WO2006079931A1

WO2006079931A1 - Vessel flushing system

Info

Publication number: WO2006079931A1
Application number: PCT/IB2006/000685
Authority: WO
Inventors: Davoud Tayebi
Original assignee: Aker Kvaerner Process Systems A.S
Priority date: 2005-01-25
Filing date: 2006-01-18
Publication date: 2006-08-03
Also published as: GB2437231A; GB0501442D0; GB0716071D0; NO20074298L; AU2006208968A1

Abstract

A flushing system for an accumulator vessel has a plurality of nozzles (16a, 16b) disposed at intervals around a perimeter (18) of the accumulator vessel and adjacent a base (10) of the vessel. The nozzles (16a, 16b) are arranged to direct a fluid supplied to the nozzles to create a rotational flow around the base (10) of the accumulator vessel so as to fluidise solid particles in the base (10) of the vessel.

Description

Vessel Flushing System

The present invention relates to an apparatus and method for flushing solids in a vessel such as an accumulator vessel or flush tank.

Accumulator vessels may be used for collection of solids, for example, in process plant where solids are separated from fluids. Examples include: a flushing system used in a hydrocarbon or produced water desander assembly; a sand washing system; solids cleaning as used in oil and gas well production and processes for recovering a hydrate inhibitor such as monoethylene glycol (MEG). During production, transport and processing of well fluids it is necessary to remove solids (e.g. sand) carried by the well-stream fluids to avoid blockage or damage to plant and equipment. Another similar operation where solids may be separated and removed is a solids cleaning system.

In these solids removal systems, solid particles (e.g. sand) enter the accumulator vessel where they accumulate. The accumulated solids need to be removed from the vessel, either continuously or periodically in a batch emptying procedure, through an opening, which is usually located at the bottom of the vessel. There are several problems associated with the emptying operation. In many applications it is very difficult to empty a vessel without fluidizing the solids so that they flow through the outlet. Designing an effective fluidized bed can be a challenging issue in itself. Another problem is that accumulated solids may stick to the vessel walls and remain inside the vessel, rather than flowing out as the vessel is emptied. In these circumstances the effective volume of the accumulator vessel is reduced. A further problem with these systems is that the accumulated solids in the bottom of the vessel tend to clog the opening, thus preventing effective emptying.

It is an object of the present invention to provide a flushing system to alleviate these problems.

According to a first aspect of the present invention there is provided a flushing system for an accumulator vessel comprising a plurality of nozzles disposed at intervals around a perimeter of the accumulator vessel and adjacent a base of the vessel, wherein the nozzles are arranged to direct a fluid supplied to the nozzle to create a rotational flow around the base of the accumulator vessel so as to fluidise solid particles in the base of the vessel.

In embodiments of the system, an outlet for removal of accumulated solids is provided in the base of the accumulator vessel.

According to a second aspect of the present invention there is provided a method of emptying solids from an accumulator vessel, the method comprising: directing a flow of a fluid through a plurality of nozzles into the base of the accumulator vessel so as to create a rotational flow around the base of the accumulator vessel, said rotational flow fluidising solid particles in the base of the accumulator vessel; and opening an outlet in the base of the accumulator vessel to allow the solid particles to be emptied.

In a preferred embodiment the outlet is arranged for the solids to flow downwards out of the vessel. Alternatively, the outlet may be arranged to provide an upward outflow of solids.

It has been found that by directing the flow of a fluid through nozzles that are positioned to produce a rotational flow around the base of the vessel, the solids can be fluidised locally in a layer at the base of the vessel. The fluidising effect forms a localised slurry of particles, which are continuously moved within the fluidised layer so that the solids flow freely out of the vessel as it is emptied.

In known fluidised bed systems, the entire bed of particles is fluidised so as to be lifted up from the base. A certain minimum fluid flow rate is required to fluidise the solid particles and this has to be provided across substantially the entire cross-section of the vessel in order to lift the bed. This means that a relatively large fluid flow rate is required. In embodiments of the present invention, effective emptying of solids can be achieved without fluidising or lifting the entire bed. As long as the fluid velocities are high enough to fluidise a layer of solids at the base of the vessel, highly effective emptying can be achieved.

