NZ506939A - Pumping apparatus with low NPSH with fluid bled from outlet back to inlet - Google Patents

Abstract

Fluid is bled from a pump outlet pipe (at 2) back to the pump inlet via lines 1 (only one shown). Bleed flows are injected at higher velocity heads (at 3) compared to bleed outlets (at 2). The nozzle unit is a separate unit engaged or engagable in the pumping system and the pumping system (and preferably the unit itself) provides the increase in total velocity head of the bleed flow. Divisional filed as 517598

Description

NEW ZEALAND PATENTS ACT, 1953 Divided out of 329969/334408 Date: 13 March 1998 / 26 February 1999 COMPLETE SPECIFICATION "Improved Pumping Apparatus and Methods" We, MOUAFAK ARIF ZAHER, a New Zealand Citizen of 47 Grande Avenue, Mt Albert, Auckland, New Zealand, and UNITEC INSTITUTE OF TECHNOLOGY, a New Zealand institution under the Education Act 1989 of Carrington Road, Mt Albert, Auckland, New Zealand do hereby declare the invention for which we pray that a patent may be granted to us, and the method by which it is to be performed, to be partici|:^r^|dp.scnbejl^n by the following statement: Office of NZ The present invention relates to improved pumping apparatus and methods, and, in particular although not solely, to pumps (centrifugal) to operate in low suction head conditions.

Pumps such as centrifugal pumps have been manufactured for many years. They have been used in boiler feed, mine drainage and petroleum/chemical industry for the pumping of different fluids.

There are many places in process industries where low NPSH (Net Positive Suction Heads) are required for pumping liquids which are at or near their boiling points. In conventional pumps as the NPSH is reduced, cavitation sets in and it eventually becomes extensive enough to cause blockage of the blade passages, efficiency breakdown, and possibly also a gradual destruction of the components of the pump. As is well know, cavitation may be expected in a flowing liquid when the local pressure falls to the vapour pressure of the liquid. Local vaporisation of the liquid will then result, causing development of a vapour filled bubble or cavity in the flow. The low pressure cavity is swept down stream into a region of higher pressure where it collapses suddenly, the surrounding liquid rushing in to fill the void. At the point of disappearance of the cavity the in rushing liquids come together momentarily raising the local pressure within the liquid to a very high value. If such cavity collapse is near the boundary wall, such high local pressure can cause erosion of the wall material. An advantage of pumps having a low NPSH is the ability to reduce holding tank heights (and save construction cost), or to pump fluids which are near their boiling point, or by running the pump faster and replacing multi-stage (booster) pumps.

Development on pumps to reduce or eliminate cavitation at inlet include work on what is known as the super cavitating type of pumps which use an inducer upstream from the pumps impeller. Such an inducer is normally driven by the axle of the impeller and hence increase the demand for power from the motor driving the pump. In general, anti-cavitation pumps can increase the inlet type number (specific speed) of a high speed pump by two to three times, thus giving a generous margin of safety in respect of permissible NPSH.

The provision of a booster pump upstream from the centrifugal pump where cavitation may result is also a common way of overcoming cavitation. The boost pump increases the head of fluid at the centrifugal pump to there by reduce the occurrence of cavitation.

Yet an alternative development to conventional pumps includes the provision of guide vanes upstream of the impeller of the pump wherein the guide vanes provide a swirling motion to the intake fluid to improve the angle of incidence of the fluid flow with Intellectual Property Office of N.Z. 0 5 m 2002 RECEIVED -3 intellectual Property Office of N.Z. 0 5 MAR 2002 RECEIVED the leading edge of the impeller. The problem with both super cavitating pump and pumps having up stream guide vanes is related to the increase in surface area over which the intake fluid is required to flow hence creating a head loss as a result of friction. The booster pump or super-cavitating pump also require a further power input.

A further alternative way in which the performance of a pump may be improved by avoiding cavitation inside of the pump itself. Such cavitation may occur under partial load conditions. In UK Patent Application GB2058218 there is described a means of drawing fluid form the volute casing of a pump, collecting it in a cylindrical reservoir and directing the fluid via openings to the impeller blades at inlet. The configuration of this method is such to induce a backflow within impeller blades which counteracts the occurrence and development of the secondary flows in such a way that cavitation loading of the impeller for part-load conditions is eliminated or reduced. This specification however deals with minimisation of the energy loss as a result of secondary flows, sometimes called re-circulating flows. These are usually in the form of vortices with a direction of rotation opposite to that of the main stream. These are usually found at the tips of the impeller vanes on the exit and inlet sides. The losses are normally accounted for as part of the total loss assumed within the major elements of the pump, which also include fractional losses.

The function of the injection of fluid of the invention of GB2058218 is analogous to providing a series of fixed guide vanes before the impeller in order to guide the fluid in a certain direction in respect of partial load vortices.

US4492516 deals with the losses as a result of impeller re-circulation by the provision of passage ways shaped to direct the flow of fluid to be reinjected at the inlet of the impeller. As such head losses as a result of secondary flows or friction are small compared to the total headloss of the energy equation of fluid flow its reduction or elimination, although having some effect, proportionally it has very little effect on such energy.

It is therefore an object of the present invention to provide improved pumping apparatus and methods, and in particular although not solely to centrifugal pumps to operate in low suction head conditions, or which will at least provide the public with a useful choice.

