The present invention relates to a compressor, in particular a liquid ring compressor.
Most compressors work with approximate adiabatic process, i.e. without exchanging heat during the compression phase. In practice, e.g. a reciprocating compressor, emit quite a lot of heat, but it is only a small part of this heat which is emitted during compression, most of it is after, or in the end phase. A turbo compressor often has very close adiabatic process.
Some, a bit more special compressors can work very close to isothermal, i.e. the heat which is generated is continuously led away and the temperature is kept unchanged. Examples of these are water driven ejectors and liquid ring compressors, where both are frequently used with vacuum. A screw compressor with oil injection works polytropic, i.e. somewhere between adiabatic and isothermal.
The isothermal process requires less energy supplied than the adiabatic. The difference increase rapidly with increasing pressure difference, as shown in the diagram in
For most objectives it is not desirable with hot gas after compression, and from this and the energy consumption, the isothermal process is preferred in theory.
When this, despite the above, is not employed today, the reason can be found in that existing isothermal or close isothermal compressors have too large hydraulic and dynamic losses. It is an exception for vacuum pumps which n reality is liquid compressors with high pressure difference, p2/p1, but with little pressure height, p2−p1. These can operate with low peripheral speeds on the liquid ring. Another problem is in the technical challenge to be able to remove heat continuously during compression.
Within vacuum both ejector and water ring compressors are frequently used. An ejector exploit the mass speed in a water jet in which the cross section expands and thereby can pull another medium with it. The ejector transform dynamic pressure to static pressure. However, an ejector system has relatively high losses in pump, in nozzle, by impact and friction. Ejectors are rarely used to anything else than the vacuum field. Within prior art the water ring compressor is closest to the compressor according to the present invention.
A liquid ring compressor consists mainly of an impeller which rotates eccentric in an outer enclosure together with a ring of water which the centrifugal force keeps in place against the periphery. The inlet is normally positioned as an opening in one or both of the end walls of the enclosure where the gas is drawn, into the gaps of the impeller. Accordingly, it is arranged openings in the end walls on the pressure side, where the compressed gas is pushed out. All the types can have stationary commutators arranged centrally within the rotor where inlet and discharge happens radially.
Liquid ring compressor does not transform the energy in the water in the same way as the ejector. The static pressure in the ring of water remains constant. The ring of water acts as a piston in every cell of the rotor. The principle for an ordinary liquid ring compressor is shown in
The static pressure in the ring of water has to be the same as the compression pressure, otherwise the water will be pressed out of the cell, i.e. the water ring will be deformed. Thereby it is given that a certain pressure height, p2−p1, require a minimum centrifugal force. A liquid ring compressor usually has considerably higher pressure height and therefore requires higher speed of rotation than a vacuum pump.
The highest loss of friction in a conventional water ring compressor arise when the rotor is touching the wall of the enclosure. The clearing must here be very small, something which involves the water against the enclosure's periphery to have the same speed as the impeller tips of the rotor. Furthermore it must be very little clearance between the sides of the rotor and the enclosure. Also in these gaps there will be high frictions.
Generally the friction losses increase with a square of the speed increase, and in practice the water ring compressor looses level of energy in relation to energy in relation to an adiabatic compressor even at relatively low pressure ratios.
Without these friction losses, the liquid ring compressor has many advantages. It is very simple and can be one stage up to relatively high pressure ratios.
It is apparent that if that the enclosure around the ring of water rotated together with this, the hydraulic friction losses would be minimal. Thus, such a compressor would for normal pressure ratios could exploit the isothermal energy advantages almost in full.
An earlier suggestion disclosed with an outer, rotating cylinder tried to solve the problem of friction, without this leading to a feasible solution. U.S. Pat. No. 5,100,300 and U.S. Pat. No. 5,370,502 describes liquid ring compressor with a cylinder which floats on a film of liquid or gas between the cylinder and the outer stationary enclosure. By floating on a liquid film, it is doubtful whether it would be achieved any reduction in the friction, and with gas it would probably not be possible to achieve sufficient bearing capacity and stability, such that the cylinder do not touch the enclosure.
In a later patent, U.S. Pat. No. 5,395,215, from the same firm, it is suggested a bearing of this cylinder in an outer enclosure, where a number of rollers are inserted in the wall of the enclosure where the cylinder is supported by the rollers. This do not seem realistic with the actual rotational speeds the rollers will achieve. A subsequent patent, U.S. Pat. No. 5,653,582, returns back to fluids as the peripheral bearing for the rotating cylinder and suggestion to the basic solution.
U.S. Pat. No. 5,251,593 discloses as the previous application that it is an intricate problem to get to, in relation to each other, eccentric bearings in combination with a stationary canals for the inlet and discharge of the gas. This publication indicates a bearing of the outer rotating cylinder on one side and the rotor on the opposite side, where a stationary plate close to the open end of the rotor has canals for inlet and discharge. It is mainly two decisive weaknesses with this design. The first is the one-sided bearing this solution gives, where the bearing load becomes uneven and too high. At the same time large axial to thrust forces arise. The other weakness is the problems with achieving a reasonable gas tight sealing between the outer rotating cylinder, and the plate where the inlet and discharge canals are positioned in a circular plate, inlaid in the open end of the rotor. It would here be gas leaks backwards from cell to cell and in addition out through the circular gap between the stationary plate and the rotor. The principle is unrealistic for practical purposes.