In one embodiment, the vessel is an upright cylindrical vessel having a flat base or a semi-flat base (i.e. a base having a flat portion). Preferably, the nozzles are mounted in the cylindrical side-wall a short distance (for example less than one fifth of a vessel diameter) above the base. More preferably, the nozzles are equi-spaced around the side-wall.

The nozzles are preferably directed downwards towards the base at an angle to the vertical. The angle is preferably less than 45 degrees and more preferably is about 30 degrees. The nozzles should also be directed at an angle to the radius of the vessel so as to set up the rotational flow. A preferred angle to the radius is 60 degrees.

The nozzles may be flat spray type nozzles. The flat spray type nozzles may provide a spray over an angle of about 100 degrees.

Further nozzles may be provided in the base. These nozzles preferably produce a spray in the form of a flat disc.

The arrangement of nozzles advantageously provides a fluidising flow covering a substantial portion of the area of the base of the vessel, and may in fact cover the entire base area of the vessel.

In an alternative embodiment, the upright cylindrical vessel has a conical section at the base having a cone wall. Preferably, depending on the size of the vessel and diameter of the cone some of the nozzles are mounted in the cylindrical side-wall a short distance (for example less than one fifth of a vessel diameter) above the top of the conical section. More preferably, the nozzles are equi-spaced around the side- wall. Further nozzles may be provided in the cone wall. These further nozzles are preferably also flat spray type nozzles. The further nozzles are preferably directed downwards at an angle to the cone wall and at an angle to the radius of the vessel.

The method of emptying the vessel may comprise supplying a fluid under pressure to any combination of nozzles. In addition fluid may be provided under pressure through the outlet of the vessel prior to emptying. This provides the advantage of preventing or breaking down the bridging of solid particles above the outlet and is particularly suitable for use on a vessel having a conical section base.

In another embodiment the vessel has a low-profile domed base. Alternatively the base of the vessel may be "semi-flat" having a flat central portion and a conical outer portion. The nozzles are preferably arranged in a similar manner to that preferred for a flat base. Alternatively the base may be a hemispherical domed end. This has many of the advantages of a conical section base in assisting the flow of solids during emptying, but has a considerably larger volume for a given vessel height, and is less prone to blockage due to bridging.

An embodiment of the invention will now be described with reference to the accompanying drawings, in which:

Figure IA is a cross-sectional elevation of the bottom of an accumulator vessel, showing flow directions of fluid supplied to nozzles;

Figure IB is a plan view of the bottom of the accumulator vessel of figure IA, showing a rotational flow pattern;

Figures 2A and 2B are plan views similar to that of Figure IB, but showing a flow coverage area of a single nozzle (Figure 2A) and overlapping coverage areas of three nozzles (Figure 2B);

Figure 3 is a sectional elevation of a model of an accumulator vessel used for experimental tests; Figure 4 is a plan view of the accumulator vessel of Figure 3, showing flow coverage areas for selected nozzles;

Figure 5 is a sectional elevation of the accumulator vessel model of Figure 3 having a conical base section; and

Figures 6A and 6B show two alternative arrangements for the outlet of an accumulator vessel.

Referring to Figure IA, the bottom part 10 of an accumulator vessel has a domed end 12 with a central, downwardly oriented outlet 14 for emptying the vessel. Nozzles 16a, 16b protrude through a peripheral wall 18 of the bottom part 10 and are directed downwards at an angle (in this case of about 30 degrees) to the vertical, so that a flow of fluid supplied to the nozzles is directed downwards into the domed end 12, as shown by the arrows 20a.

Referring to Figure IB, three nozzles 16a, 16b, 16c are shown equi-spaced at angles of 120 degrees around the peripheral wall 18. The nozzles are not only directed at a downward angle as shown in Figure IA, but, as shown by the arrows 20b, are arranged to set up a rotational flow around the vessel when fluid is supplied simultaneously to all three nozzles 16a, 16b, 16c.