Accordingly in a first aspect the present invention consists in a pumping system comprising means defining a pump having an inlet and an outlet, an inlet conduit connected to said inlet of said pump for the inlet flow of fluid to said pump intellectual Property Office of HZ. 0 5 MAR 2002 -4- RECEIVED a delivery conduit connected to said outlet of said pump for the outlet flow of fluid from said pump means adapted to bleed at least part of the outlet flow (hereinafter the "bleed flow"), a nozzle unit engaged to said inlet conduit and positioned upstream of said inlet of said pump, means to deliver said bleed flow to said nozzle unit, said nozzle unit including at least one means to inject the bleed flow into the inlet conduit, means capable of increasing the total velocity head of bleed flow to, in use, inject said bleed flow into said inlet flow to said pump at a higher total velocity head than the total velocity head of bleed flow at said means adapted to bleed.

Preferably said means to deliver is a conduit.

Preferably said means to inject is a nozzle.

Preferably said means capable of increasing velocity head is provided by a reduction of total cross-sectional flow area of said means to deliver to thereby increase the velocity pressure head of said bleed fluid flow prior to being injected into said inlet conduit.

Preferably said means capable of increasing velocity head is provided at said nozzle in the form of a reduced total cross-sectional flow area for bleed flow through said nozzle.

Preferably said nozzle comprises a passage for fluid with an inlet for bleed flow to be received from said means adapted to bleed (via conduit) and an outlet connected for bleed flow delivery to said inlet conduit, wherein said inlet of said nozzle is of greater cross sectional area that said outlet of said nozzle.

Preferably said passage of said nozzle is gradually tapered over the length of said passage between said nozzle inlet and said nozzle outlet such that the cross sectional flow area of said outlet is smaller that the cross sectional flow area of said inlet of said nozzle.

Preferably the ratio of said cross sectional flow area of said nozzle inlet to the cross sectional area of said to said nozzle outlet is substantially four.

Preferably said passage of said nozzle is substantially circular in cross sectional area and the ratio of inlet diameter to outlet diameter is substantially two.

Preferably the ratio of the length of said passage to the diameter of said nozzle inlet is substantially two.

Preferably said nozzle is engaged with said inlet conduit to, in use inject bleed flow fluid into the inlet conduit through said nozzle outlet at substantially right angles to the intciteclua! Property j Office of N.Z. 0 5 MAR 2002 -5- RECEIVED main flow direction of fluid in said inlet conduit.

Preferably said nozzle is engaged with said inlet conduit to, in use inject bleed fluid into the inlet conduit through said nozzle outlet to induce a rotational flow to inlet flow of fluid approaching said pump.

Preferably there are two and preferably four nozzles for injecting fluid into said inlet conduit.

Preferably said nozzles are each positioned the same distance upstream of said pump.

Preferably said nozzles are equi-spaced about the perimeter of said inlet conduit.

Preferably said pump is a centrifugal pump.

Preferably said pump is a centrifugal pump and said rotational flow is co-rotatory with the rotational direction of the pump impeller.

In a further aspect the present invention consists in a nozzle unit for injecting fluid into a main inlet flow stream of a pump comprising; a conduit section capable of insertion into a main inlet conduit defining the flow boundary to said inlet flow stream of said pump said conduit section comprising an inlet, an outlet a wall defining region between said inlet and outlet to, in use, provide a substantial continuation of the boundary to the main inlet flow stream of fluid, and at least one aperture through said wall defining region an injection nozzle for or defined by said at least one aperture to in use deliver a fluid bled from the outlet flow of said pump into said inlet flow stream, said injection nozzle being a conduit having an inlet and an outlet said outlet at said wall defining region wherein the flow area of said inlet to said conduit is larger than said outlet to said conduit.

Preferably said conduit of said nozzle is secured to said conduit section such that in use the flow path of injected fluid at said outlet to said conduit is at substantially right angles to the main inlet flow stream of a said pump.

Preferably said wall defining region of said conduit section is defined by a substantially constant diameter bore corresponding to the diameter of said main inlet conduit.

Preferably said conduit of said nozzle is secured to said conduit section such that n itvii-wwiuui r »ty O'fior of M.2 0 5 MAR 2902 RECEIVED in use the flow path of injected fluid at said outlet to said conduit is at substantially right angles to the main inlet flow stream of a said pump and at a tangent to said boundary of said main inlet conduit.

Preferably said inlet of the conduit of said nozzle is of a flow area substantially four times that of the flow area of the outlet to said conduit.

Preferably said conduit of said nozzle is in cross section substantially circular.

Preferably said conduit is gradually tapered between said inlet and said outlet over a distance substantially twice that of the diameter of said inlet to said conduit.

This invention may also be said broadly to consist in the parts, elements and features referred to or indicated in the specification of the application, individually or collectively, and any or all combinations of any two or more of said parts, elements or features, and where specific integers are mentioned herein which have known equivalents in the art to which this invention relates, such known equivalents are deemed to be incorporated herein as if individually set forth.

The invention consists in the foregoing and also envisages constructions of which the following gives examples.

