe bag if the latter is full. Demand valves which exactly match internal to external pressure are well known in prior art:- the two-stage demand valve used in SCUBA (Self-Contained Underwater Breathing Apparatus) is an example. If such

a valve is fixed at the mouth of the bag and senses the pressure in the latter through a tube -

preferably this is closed by a flexible diaphragm to prevent the ingress of water - it will admit air

only when the air/water interface rises, i.e. the bag is less than full. A short tubular extension to the mouth of the bag, as shown in Fig 5, will improve the reliability. If the air/water interface descends due to the bag rising, the demand valve will close and the excess air will emerge from the open bottom as usual.

It will be clear to those skilled in the art that the maximum flow rate required of such a valve will be determined by the size ofthe bag and the maximum rate ofthe descent envisaged. If the rate of descent is too high, so that the air supplied by the valve cannot maintain the volume against the increasing pressure, the buoyancy will fall, and the bag will descend out of control. It is therefore essential to use a valve capable of passing a sufficient flow.

Heavy loads, such as wrecks, are often partly supported by so-called "constant volume" air-lift bags while they are lifted by another method. Unlike the air-bags previous discussed, these are totally enclosed, with no open bottom; however, they are equipped with an over-pressure valve. When used to lift a wreck, for example, these bags are attached to the load on the bottom, and fully inflated. As the load is lifted, for example by a crane, the expanding air escapes through

the over-pressure valve.

Conventional constant volume air bags cannot be used for emplacement, since there is no

means of replacing the lost air, indeed, if a wreck lifted with their help is forced down in the water by as little as 0.5m, the resultant decrease in buoyancy may be enough to sink it again. This limitation can be overcome by supplying air from the surface through a demand valve similar to

those already discussed. Preferably, the demand and over-pressure valves should be arranged, by means which are well known in prior art, to maintain the intemal pressure in the bag slightly higher than the external pressure. This will keep the buoyancy of the bag constant, regardless of the depth, providing that the demand valve can supply at the necessary pressure and flow-rate. A constant volume bag provided with an air supply and equipped with a demand valve represents a significant improvement on existing constant-volume air-lift bags, which can only be used for lifting, and falls within the present invention. However, the lack of buoyancy control is a

disadvantage.

The lift of a closed air bag can be changed by mechanically deforming as in the first aspect of this invention. Fig 6 is provided by way of example and shows one way in which this can be done, but the invention is not limited to this form. An electrically-operated winch is mounted inside the bag and draws opposite ends together; clearly other geometrical arrangements could be devised on the same principle. As the bag is collapsed, air is driven out of the exhaust valve; conversely, as it is allowed to expand, the demand valve maintains the intemal pressure.

Submarines admit water to ballast tanks to decrease buoyancy and expel this water with compressed air to retum to the surface. This principle can be applied to constant volume air bags, but since the intemal pressure will be slightly above the extemal pressure, the water must be actively driven in. A pump on the surface may supply this water through a second hose, but it is preferable to pump it in directly. Since the difference between the intemal and extemal pressures is small, the choice of pump and prime mover is not critical. The preferred method is to operate a simple diaphragm pump, which can be purchased from many manufacturers, from the existing air supply. Since the pumping speed is relatively low, flow control is not required, and a simple on off electro-magnetic valve in either the air supply to the pump or the water outlet will suffice. Air is normally supplied at relatively high pressure, since this allows the diameter ofthe hose to be reduced. This pressure must be reduced for SCUBA and similar demand valves to operate effectively, so a first reduction stage is normally fitted. This first reduction stage can also supply the diaphragm pump, since the maximum supply pressure of these is limited.

The excess pressure inside the bag will expel the ballast water automatically when a valve

in the bottom ofthe bag is opened. Tbis valve must be externally controlled, and it must also be

relatively large to secure a good flow of water since the excess pressure inside the bag is fairly small. Although the means of operating the valve is not critical, the preferred method is pneumatic actuation from the air supply, controlled by an electro-magnetic valve.