Referring to Figure 2A, a coverage area 22b is shown for one nozzle 16b. The nozzle 16b is shaped to provide a flat spray pattern, radiating outwardly over a certain angle (in this case of about 100 degrees). The flat spray pattern of the nozzle 16b has a central spray direction 24b, which is directed at an angle to the radius of the vessel. This results in the coverage area 22. The coverage area is the area, in plan view, of fluidisation of solid particles in the vessel bottom 10.

Referring to Figure 2B, coverage areas 22a, 22b, 22c are shown for the three nozzles 16a, 16b, 16c. The combined coverage of the three nozzles covers the entire cross- section area of the vessel bottom 10. There are three areas of overlap, area 26ab of nozzle coverage areas 22a and 22b, area 26bc of nozzle coverage areas 22b and 22c, and area 26ac of nozzle coverage areas 22a and 22c. The momentum of the flow from nozzle 16b is centred on the central direction 24b, which is at an angle of about 60 degrees to the radius of the vessel. Due to the symmetry of the arrangement the momentum from each of the other nozzles 16a, 16c is at a similar angle, and the combined effect is to set up a rotational flow around the vessel (as indicated by the arrows 20b in Figure Ib). Although it may be advantageous for the flow coverage areas 22a, 22b, 22c, to cover the entire cross-section, this is not essential to achieve the rotational flow pattern. It will also be appreciated that a rotational flow pattern may be achieved with two nozzles or with more than three nozzles, that it is not essential for the nozzles to be equi-spaced, and that the angles at which the nozzles are directed may be varied to suit the particular vessel geometry. However, the three nozzle equispaced arrangement has been found to be particularly effective.

When the accumulator vessel is full of sand and fluid is supplied to the three nozzles 16a, 16b, 16c, the sand becomes fluidised to form a sand/fluid slurry in the coverage areas. Fluidisation of a bed of solid particles occurs when the fluid velocity exceed a certain minimum velocity (determined by the properties of the fluid and solid particles). In the arrangement shown in Figures IA, IB, 2A and 2B, the solid particles become fluidised within the coverage areas in the bottom 10 of the vessel. The flow of the fluidised slurry is rotational around the vessel. As the following experimental results testify, this rotational fluidising flow is particularly effective in flushing the solid particles when the solids are being emptied through the central outlet 14.

When supplying fluid to the nozzles 16a, 16b, 16c, the vessel may be emptied at a continuous steady rate. Alternatively, the vessel may be emptied in a batch emptying operation. It may be advantageous for the fluidising flow to be turned off when the vessel is being emptied. The absence of fluidising flow during the emptying operation need not be detrimental, provided that the solids have not had time to settle before the outlet 14 is opened and the emptying outflow commenced. Figure 3 shows a lower portion of a model of a cylindrical accumulator vessel 30 used for experiments with sand and water. The accumulator vessel 30 has a flat end cup 32 at its base. The inner diameter of the vessel 30 is 500 mm while its total height is 1000 mm. Three flat spray sidewall nozzles 34a, 34b, 34c (not shown) are fixed in a sidewall 35 around the lower part of the accumulator vessel 30, and three further nozzles 36a, 36b (not shown), 36c are fixed in a bottom end wall 38. A central outlet 40 is provided for emptying the vessel and a valve 42 allows the outlet 40 to be opened and closed. A separate side entry 44 with a valve 46 is also provided to enable water under pressure to be supplied to the vessel through the outlet 40.

The flat spray sidewall nozzles 34a, 34b, 34c are disposed around the sidewall 35 with a 120° displacement with respect to each other. The further nozzles 36a, 36b, 36c are disposed in the bottom end wall 38 with a displacement of 120° with respect to each other. The disposition of the further nozzles is rotated 45° with respect to the sidewall nozzles. The sidewall nozzles 34a, 34b, 34c are mounted "almost tangentially" (i.e. so that they only protrude a very short distance into the vessel 30) and pointing down towards the bottom end wall 38. As indicated in Figures 3 and 4 the flushing coverage area for each nozzle is mainly in front of it along the bottom end wall 38 and the sidewall 35. Using this configuration provides a rotational flow of flushing water in the core region of the vessel. In addition the sidewall 35 is subject to continuous washing.