Preferred forms of the present invention will now be described with reference to the accompanying drawings in which; Figure 1, by way of example, is a plan view diagrammatically showing a pump and a bleed control adapted to provide a drawing off of liquid from the output of the pump and the injection thereof into the input of the pump, the broken line "XX" showing an axis about which in one rotational direction the impeller of the pump rotates and the preferred co or counter rotatory motion on the input or suction side of the pump is preferably induced, Figures 2-4 illustrate arrangements which by way of experimental results are explained herein, wherein pump is placed in respect to a fluid source reservoir which for figures 2 and 3 provide a suction head of 105mm and 557mm respectively and a suction lift for figure 4 of 875mm, Figure 5 to 7 illustrate in section a preferred configuration of a nozzle unit and delivery conduit, wherein the nozzle unit has two or four nozzles for injecting the bleed flow from the output flow of the pump at points on the suction side, Figure 8 is a schematic view of the flow of fluid from suction to the delivery side including the flow diversion from the point of bleeding to the point of injection, Figure 9 illustrates the dimensions of a nozzle, Figure 10 illustrates a sectional view of an impeller wherein there has been illustrated the flow velocities of fluid in respect of the impeller entrance, Figure 11 illustrates a sectional view of an impeller wherein there has been illustrated the flow velocities of fluid in respect of the impeller entrance, wherein a pre-rotation adds to the velocity vector Cul, Figure 12 is a schematic view of the testing facilities used wherein the fluid for delivery to the pump has a delivery head X, Figure 13 is a plot of performance characteristics of test results conducted as herein after defined wherein a four nozzle unit with a nozzle diameter of 12mm was used wherein the fluid prior to the pump had a suction lift of 875mm, for test pump A, Figure 14 is a plot of performance characteristics of test results conducted as herein after defined wherein a four nozzle unit with a nozzle diameter of 12mm was used wherein the fluid prior to the pump had a suction head of 557mm, for test pump A, Figure 15 is a plot of performance characteristics of test results conducted as herein after defined wherein a four nozzle unit with a nozzle diameter of 12mm was used wherein the fluid prior to the pump had a suction head of 105mm, for test pump A, Figure 16 is a plot of performance characteristics of test results conducted as herein after defined wherein a four nozzle unit with a nozzle diameter of 40mm was used wherein the fluid prior to the pump had a suction lift of 1500mm, for test pump B, Figure 17 is a plot of performance characteristics of test results conducted as herein after defined wherein a four nozzle unit with a nozzle diameter of 40mm was used wherein the fluid prior to the pump had a suction head of 1300mm, for test pump B, Figure 18 is a plot of performance characteristics of test results conducted as herein after defined wherein a four nozzle unit with a nozzle diameter of 40mm was used wherein the fluid prior to the pump had a suction head of 500mm, for test pump B, Figure 19 is a plot of relative performance of pumps A and pumps B as herein described as part of the testing as a non dimensional comparison with both head and flow rates in respect of a conventional pump.

Figures 20-23 show the optimum bleeding required at three different positions (875mm suction lift, 557mm suction head) using different shapes of nozzle units. Figures 24-25 are as herein described with reference to the formulas.

Figures 26 is a perspective view of one preferred form of a nozzle unit.

As used herein "counter rotatory" or "co-rotatory" to the impeller motion does not imply any correspondence in rotational speed of (a) the rotary motion or the bleed stream or nozzle induced rotating liquid to that of (b) the rotating speed of the impeller or to the axial orientation of the impeller with respect to the flow rotation at injection.

Nomenclature Intellectual Property Office of N.Z. 0 5 MAR 2002 RECEIVED A cross sectional area of nozzle, Bn bleeding number = N x A x Cd / Qw, b, vane inlet breadth, b2 vane exit breadth, C Absolute velocity, Clu circumferential velocity component on entrance of impeller, C2u circumferential velocity component on exit, Ca axial velocity in entrance channel, Cd fluid velocity at exit from nozzle, dn nozzle diameter, Dd delivery pipe internal diameter, Ds Suction pipe internal diameter, dB bleed pipe diameter to nozzle, H0 head at operating point for conventional pump, AH total head difference between conventional and new pump arrangement, Ln nozzle tapered length, N number of nozzles, n speed, p, static pressure at entrance of impeller, pa ambient pressure, Ap swelling, Apc total swelling, Apt theoretical swelling, Apz pressure fall in valve, Pc total power, Pb power due to secondary losses, Pm power loss due to friction in bearings (mechanical), Q volume of a stream of fluid, Qd volume of stream of fluid through nozzle, Q0 volume of stream of fluid at operating point, Qw volume of a stream of fluid through the impeller, r, inlet radius to the pump, r2 exit radius to the pump, R, r radius, Rd distance from nozzle exit to axis of impeller, Re Reynolds number, Intellectual Property Office of N.2. 0 5 MAR 2002 RECEIVED Rin radius at entrance, Ro Rossby number (Vr / (Rn)- dimensionless group, Rex radius at exit, U, rotational velocity at entrance to impeller, u2 rotational velocity at exit from impeller, V velocity of fluid at radius r, w relative velocity, (3 entrance angle, P density of fluid, CO angular velocity, X hub/tip ratio, 9 coefficient of flow Where herein mentioned with reference to the nozzles and by example 2Nx40mm dB means two nozzles each with a nozzle diameter (dB) of 40mm, 2 x 2 N refers to two nozzle units in series each of two nozzles. FOV fully opened valve '/2 OV Vi open valve 1/4 OV 1/4 open valve The present invention will now be described to show how development and control of the suction or input side fluid injection has allowed the adoption of higher operating speeds without embarrassment of high NPSH requirements. It will also be shown that preferably such injection is provided with a rotary motion ahead of the impeller and by transferring part of the energy of fluid at the delivery side of the pump, back to the suction side using at least one nozzle(s) placed in the suction line, improves the overall efficiency of the pump arrangement. The control action preferably takes place through the aid of a valve system, which when it is closed position would allow a 100% delivery side discharge. Changes in discharge, pressure and power will occur during the action of opening the control valve system, however these are offset by improved intake flow at low NPSH.