Closed bags are not subject to the geometrical constraints of open-bottom bags; they are usually cyjiidrical, since this shape allows a larger volume with the same radius of curvature:- the latter determines the stress for a given internal pressure. These cylindrical bags are normally used in the horizontal position to minimise pressure differences between different parts ofthe envelope; nevertheless, their length is limited by the maximum permissible stress in the fabric and the maximum angular excursion from horizontal expected. If the bag is not exactly horizontal, ballast water will drain to the lower end, so to remove it completely, it is necessary to have a ballast dump valve at each end. If these are operated together by a single control, the upper valve will vent air before all the ballast water has been expelled. This can be avoided by operating the valves selectively, so that only the lower valve is opened, but the preferred method is to equip each valve with a buoyant flap, as shown in Fig 7. This is held clear ofthe inlet as long as there is water

around the valve, but when all the water has been jettisoned, the flap covers the inlet and prevents air escaping. The buoyancy ofthe flap must be provided by some pressure-resistant medium: - "Syntactic Foam", hollow glass spheres bonded together with synthetic resin, is an example.

If a constant volume lift bag modified in the way discussed tilts away from the horizontal

at all, the ballast water will drain to the lower end and accentuate the tilt, making the bag unstable in pitch. Suspending the load well below the bag or attaching two or more bags to a common beam will improve the stability, but this feature, coupled with the relatively slow decrease in buoyancy that can be achieved with the ballast pump, make this form of bag less suitable for recovering objects from the sea floor than the parachute type. On the other hand, the elongated shape and the possibility of multiple attachment points to a load is advantageous in several sub-sea

operations, one of which is discussed below, in order to illustrate various technical features ofthe

inventioa This is provided by way of example only; the invention is not limited to this particular application.

Apart from reeling narrow pipes from a specially designed ship, present methods of installing undersea pipelines can be divided into "Bundle Laying" and "Barge Laying", which is in turn sub-divided into "S-Laying" and "J-Laying". The former is used to instal several narrow pipes together in relatively short lengths and the latter to instal single wider pipes. Those skilled in the art will be familiar with these processes and will appreciate that a means of altering the buoyancy under external control will greatly improve both methods. In bundle laying, the carrier pipe and its chains would become redundant - although it might be retained in some circumstances purely to protect the bundle. A bundle can be temporarily parked on the bottom, for example during bad weather, some parts ofthe bundle can be made buoyant, while leaving others firmly on the bottom, during final installation- In barge laying, the ability to control the buoyancy down the length ofthe pipe enables the optimum catenary curve to be maintained without applying heavy tensile loads or having long and costly "Stingers". This in turn reduces the required wall thickness ofthe pipe and removes the need for the laying barge to kedge forward, as required in the existing system.

Fig 8 shows the main features ofa "Pipeline Buoyancy Module" (PBM), a constant volume lift bag according to the present invention and adapted for laying undersea pipelines. A series of these units is attached to the pipeline at suitable intervals by the grippers shown. Preferably, these grippers should be capable of being released remotely, by means described in prior art, so that the modules can be recovered for re-use once the pipe is in place. The units are linked by an umbilical carrying the air supply and the cable carrying the command signals; electrical power to operate the various control valves could be provided by a further cable, or, in the preferred system, by small batteries on each module. Since it may be necessary to address more than a hundred buoyancy units individually, optical fibre cables, which can carry very large amounts of infoπnation are the preferred method for transmitting commands; the same cable can also carry information on depth, etc. from each module back to the central controller on the surface. The modules are also linked by a strong rope attached to strong points on the bags in order to prevent tensile loads falling on the umbilical

In bundle laying, as each seαion ofthe bundle is completed, the appropriate number of PBMs are attached and the section is pushed out into the sea, where it is allowed to rest lightly on the bottom. When the bundle is completed, its buoyancy is slightly increased and it is towed into place in the normal way. Once in the correct position, the bundle is allowed to sink, and the ends towed into place by standard methods; the ability to make some parts ofthe bundle positively buoyant while the rest remains on the bottom will greatly assist this operation. Finally, the release units are operated and the string of PBMs rises to the surface where they can be collected for re¬ use.