The further nozzles 36a, 36b, 36c are full spray deflector type nozzles creating a flat spray disc just above the bottom end wall 38. Each further nozzle has a deflector cap 37 having the two main functions of creating a flat spray disc and preventing sand particles from entering and blocking the nozzle.

To achieve optimal coverage and fluidisation effect the sidewall nozzles 34a, 34b, 34c are angled downwards at about 30 degrees with respect to the vertical sidewall 35. The flat spray sidewall nozzles 34a, 34b, 34c used in this study have a spray angle of 102°. Figure 4 illustrates the coverage area 48 of a flat spray sidewall nozzle 34b and the coverage area 49 a full spray deflector type further nozzle 36a. Figure 5 shows an alternative design of a model of an accumulator vessel 50 for use in the experiments having a conical section end cup 52, which has a cone wall 57. The inner diameter of the vessel 50 is 500 mm while its height is 1000 mm. The angle of the cone is 45°. Three flat spray sidewall nozzles 54a, 54b, 54c (not shown) are fixed in a sidewall 55 around the lower part of the cylinder and 100 mm above the top of the conical section 52. Three cone nozzles 56a, 56b, 56c (not shown), also of the flat spray type, are disposed around the middle region of the conical section 52. The same outlet arrangement is provided for as for the flat end cup vessel 30 of Figure 3 - including the central outlet 40, valve 42, side entry 44, and side entry valve 46.

The sidewall nozzles 54a, 54b, 54c are disposed at 120° with respect to each other around the sidewall 55. The further flat spray nozzles 56a, 56b, 56c are also disposed at 120° with respect to each other and rotated 45° with respect to the sidewall nozzles. Both the sidewall nozzles 54a, 54b, 54c and the cone nozzles 56a, 56b, 56c are mounted "almost tangential" to the sidewall 55 and cone wall 57 and pointing downward and to one side. The configuration of nozzles results in an effective rotational flow around the base of the vessel. In addition the sidewall 55 is subjected to continuous washing and lifting up of adjacent solid particles. The flushing coverage area for each nozzle is mainly in front of it along the cone wall 57 and the sidewall 55.

The cone nozzles 56a, 56b, 56c are fixed with a 90° bend at the cone wall and are directed downwards at 60° with respect to the vertical axis of the tank. The sidewall nozzles 54a, 54b, 54c are angled downwards at 30° with respect to the sidewall 55. All the flat spray nozzles have a spray angle of 102°.

In the accumulator vessel configurations described above in relation to Figures 1 to 5, the outlet 40 is shown disposed below the vessel base for downward flow of solids during the emptying procedure. In Figure 6A an alternative outlet arrangement is shown with an outlet pipe 60a aligned axially within the vessel 10 and having an opening 62a located a short distance above the centre of the vessel base. By providing a pressure difference between the vessel 10 and the outlet of pipe 60a the solids are removed by means of an upward flow. A similar arrangement is shown in Figure 6B, but here the outlet pipe 60b has a bend 64 so that it passes through a side wall of the vessel 10 from where the solids are removed.

To examine the performance and efficiency of various nozzle configurations and accumulator design several tests were carried out with both the flat bottom end and conical section vessels under the same, or similar experimental conditions.

FLAT END CUP Case 1

The experiment was carried out with 100 μm sand particles in the vessel 30 and fresh water supplied to all three flat spray nozzles 34a, 34b, 34c and all three further nozzles 36a, 36b, 36c. The pressure of the water supplied to all the nozzles was kept constant at 3 bar. The initial height of the sand bed was 35 cm. The sand was packed in 2 hours before the start of fluidisation. The whole bed was fluidised after 5 minutes. Sand concentration out of the tank at the start of the emptying process was measured at 48 vol%. Sand concentration in the slurry after 10 minutes was 13 vol% and after 20 minutes about 1 vol%. The level in the vessel was kept constant. This means that the volume of flush water into the vessel was equal to the volume of sand slurry out of the vessel. The concentration of the sand in the slurry at the end of the experiment was about 1 vol%. The whole content of the tank was fluidised. The vessel 30 was successfully emptied.