As is well known for incompressible flow of uniform density fluid the static pressure head plus velocity pressure head plus any losses as a result of, for example friction, remains constant .i.e p v2 ~+—+"L0sses=C V 2 g PtQDpiiv Office of h.z.' 0 5 MAR 2002 Received It is known that on a suction side of a pump as result of frictional losses from the walls of the conduit, the head loss can be a factor in contributing to reduce the total local pressure of the fluid where at some point in the pipe this pressure becomes, less than the vapour pressure of the fluid resulting in cavitation. Losses may also be incurred by secondary flows such as eddies which may form at the impeller entrance of a pump. These eddies can also reduce the total local pressure of the fluid and contribute with other factors to the formation cavitation. However the present invention does not in a primary aspect deal with the total local pressure drop as a result of these losses. It is only the secondary results of a preferred embodiment of the present invention which may also increase the total local pressure by reducing or eliminating these losses.

The prior art which deals with modifying the flow of intake fluid normally deals with reducing the head losses as such a result of such secondary flows. The known use of booster pumps positioned upstream from the main centrifugal pump increase the pressure head to ensure the total local pressure of fluid upstream from the impeller remain above the vapour pressure of the fluid. The current invention deals with increasing the velocity head: of the fluid at inlet, achieved by bleeding from the delivery side, fluid at an increased static pressure (as a result of the pump) to the suction side in such a way to increase its velocity head which is injected into the suction side and hence increases the acceleration of the fluid at suction and therefore increases its total local pressure.

With reference to figure 1 a conduit 1 is positioned between the discharge and inlet of the pump. The pump in this instance is a centrifugal pump wherein the axis of rotation of the impeller is substantially co-axial with the inlet pipe. The present invention does have applicability to other pumps including centrifugal pumps wherein the axis of rotation of the impeller is at right angles to the inlet pipe.

The conduit 1 provides a connection between the point of bleeding 2 and the at least one nozzle 3 of the arrangement. The conduit 1 is preferably a circular cross section and may be provided with a valve between points 2 and 3.

The nozzles are placed in the suction pipe or at a suitable point where the best results in AH are achieved for the particular pump being utilised. This position does vary from pump to pump and needs to be determined experimentally. The present invention is not necessarily limited to there being only one point of injection into the suction pipe.

Oifee of i\j.z. 0 5 MAR 200? Indeed this is not preferred and it has been shown by experimentation that at least two and preferably four nozzles in a nozzle unit provide the best AH. Accordingly each nozzle of a nozzle unit has preferably a separate conduit 1 to provide the fluid connection between the bleeding means of the output and the input pipe. Hence preferably there are also corresponding numbers of points of bleeding on the upward pipe. These are preferably positioned about a pipe at approximately equal space there around. More than one nozzle unit, placed in series may also be appropriate.

The fluid of higher static pressure head on the delivery side (or out put side) of the pump is bled at a rate through conduit(s) 1 to at least one nozzle positioned upstream of the pump to bleed the fluid from the discharge line to the suction line. In the preferred form of the invention, prior to the point of injection of the bleed fluid the velocity head energy is increased. This is preferably achieved by a reduction in cross sectional area of the fluid passage. As opposed to increasing the energy of the suction line fluid by bleeding fluid into the suction line with a high static pressure head, the injection of fluid with a high velocity head has less demand on the delivery side total pressure head. As the principle of a centrifugal pump is to convert velocity head of the inlet fluid to static pressure head of the outlet fluid, bleeding fluid with a high static pressure head back into the delivery side would relatively reduce the output while achieving of higher total pressure head, higher overall efficiency and lower power input. These were proved experimentally.

As the kinetic energy of the fluid is dependent on the V2 (the velocity of injected fluid) the addition of energy as a result of an increase in velocity from the delivery side to the suction side is a more effective way of increasing the energy of the flow entering the impeller. In this respect 2:1 ratios of flow path diameter have shown experimentally to provide the best AH.

In the preferred form the reduction in cross-sectional area occurs at the nozzle unit immediately before the point of inlet into the suction flow for reason for which are herein after described. With reference to figures 5-7 wherein there is shown in cross-section preferred configurations of nozzle units 5 wherein in figures 5 and 6 the nozzle units have two nozzles and in figure 7 there are four nozzles. With reference to figure 9 there is illustrated the momenclature referred to herein in relation the dimensions dn and dB which are proportional to the area of the nozzle at its inlet and outlet and hence the velocity of fluid at each point is also dependent thereon. Preferably the cross-sectional flow area of the nozzles, inlet and outlet conduits and other conduits are substantially circular. Other shaped cross-sections may also be used.