Despite may studies and suggestions over many years, it evidently has not been possible to reach a design which fulfil the requirements to function satisfactory. Thus at present there exists no liquid ring compressor with such co-rotating rotor. The above mentioned publications indicates that one has been tied up to the starting point for a rotor and communicator system like those in conventional vacuum pumps and compressors for relatively low pressure, with the above mentioned limitations in speed. This is reflected in relatively wide rotors with communicator on each side, which lead to long bearing distance and high bearing loads. In a compressor with liquid ring in the outer co-rotor, the geometry will be wrong, which will lead to bearing relationship which is unsuitable for existing bearing types. With communicator on each side, it becomes four sections with gaps where there exists leakage from the zones on the pressure side.
The compressor according to the present invention has the objective to solve this problem which up to know has prevented a water ring compressor to exploit the above mentioned advantages with a co-rotor for the liquid ring. Another objective is to achieve almost isothermal compression with a new, very efficient direct injection of liquid into the gas during the whole compression stage.
Water as injection liquid has very good thermal properties, and is desirable to use with those gases which allow this. But, as for pumps and the like, the design for a liquid compressor with a co-rotor require a distinct division between water and the bearing of the co-rotor. From the development of screw compressors with water injection it is known and it has been problems with sealing on the pressure side of the screws. Firstly, water has small to little lubricating effect on the sealing which must have relatively high pressure towards the axle and therefore high wear. Furthermore, water penetrates easily through even the finest caps, and especially high pressure. Below it will be evident that the compressor according to the present invention solves the sealing problem by eliminating the reasons for them. The aforementioned objectives will be satisfied with the liquid ring compressor according to present invention as it is defined in the attached claims.
The invention will now be described, by way of example, with reference to the accompanying drawings, in which
The main parts in
With the largely reduced friction in the water ring due to the co-rotor, is it possible to make the rotor considerable narrower at the same time that the discharge volume is compensated with a considerable increase in speed. Thereby the inner pressure in the water ring is increased and the compressor can deliver with very high pressure.
A short rotor get little bending force from the gas pressure and is thereby allowed to be fixed to a flange on its axle only at one end wall and thereby being able to have a simple commutator in the entire width of the rotor. It is then only created two leakage gap between the commutator and the rotor. These gaps are the only place where leakage from the pressure side will find place. It can leak actually to both sides from the gap and along the periphery from the pressure discharge against the inlet, especially in the direction of rotation. Even in very small gaps, pure gas without liquid will with the present pressure be able to leak in considerable amounts, with smaller amounts deliver and lower efficiency as a result.
The surface of the rotor 6 on the inside towards the commutator is at its ends 63 smooth, with interweaving canal openings 62 to each individual cell. On
The liquid ring compressor according to the present invention could be designed with hydro dynamic bearing for the co-rotor. These could then be lubricated and cooled with the same liquid which was used for injection. But with the starting point with necessary axle diameter and speed, research shows however that the friction losses in such bearings then will be very high and some of the advantages with a co-rotor are lost. With higher pressure the bearing size increases farther and the losses in them become unacceptable.
On the other hand the same relationship seems to be acceptable for relatively large ball or roller bearings, but at the same time this leads to new problems around the bearing sealing. Bearings with integrated seals can not operate close to the necessary speeds and there do not exist any static seals which allows this, or that will achieve acceptable lifetime. Labyrinth seals however are touch-free and can operate with high speeds, but do not give any static sealing. These seals assume there are no differential pressures across the seal.
To prevent differential pressure across the bearing a co-rotor is aired to the compressor enclosure through the holes 81 as shown in
This is led to the need for designing and completely new lip seal 82, shown in more detail in
It is evident that the lip during operation places itself against the edge of the openings in the co-rotor and walls so it is relatively small movements the lip bends from being in contact with the axle until it is not. This give little fatigue effect even with frequent start and stop.
The seal is in other words static at low speeds and seems dynamic at higher speeds, where its purpose is only to prevent bearing grease to be projected out. Bearing of the co-rotor will with this system get about the same surrounding relationship and security as if they operated in air.
It is the diameter on the bearing of the rotor and the eccentricity between the rotors which decide the diameter on the bearing axles of the co-rotor because the bearing of the rotor as shown is inserted in these. The load on the bearing of the rotor becomes the same as for the co-rotor. To withstand this load, at the same time as giving smallest possible dimensions for the axle to the co-rotor, so-called needle bearings are used for the rotor. The purpose and the necessity by integrating the bearing of the rotor in the axle of the co-rotor, is to achieve short as possible bearing distance which give smallest axle diameter. For the bearing for this axle the peripheral speed allows ordinary static seals, and the bearing can be lubricated by oil.
To prevent the creating of trapped gas pockets in the cells of the rotor 61, are these as shown in
The commutator is fixed to the one, stationary axle 8 for the co-rotor. The axle connects the canals of the commutator with the respective inlet and discharge for gas and injection liquid.
When the compressor according to the present invention is used for other gases than air, e.g. cooling system or in a petrochemical plant, it can be useful to use the actual gas in the liquid phase for injection and as liquid ring.
At an expected, reasonably lower energy need than for a turbo compressor the compressor will according to the present invention be very suitable as compressor in a gas turbine plant which operates with relatively high pressure relationship. The air from this will indeed in contrast to the turbo compressor, be cold, but it is necessary to note that the heat which the turbo compressor delivers is taken out of the axle of the turbine and reduces the output effect accordingly at the same time as the warm air to do not make it possible for heat recovery from the turbine exhaust. By use of the compressor according to the invention, the air from the compressor can be heat exchanged with the exhaust gas and almost reach the same temperatures as after a turbine compressor.
Number | Date | Country | Kind |
---|---|---|---|
20021844 | Apr 2002 | NO | national |
Number | Date | Country | |
---|---|---|---|
Parent | 10511753 | May 2005 | US |
Child | 12164908 | US |