Case 2

The accumulator vessel 30 was filled with 20 cm of 100 μm sand, and water to a height of 80 cm. Both the sidewall nozzles 34a, 34b, 34c and the further nozzles 36a, 36b, 36c were used. The bed had been at rest for 30 minutes fore the start of the fluidisation process. The water pressure behind all the nozzles was kept constant at 3 bar. The content of the bed was mostly fluidised after 10 minutes, but it was first fully fluidised after 20 minutes. The fluidised bed height at 3 bar after 20 minutes was 95 cm. There was a quiet zone with clear water at the top of the bed. The height of this zone was about 10 cm. Case 3

The accumulator vessel 30 was filled with 20 cm of 100 μm sand, and water to a height of 80 cm. Only the sidewall nozzles 34a, 34b, 34c were used. The bed had been at rest for 30 minutes before the start of fluidisation. The water pressure behind all nozzles was kept constant at 4 bar. The content of the bed was mostly fluidised after 10 minutes. The fluidised bed height at after 10 minutes was 50 cm. There was a quiet zone with clear water at the top of the bed. The height of this zone was about 40 cm. The bed was almost fully fluidised after 20 minutes. The fluidised bed height after 20 minutes was 57 cm. Whole bed content was moving except a small region at the bottom centre. The vessel 30 was emptied in 6 minutes. There were no particles left at the bottom or inside the vessel after emptying. It was an effective and good sand removal operation.

Case 4

The accumulator vessel 30 was filled with 35 cm of 100 μm sand, and water to a height of 80 cm. Both the sidewall nozzles 34a, 34b, 34c and the further nozzles 36a, 36b, 36c were used. The bed had been at rest for 30 minutes before the start of fluidisation. Then, all the nozzles were closed. The water above the sand bed was removed from the top of the tank. Two litres of Nome oil was added to the bed. The bed was then fluidised again using the bottom and the side nozzles. To mix the oil into the sand/water slurry a pump was used. The oil was sucked from the top layer through a recycle tube and then injected to the bed at the bottom of the tank. The recycle tube was moved around the bed to achieve a more efficient mixing. This process was continued for 45 minutes. After mixing the oil into the sand/water slurry the mixture was left to rest for 10 days. During this time the oil/water height in the tank was 95 cm while the height of the wet oily sand bed was 35 cm. All the nozzles were closed and the water flow behind them was shut down.

After 10 days the bed was then fluidised using both the sidewall nozzles 34a, 34b, 34c and the further nozzles 36a, 36b, 36c. The water pressure behind all the nozzles was kept constant at 3.7 bar during the test. The bed started to fluidise after about one minute and it was fully fluidised after about 5 minutes. The bed height after 5 minutes was 50 cm. The boundary between the fluidised region with sand particles and clear water above it was very clear. The interface between the fluidised bed of particles and the water above it was also quite clear with a wavy form interface.

The vessel 30 was emptied 10 minutes after the start of fluidisation. During the emptying process all the nozzles were operating at reduced capacity. The pressure behind the further nozzles 36a, 36b, 36c was 2.5 bar and the pressure behind the sidewall nozzles 34a, 34b, 34c was 3.5 bar. During the last 2 minutes of the emptying process the pressure was further reduced to approximately 0.5 and 2.5 bar respectively. When the slurry level was about 10 cm above the bottom the water flow to all nozzles was closed down. The vessel 30 was successfully emptied with both the bottom end and the sidewalls free of sand.

At the start of fluidisation none of the nozzles were blocked by sand and all were fully operative. The experiment was carried out without the vessel 35 or its content having been moved.

One conclusion from the observations during these experiments is that it is possible to ensure effective emptying using only the sidewall nozzles 34a, 34b, 34c.