In a broad sense the accelerated (i.e. increased kinetic energy) bleed fluid may be IriiGiiGc'ua! P'opGitv Office of N.Z. 0 5 MAR 2002 mjected into the suction side of the flow at any angle. However in the preferred form such injection may occur at right angles to the flow direction (whether in a radial or tangential sense) and preferably occurs in a tangential direction. The preferred tangential injection of fluid is preferably achieved by a nozzle unit as by example shown in figure 5. Such injection reduces any losses as a result of the mixing of the main flow with the injected fluid. The preferred highly swept back design and preferred uniform flow of fluid through the nozzle will decrease such nozzle losses. In the preferred form of the invention, the injected fluid simultaneously imparts a rotatory flow (i.e. a swirl) to the flow of fluid prior to it reaching the impeller. This is preferably achieved by the tangential injection of fluid and can result in improved intake angles (i.e. the angle of incidence of fluid in respect of the impeller blades and may also have an effect on reducing the losses as a result of secondary flows).

For at least a centrifugal pump, action of centrifugal forces, caused by rotating blades, will create change in pressure on the suction side, which increases radially towards the pipe walls. The pressure distribution in planes perpendicular to the pipe or input flow axis ahead of the impeller defines (for equal pressure curves), a parabola which gets steeper in regions closer to the impeller. As particles of fluid get closer to the impeller, a rotational motion is created because the absolute velocities of these particles are the sum of both axial and radial components. This motion does not take place suddenly, but starts at a certain distance from the pump entrance. At a certain section ahead of the impeller, particles with axial motion change gradually to particles with rotational motion in vicinities closer to the impeller suction side. This change is accompanied by change in momentum from approximately zero to values equivalent to those particles near the impeller. The present invention preferably envisages the use of a plurality of nozzles although one nozzle may suffice in certain instances. Their location and number are matters to be determined by experimentation in relation to particular pump parameters and geometries so as to optimise effect and performance. Likewise with the control systems and/or the nature of the bleed streams. However with reference to the examples herein after described we have shown what has experimentally been found to be favourable conditions of operation for particular pumps and system parameters.

The bleed flow from the delivery side of the pump is transmitted via a bleeding means 2, through a connection conduit 1 to the suction side and with nozzled 6 entry results in injection of bled fluid into the suction side.

The conversion of pressure energy to kinetic energy between the point of bleeding at 2 from the discharge line to the suction line is preferably achieved at the nozzles substantially immediately prior to fluid injection into the suction line. This is preferably achieved by a tapering of the nozzles which with reference to figure 9 preferably has a ratio defined by dB/dN occurring over a distance L, where dN and dB are the nozzle outlet and inlet diameter respectively and bleed conduit diameter and their respective areas are directly proportional to such diameter. It must however be appreciated that the bleed flow area need not necessarily be of a circular nature and may be of any suitable shape cross-section.

From experimentation we feel that the best improvement (or change in total head delivery AH) in head of a pump of the present invention compared to a conventional pump is where the ratios of d„/dN and dB/dL are substantially as follows: We believe that these ratios are the preferred ratios independent of the number of nozzles, (i.e. means to inject) bleeding rate of flow (Qd) and pump size.

As mentioned above it is desirable, in order to achieve the best performance, to experimentally investigate the flow of particular spec pump. This is hereinafter referred to as the calibration test. Armed with the insight gained is to predict the performance of new designs.

Both the optimal range of Bn and the value of N are determined from tests. Both Cd and Qw should then be calculated using given parameters for each specific pump which can then be used for different spec pumps.

The pumps selected represented three duties in which the operators can have cavitation troubles with conventional pumps. The conventional duties of the pumps used to describe the examples were: dB/dN = 2:l dB/dL =1:2 Examples Product Pumping temp. Specific gravity Vapour press.

Capacity Differential head NPSH A Water Ambient 1.00 2.55x102 kg/m2 0.36 Psi Abs. 0.8 L/s 2 m 0.5 m B Water Ambient 1.00 2.55x102 kg/m' 0.36 Psi Abs. 20 L/s 4.5 m 1.5 m C Water Ambient 1.00 52.8x102 kg/m2 7.51 Psi Abs. 5 L/s 3.5 m 0.75 m fractectsjai Property Off'ce of HZ. 0 5 MAR 2002 RECEIVED The first item in the design process was to decide on the optimum diameter (dn) and the positioning of the nozzle unit on the suction side: Example of Test Pump A (i) The constant head water tank: It is an open tank: 1x0.7x2.2 cu.meters, elevated 100 mm from ground level. Delivery from tank at suction side is through a circular opening 27 mm in diameter. The tank is made of galvanised mild steel, 4 mm sheet thickness. A container of 20 liters is positioned near the top side of the tank to measure the volume of discharged water per unit time. (ii) The suction line: Is 297 mm long made of mild steel. The internal diameter is 27 mm. Arrangement for spacers with different lengths is provided to facilitate positioning of nozzle units at the desired distance from the impeller. Suction valve is mounted at a distance of 85 mm from the tank edge. (iii) The pump: It is a model pump with a centrifugal impeller and single discharge volute. There is a conical diffiiser at the higher-pressure end of the pump. The pump has the following dimensions: r, = 10 mm, r2 = 30 mm b, = 8 mm, b2 =12 mm Number of vanes = 7 Model No.: 45-00701-01 Manufacturer: AETTS-Australia H (head) = 2 m at Q (discharge) = 0.8 L/s (iv) The electric motor: The drive motor is asynchronous and runs at the very nearly constant speed of 1400 r.p.m, with the following characteristics: Horse power HP = 1/3, Phase =1, Amperes =1.1, Hertz Hz = 50, Voltmeter/ammeter is used to measure the voltages and currents consumed. (v) Temperature detector: Made use of a mini-surface detectors, having a measuring range between (-50°C) up to (+250°C). Readings at suction and delivery sides were recorded for each experiment. (vi) The nozzle units: Nozzles on the suction side of each test, are made of aluminum so that the effect of corrosion and fr-uiv r •-!-.- o ;f(V G.lic,- 05 w,.4. 0 5 MAR 2002 abrasion on the change of friction losses would be negligible. The following table gives the dimensions of the used nozzles: Dimensions of nozzle units used with pump A No of nozzles Bleed pipe diam.