In barge laying, PBMs are attached to each section as it is welded on and pushed over the stem ofthe laying barge. As laying progresses, each section descends slowly to the bottom and the buoyancy is adjusted to maintain the correct curve. Once a given section is firmly in place on the bottom, the PBMs are released. As before, they rise in a string to the surface where they can be collected and returned to the laying barge for re-use.

Although modem air-pumps allow air to be pumped down to the deepest parts of the ocean, the buoyancy of any air lift bag will decrease due to the increasing density of air. This can be compensated for by having the bag virtually full of ballast and dumping it as the depth increases. This will allow lifting bags made according to the invention to be used to depths of at least 1km.

Claims

Claims

1. A device to provide buoyancy comprising a flexible envelope which can be attached to the object to be manipulated, a supply of gas sufficient to keep the envelope full at all times and a means whereby excess gas can escape.

2. A device as in Claim 1 in which the gas is supplied through a pipe from the surface.

3. A device as in Claim 1 in which the gas is supplied from tanks attached to the envelope.

4. A device as in Claim 1 in which the excess gas can escape through the open bottom ofthe envelope.

5. A device as in Claim 1 in which the envelope is closed and the excess gas escapes through one or more over-pressure valves.

6. A device as in Claim 1 in which the gas supply is regulated by a demand valve so that the pressure in the envelope is always equal to, or slightiy above, the local pressure.

7. A device as in Claim 1 which includes mechanical means to deform the inflated envelope in order to vary its effective volume.

8. A device as in Claim 7 in which the actuator to deform the inflated envelope is powered electrically.

9. A device as in Claim 7 in which the actuator to deform the inflated envelope is powered pneumatically.

10. A device as in Claim 7 in which the force to deform the inflated envelope is wholly or partly derived from the lifting force ofthe device itself.

11. A device as in Claim 10 in which lines are attached to two different parts ofthe envelope and wound onto drums.

12. A device as in Claim 11 in which the drums are attached to a common shaft and the lines are wound onto them in opposite directions.

13. A device as in Claim 12 in which the radii ofthe drums are varied in order to wholly or partly balance the torques due to the lines.

14. A device as in Claim 11 in which the drums are controlled by brakes.

15. A device as in Claim 14 in which the brakes are controlled by a common lever which releases one ofthe brakes depending on the direction it is moved.

16. A device as in Claim 5 in which buoyancy can be reduced by injecting water into the

inflated envelope.

17. A device as in Claim 16 in which the water is supplied from a pipe to the surface.

18. A device as in Claim 16 in which the water is pumped in from the surroundings.

19. A device as in Claim 16 in which the pump is electrically operated.

20. A device as in Claim 16 in which the pump is pneumatically operated.

21. A device as in Claim 16 in which the envelope is provided with valves which can be opened remotely to allow the water in the envelope to be expelled.

22. A device as in Claim 16 in which the valves are closed automatically when all the water is expelled.

23. A device as in Claim 16 in which remote operation ofthe valves is achieved by electrical means.

24. A device as in Claim 16 in which remote operation ofthe valves is achieved by pneumatic means.

25. A device as in any preceding Claim in which the attachment ofthe envelope to the load can be remotely released.

26. A device as in any preceding Claim in which command signals from the surface are transmitted through a cable.

27. A device as in any preceding Claim in which command signals from the surface are

transmitted acoustically.

28. A device as in any preceding Claim in which the buoyancy is controlled by an on-board controller which produces command signals according to information received from local sensors.

29. A system for dealing with elongated loads, such as a pipe, in which a series of devices according to any preceding claim are attached to the load and controlled from the surface.

30. A system as in Claim 26 in which the devices are supplied with air from a common pipe.

31. A system as in Claim 26 in which the command signals are transmitted through cables.

32. A system as in Claim 26 in which the command signals are transmitted acoustically.