CONICAL END CUP

Case 5

The accumulator vessel 50 was filled with lOOμm dry sand. The water height was 70 cm. It was left to settle and rest for two days. AU six nozzles 54a, 54b, 54c, 56a, 56b, 56c were used. The water pressure behind all the nozzles was kept constant at 3.7bar. The bed was fluidised after 4 minutes with a height of 60 cm. The whole bed was fluidised and everything was moving. The fluidised bed height after 10 minutes was 85 cm.

After fluidisation the bed was emptied. The vessel 50 was fully emptied of both sand and water after about 7 minutes. However, at the start of the emptying process the sand/water mixture passing through the outlet 40 was very dense resulting in a blockage of the outlet. The reason was that most of the large sand particles accumulated just above the outlet at the bottom of the cone because it is difficult for fluidisation water to reach this part of the cone. However, the particles in this region can be fluidised from below using pressurized water through the outlet 40 just before opening the outlet valve 42. The problem was therefore solved by closing the outlet valve 42 and then opening the side entry valve 44, allowing pressurised water to flow in for about 20-30 seconds. The tank was then emptied without any further problem.

Case 6

The accumulator vessel 50 was filled with 80 kg of 100 μm dry sand and water. The water height in the bed was 80 cm. The bed was left to settle and rest for 30 minutes. The bed was the fluidised and emptied. All six nozzles 54a, 54b, 54c, 56a, 56b, 56c were used. The water pressure behind all the nozzles was kept constant at 3 bar. Particles started to move around after about 30 seconds. The bed was fluidised after 3 minutes.

Case 7

The accumulator vessel 50 was filled with sand, water and oil. Two litres of Nome oil was filled into the tank. The fluidisation water was then started using the cone nozzles 56a, 56b, 56c. The water pressure behind the nozzles was kept constant at 1 bar. 100 μm sand was then gradually added to the bed while it was moving. The fluidisation water helped to mix and lift up the bed level. When the bed height passed the sidewall nozzles 54a, 54b, 54c they were also opened to help the fluidisation. The water pressure behind them was also kept constant at 1 bar. Five additional litres of oil was added to keep the bed oil-rich. 80 kg of dry sand was filled into the vessel 50 while it was fluidising. The bed was allowed to rest for about one hour before the water above the sand bed and below the oil layer was removed. The bed was then fluidised again using all the nozzles 54a, 54b, 54c, 56a, 56b, 56c.

A slurry a pump was used to mix the oil into the sand/water. The oil was sucked from the top layer through a recycle tube and then injected to the bed at the bottom of the vessel 50. The recycle tube was moved around the bed to achieve a more efficient mixing. This process was continued for 30 minutes. After mixing the oil into the sand/water slurry the mixture was left to rest for 11 days. The oil/water height in the tank was 95 cm and the wet oily sand bed height was 17 cm above the top of the conical section 52. AU the nozzles were closed and the water flow behind them was cut off.

The bed was fluidised after 11 days of storage. AU six nozzles 54a, 54b, 54c, 56a, 56b, 56c were used. The water pressure behind all the nozzles was kept at a constant 3 bar. The bed was fluidised after 2-3 minutes. The height of the fluidised bed after 3 minutes was 36 cm. Sand particles started to move around just after the start of fluidisation. The whole bed was fluidised and everything was moving after about 3 minutes. Fluidised bed height after 6 minutes was 45 cm above the top of the conical section 52.

After fluidisation the bed was emptied. The tank was successfully emptied of sand, water and oil after about 5 minutes. As with the previous experiments the side entry valve 46 was opened for about 20-30 seconds to allow pressurised water in, before emptying the tank. The tank was then emptied without any problem.

Several samples were taken during the emptying process to measure the volumetric concentration of the sand. One could see that at the start of the emptying process the sand/water/oil mixture out of the tank was very dense. The sand came out as a sausage. Its volumetric concentration was close to 100%, but it didn't cause blockage of the outlet. However, with this sand concentration, further problems are likely to arise downstream of the accumulator vessel in a process plant.