Nozzle head diam.

Length of taper (N) to nozzle (dB) mm (dn)mm (Ln) mm a 2/4 3 1.5 6 b 2/4/two series 6 3 12 c 2/4/two series 12 6 24 d 4 12 3 24 e 2/4/two series 19 9.5 38 f 2/4 19 4 38 (vii) The discharge line: It is 2.620 m long ( for positions x (see figure 2, 3) 105 and 557 mm), and 4.558 m long (for suction lift position see figure 4). Pipe is made of PVC, 3 mm thick, and with internal diameter of 25 mm. The discharge valve is located at a distance of 300 mm from pump exit.

Calibration test for pump A: These tests were performed to study the behavior of the pump performance using bleeding system from the delivery side to suction side. Tests were performed in steady, non pulsating conditions, and almost constant temperature. The variables in the test were a) the nozzle number N b) the nozzle diameter dB c) the suction dop/lift d) delivery valve position e) bleed valve position (and hence Qd) f) the upstream position of the nozzle from the impeller. g) swirl direction of suction flow The tests for pump A involved three positions X, pump: (refer to Figure 2-4) -When level of water tank is 105 mm above the suction valve.

-When level of water tank is 557 mm above the suction valve.

-With a suction lift of 875 mm from centre of suction pipe.

Two types of nozzle units were used m the first test. These were: fiuc«ectuai Property Office of N.Z. 0 5 MAR 2002 DECEIVED - 2 nozzles x 3mm dB - 4 nozzles x 3mm dB Each of these units was tested to find the optimum resolutions, when working on the following parameters: -Delivery valve: Fully open, lA open, lA open, and fully shut.

-Bleeding valves: Fully closed, fully open, Vi open, and lA open.

-Nozzle position from impeller: 68mm, 85mm, 100 mm, 125 mm, and 200 mm.

-Direction of rotation of flow from nozzle included rotation in the same direction as that of the impeller, and another against the direction of rotation of the impeller.

In the second group of tests, four types of nozzle units were used. These were: -2 nozzles x 6mm dB -4 nozzles x 6 mm dB -2x2 nozzles in series [ at 15, 40, and 60 mm apart ] x 6 mm dB -2x4 nozzles in series [ at 15, 40, and 60 mm apart ] x 6 mm dB Each of these units was tested in the following positions and combinations: -Delivery valve: Fully open, Vi open, % open, and fully shut.

-Bleeding valves: Fully shut, fully open, lA open, and lA open.

-Nozzle position from impeller: 65 mm, 85 mm, and 100 mm.

-Direction of rotation of flow from nozzle: With the same direction of rotation as that of the impeller.

(C) In the third group of tests, five types of nozzle units were tested. These were: -2 nozzles x 12mm dB -4 nozzles x 12 mm dB -2x2 nozzles in series [ at 40 mm apart ] x 12 mm dB -2x4 nozzles in series [ at 40, and 80 mm apart ] x 12 mm dB -4 nozzles x 12 mm (3 cp) dB Each of these units was tested in the following positions and combinations: -Delivery valve: Fully open, XA open, lA open, and fully shut.

-Bleeding valves: Fully shut, fully open, XA open, and % open.

-Nozzle position from the impeller: 65mm.

-Direction of rotation of flow from nozzle: With the same direction of rotation as that of the impeller. iuc!:cc!udi Property C:fco Of i\l.Z. 0 5 MAR 2002 r f r. f i v f n (d) In the fourth group of tests, six types of nozzle units were tested. These were: - 2 nozzles x 19 mm dB - 4 nozzles x 19 mm dB -2x2 nozzles in series [ at 40, and 80 mm apart ] x 19 mm dB -2x4 nozzles in series [ at 40, and 80 mm apart ] x 19 mm dB - 2 nozzles x 19 mm (4 cp) dB - 4 nozzles x 19 mm (4 (p) dB Each of these units was tested in the following positions and combinations: -Delivery valve: Fully open, XA open, lA open, and fully shut.

-Bleeding valves: Fully shut, fully open, lA open, and lA open.

-Nozzle position from the impeller: 65 mm.

-Direction of rotation of flow from nozzle: With the same direction of rotation as that of the impeller.

Test Observations: The following observations were derived from the calibration test for pump A: (1) Direction of rotation of flow from nozzle unit: The analysis of data for both positive (with the direction of impeller motion) and negative (against the direction of impeller motion) rotations of flow from nozzle units demonstrated an appreciable pressure detection when compared to conventional pumping. On the other hand, the Head-Discharge characteristics was found to be higher in the case of positive rotation as compared to negative rotation, affirming more stable conditions, and reduced separation at entrance of the impeller. (2)Location of nozzle units ahead of the impeller: Test results showed that better values for AH are reached with nozzle unit approaching closer to the impeller. This can be explained by the fact that, as the particles of rotating fluid get closer to the impeller, the absolute velocity, being the sum of both the axial and radial components of the rotating flow, is increased due to the influence of centrifugal forces caused by rotating blades. Consequently, creating a higher change m pressure on the suction side. (3) Shape and design dimensions for the nozzle unit: The role of the nozzle has been made easier by giving it the best shape to achieve the highest possible AH. A single 4 nozzle unit gave the best increase in AH. Using this shape the mixing process ensured a reasonable uniform flow to the impeller. This was in contrast to other shapes in which substantial losses occurred inside the nozzle channels, giving lower values of AH. Optimum dimensions recommended for the nozzle shape: dB : dN = 2 : 1 dB : L = 1 : 2 0 5 MAR 2002 received Test facilities for pump (B): Constant head water tank: Delivery from tank at suction side is through a circular opening of 100 mm diameter. A container of 150 liters is positioned at the top of the discharge tank. The container dimensions is 750 mm § x 960 mm height, made of polyethylene.

Two pipes are connecting both tanks. The pipes has the following dimensions: Diameters: (1) 96 mm Lengths: (1) 1410 mm material: mild steel (2) 150 mm (2) 1456 mm The suction line: Is 621 mm long made of mild steel with internal diameter of 103 mm and 6 mm thickness. A suction valve is mounted at a distance of 225 mm from the tank edge.

The pump: A single stage mixed flow pump. The pump has the following dimensions: r, = 30 mm, r2 = 60 mm = 90 mm, b2 =90 mm Number of vanes = 7 Model No: UniBloc, 80-169 Manufacturer: Applied Pumping Technologies LTD (APT) Serial No: J787 H (head) = 4.5 m at Q (discharge) = 20 L/s NPSH = 1.5 m The electric motor: Runs at a constant speed of 1390 r.p.m, having the following characteristics: Horse power HP = 2, Phase =3, Amperes = 0.2, Hertz Hz = 50, Voltmeter / ammeter detector reads = 230 V Kw = 1.5 Manufacturer: CEG - Italy, MOT 3IEC34 No : 0229912 The nozzle unit: Four nozzles, dB = 40 mm, dN = 20 mm, and L = 80 mm.

The discharge line: It is 4775 mm long ( for positions 500 mm and 1300 mm), and 3895 mm long (for suction lift position) Pipe is made of mild steel, 5 mm thick, and with internal diameter of 79 mm. The Jf ^ ^ i 9 ^ Oif-CtJ CT ft r tMn discharge valve is located at a distance of 395 mm from pump exit. (viii) Pressure gauge/manometer: Open tube manometers were used to measure the pressure head at the suction side.

Gauge manometers were used in measuring the pressure head at the delivery side. The gauge was placed at a distance of 650 mm from the pump exit. The gauge was ranged from 0 - 2.5 bars.

Tests were performed to study the behavior of the pump performance using bleeding system from the delivery side to suction side. Tests were performed in steady, non pulsating conditions, and almost constant temperature. Re-cords showed a negligible rise in temperature between the suction and delivery flows, in average not exceeding 0.5 °C. One effect could be the hot air discharged from the pump electric motor and directed towards the pump impeller.

Motor was operating at the service factor horsepower.

The tests involved three positioning of pump: -When level of water tank is 500 mm above the suction valve.

-When level of water tank is 1300 mm above the suction valve.

-With a suction lift of 1500 mm from centre of suction pipe.

The dimensions of the nozzle unit is: - 4 nozzles x 40 mm dB x 20 mm dN x 80 mm L The nozzle unit was tested for the following parameters: -Delivery valve: Fully open, Vi open, 'A open, and fully shut.

-Bleeding valves: Fully closed, fully open, lA open, and % open.

-Nozzle position from impeller inlet: 130 mm.

-Direction of rotation of flow from nozzle: The same direction of rotation as that of the impeller.

Test facilities for pump (C): The same test facilities used for pump (B) were also used for pump (C), with the following modifications: Constant head water tank: Delivery from tank at suction side is through a circular opening of 63 mm diameter.

The suction line: {r-uiicciuai P'C|. Oit'ce ot 0 5 MAR 2032 Is All mm long made of mild steel with internal diameter of 63 mm and 4 mm thickness. A suction valve is mounted at a distance of 225 mm from the tank edge.

The pump: A single stage centrifugal glanded pump, having the following dimensions: Manufacturer: Perfecta - D.M.Wallace LTD Type: NCP 6-125 H (head) = 3.5 m at Q (discharge) = 5 L/s NPSH = 0.75 m The electric motor: Runs at a constant speed of 1450 r.p.m, having the following characteristics: Horse power HP = 0.4, Phase =3, Amperes = 3.5, Hertz Hz = 50, Voltmeter / ammeter detector reads = 405 V Kw =0.3 Manufacturer: Perfecta - D.M.Wallace LTD No : 39277/300 4/230/400 Type: BGP6 - 125 The nozzle unit: Four nozzles, dB = 25 mm, dN = 12.5 mm, and L = 50 mm.

The nozzle unit is placed at a distance of 230 mm from pump centre.

The discharge line: It is 4385 mm long ( for positions 500 mm and 1300 mm). The pipe is made of mild steel, 4 mm thick, and with internal diameter of 63 mm. The discharge valve is located at a distance of 290 mm from pump exit.