The tests show that the accumulator vessel design with a flat end cup or a conical end cup including the flushing arrangements described, work very well. The flushing arrangement fluidises the tank content rapidly and efficiently creating a rotating core region inside the vessel. However, a flat end cup or a semi flat, domed end cup, with the flushing arrangement described, has many advantages. It provides a more uniform operating condition both inside the vessel and for the downstream equipment and process. It is less prone to blockage of the outlet due to the high sand concentration at the start of the emptying process. Such a design also results in a larger effective vessel volume, lower total height and weight. In this design, the fluidised slurry will be in a continuous rotational movement, resulting in a more uniform distribution of particles in the vessel. The more uniformly distributed slurry will flow continuously towards the outlet. When the bed is fluidised and the particles are more uniformly distributed in the vessel, the outflow will have a uniform solids concentration, which results in a more stable operation and a more robust system.

A flattened dome shaped end cup is almost like a flat dish while a conical end cup, to provide a natural fall for the solids, must have a minimum cone angle of 45°, resulting in a cone height of at least one radius. This is significant in relation to the total height of an accumulator. Another important issue is the available useful volume in the vessel. This volume is much bigger in a vessel with a flattened dome shaped end cup than for a conical end cup with the same total height. Using a flattened dome shaped end cup may also result in a lower total weight of the accumulator vessel.

The experiments show that both vessels with the corresponding end cups and flushing systems work generally well. However, when using a conical end cup the main problem occurs at the start of the emptying process. Due to the larger angle of the inclination in a cone and a volume reduction of the cone towards the outlet, all particles push against each other causing bridging of particles over the outlet. This phenomenon may result in the blockage of the outlet which can prevent the solids from exiting the vessel.

Apart from the problems in the start phase of the emptying of the vessel, a conical shape end cup will result in a shorter emptying time period and no particles will remain in the vessel. Due to the natural fall of particles in a conical end cup one does not need to fluidise the bed with the same intensity as for a flat end cup during the emptying period.

Based on the results of the experiments a hemispherical end cup may provide a better performance by using the best aspects of the conical and flat end cups. Because of the larger volume above the outlet in a hemispherical end cup there will be no bridging or clogging around the outlet (as occurs in a conical end cup). There will also be a more uniformly distributed solid/liquid slurry leaving the vessel similar to that in a flat end cup. Due to the larger angle of the inclination in a half spherical end cup the tank will empty faster than in the case with a flat end cup and no particles will be left behind. These effects result in a more stable and robust system.

Claims

Claims:

1. A flushing system for an accumulator vessel comprising a plurality of nozzles disposed at intervals around a perimeter of the accumulator vessel and adjacent a base of the vessel, wherein the nozzles are arranged to direct a fluid supplied to the nozzle to create a rotational flow around the base of the accumulator vessel so as to fluidise solid particles in the base of the vessel.

2. A flushing system according to claim 1 wherein an outlet for removal of accumulated solids is provided in the base of the accumulator vessel.

3. A flushing system according to claim 2, wherein the outlet is arranged for the solids to flow downwards out of the vessel.

4. A flushing system according to claim 2, wherein the outlet is arranged to provide an upward out-flow of solids.

5. A flushing system according to any preceding claim wherein the nozzles are configured to fluidise solids at the base of the vessel without fluidising or lifting an entire bed of solids in the vessel.

6. A flushing system according to any preceding claim, wherein the vessel is an upright cylindrical vessel having a flat base.

7. A flushing system according to any of claims 1 to 5, wherein the vessel has a low-profile domed base.

8. A flushing system according to any of claims 1 to 5 wherein the base of the vessel is "semi-flat", having a flat central portion and a conical outer portion.

9. A flushing system according to any of claims 1 to 5 wherein the base comprises a hemispherical domed end.

10. A flushing system according to any of claims 6 to 9, wherein the nozzles are mounted in the cylindrical side-wall a short distance, preferably less than one fifth of a vessel diameter, above the base.

11. A flushing system according to any of claims 6 to 9, wherein the nozzles are equi-spaced around the side-wall.

12. A flushing system according to claim 10 or claim 11, wherein the nozzles are directed downwards towards the base at an angle to the vertical.

13. A flushing system according to claim 12 wherein the angle is less than 45 degrees and preferably is about 30 degrees.

14. A flushing system according to any of claims 6 to 14 wherein the nozzles are directed at an angle to the radius of the vessel so as to set up the rotational flow.

15. A flushing system according to claim 14 wherein the angle to the radius is 60 degrees.

16. A flushing system according to any of the preceding claims, wherein the nozzles are flat spray type nozzles.

17. A flushing system according to claim 16, wherein the flat spray type nozzle provides a spray over an angle of about 100 degrees.

18. A flushing system according to any of the preceding claims, wherein further nozzles are provided in the base.

19. A flushing system according to claim 18 wherein the further nozzles produce a spray in the form of a flat disc.

20. A flushing system according to any of the preceding claims wherein the nozzles are arranged to provide a fluidising flow covering a substantial portion of the area of the base of the vessel.

21. A flushing system according to claim 20 wherein the nozzles are arranged to provide a fluidising flow covering the entire base area of the vessel.

22. A flushing system according to any of claims 1 to 5, wherein the accumulator vessel comprises an upright cylindrical vessel having a conical section at the base, the conical section having a cone wall.

23. A flushing system according to claim 22 wherein at least some of the nozzles are mounted in the cylindrical side-wall a short distance, preferably less than one fifth of a vessel diameter, above the top of the conical section.

24. A flushing system according to claim 23, wherein the nozzles are equi-spaced around the side- wall.

25. A flushing system according to any of claims 22 to 24, wherein further nozzles are provided in the cone wall.

26. A flushing system according to claim 25, wherein the further nozzles are flat spray type nozzles.

27. A flushing system according to claim 25 or claim 26, wherein the further nozzles are directed downwards at an angle to the cone wall and at an angle to the radius of the vessel.

28. A method of emptying solids from an accumulator vessel, the method comprising: directing a flow of a fluid through a plurality of nozzles into the base of the accumulator vessel so as to create a rotational flow around the base of the accumulator vessel, said rotational flow fluidising solid particles in the base of the accumulator vessel; and opening an outlet in the base of the accumulator vessel to allow the solid particles to be emptied.

29. A method according to claim 28 comprising supplying a fluid under pressure to any combination of nozzles.

30. A method according to claim 29 wherein fluid is provided under pressure through the outlet of the accumulator vessel prior to emptying.

31. A method according to any of claims 28 to 30, wherein the outlet is arranged for the solids to flow downwards out of the vessel.

32. A method according to any of claims 28 to 30, wherein the outlet is arranged to provide an upward out-flow of solids.

33. A method according to any of claims 28 to 32 wherein the step of directing fluid flow through a plurality of nozzles fluidises solids at the base of the vessel without fluidising or lifting an entire bed of solids in the vessel.

34. A method according to any of claims 28 to 33, wherein the fluid flow is directed through the nozzles downwards towards the base at an angle to the vertical.

35. A method according to claim 34 wherein the angle is less than 45 degrees and preferably is about 30 degrees.

36. A method according to any of claims 28 to 35 wherein the fluid flow through the nozzles is directed at an angle to the radius of the vessel so as to set up the rotational flow.

37. A method according to claim 36 wherein the angle to the radius is 60 degrees.

38. A method according to any of claims 28 to 36, wherein the nozzles are flat spray type nozzles, the fluid flow being directed through the nozzles to provide a spray over an angle of about 100 degrees.

39. A method according to any of claims 28 to 39 wherein fluid is directed through further nozzles in the base of the accumulator vessel.

40. A method according to claim 39 wherein the further nozzles produce a spray in the form of a flat disc.

41. A method according to any of claims 28 to 40 wherein fluid flow is directed through the nozzles to provide a fluidising flow covering a substantial portion of the area of the base of the vessel.

42. A method according to claim 41 wherein fluid flow is directed through the nozzles to provide a fluidising flow covering the entire base area of the vessel.