Pressure gauge/manometer: Open tube manometers were used to measure the pressure head at the suction side.

Gauge manometers were used in measuring the pressure head at the delivery side. The gauge was placed at a distance of 440 mm from the pump exit. r, = 30 mm, r2 = 60 mm Number of vanes = 7 Model No: EV6-1601150 b, = 8 mm, b2 = 10 mm The tests involved two pump positions: -When level of water tank is 500 mm above the suction valve. 0 5 MAR 2002 -When level of water tank is 1300 mm above the suction valve.

The nozzle unit was tested for the following parameters: -Delivery valve: Fully open, XA open, lA open, and fully shut.

-Bleeding valves: Fully closed, fully open, lA open, and % open.

-Nozzle position from impeller centre: 230 mm.

-Direction of rotation of flow from nozzle: The same direction of rotation as that of the impeller. 6. Test Results and Performance Characteristics The three pumps were tested at Unitec Laboratories. Details of these pumps are: Pump ABC Inlet diameter mm 20 60 60 Flow L/S 0.5 20 5 Cent, tip mm 60 120 120 Diameter Head m 2 4.5 3.5 Figures (20-22) shows the optimum bleeding required at three different positions [ 875 mm suction lift, 557 / 105 mm suction head ] using different shapes of nozzle units. It was found that the optimum bleeding occurred in the range between BN = 0.03 - 0.04 It was also noticed that the flow tends to be unstable as the Bleeding Number values approach BN = 1.

Complete characteristic curves including the head flow, efficiency, and break-horsepower characteristics for the conventional and new arrangement of pumps: A (using best nozzle arrangement of: 4N x 12 mm dB ), B (using nozzle unit: 4N x 40 mm dB ), are shown in figures 13-18. The efficiency given is the overall pump efficiency including motor, which means a peak efficiency of 80 percent for pump A, 86 percent for pump B, as compared with conventional pumps of 70 percent, 70 percent.

The relative performance of the three pumps are shown by a non-dimensional comparison in Figure 19, where both head and flow are related to conventional pump operating (duty) points head (H0) and flow (Q0). It can be seen that pumps A+B have substantially similar characteristics. As would be expected pump (B) has a larger operating range.

Application of bleeding system using a suitable nozzle arrangement unit placed on the suction Property" Oif'co of 'w.z. 0 5 MAR 2002 side, has allowed for the first time the control of the pump characteristics from the suction side. Comparison between conventional and new pump application suggest how effective the bleeding units are. The inequalities of pressure head have now been smoothed out, the overall energy loss has in this instance been reduced, and the ideal useful lift has risen from 2.1 to 2.55 m for pump (A), from 4.4 to 5.5 m for pump (B), and from 2.27 to 3.1 m for pump (C) without increase of energy input. The pump efficiencies has also increased in average from 70 to 82.6 percent.

The pump used in test has a centrifugal impeller and single discharge volute. The pumps head (H) is 2 meters at discharge (Q) =0.8 L/s. The pump was tested in three basic levels relative to the water level inside the tank (105 mm, 557 mm suction head, and 875 mm suction lift).

Eleven different combinations of bleeding nozzle umts having diameter between 3 to 19 mm were used for testing pump (A).

Test results showed that the conventional pump efficiency was 70 percent when operating under rated conditions, and require 0.26 hp to drive it. The new pump, at optimum bleeding, gave an efficiency of 80 percent and requires only 0.22 hp. That is to say, there is a 10 percent increase in overall efficiency with a 16 percent reduction in power.

The new pump rated capacity was found to have a value of 0.98 L/s at the same pumping conditions of the conventional pump which is rated at 0.73 L/s. It would take the new pump 2.8 hrs to fill a tank of 10x103 liters, and 3.8 hrs to fill the same tank using the conventional pump, a saving of one hour.

The maximum value for the bleed causing the rotation ahead the impeller on the suction side may be determined by the dimension less group referred to herein as the Bleed number BN: Qd _NACd„ N Qw Qw N The basic assumption of the dimensionless group (Bleeding Number) is: BN = NACd/Qw --(1) Where, N = Number of nozzles used.

Greater ranges of control is expected when compared with effect of rotating action alone.

Figure 24 represents trends of having both forced rotation and a natural pre-rotation ahead of the impeller. Both the forced and pre-rotation curves are best determined ' iJsil Pfopsity 0:f»eo ot i\s.z. 0 5 MAR 2002 experimentally.

Theoretically, pressure due to rotation could, in approximation, be calculated for radial pumps as follows: Apt = P ( u2 C2u - u, Clu) —(2) and for axial pumps: Apt = pu( C2u- Clu) --(3) where, u, = u2 = u and, the circumferential velocity Cul could, in approximation, be calculated from the principle of momentum as follows: pQ„C(1R4 = pj2jtRdrC..C11R -"(4) Taking distribution of velocity ahead an axial impeller to follow, in approximation, the free vortex principle: R Cu = constant —(5) Similarly, for radial impeller: Rj Clu = constant —(6) Substituting, we get for axial impeller: Qd Cd Rd = Cu R { 2 7t R dr Ca = Qw Cu R Where, Rd = distance from axis of nozzle to axis of impeller.

Similarly, for radial impeller we get: Qd Qi Rd ~ Qw R Clu Clu: is the circumferential velocity which should be checked on experimental basis.

To design a new type of pump we need to know, with aid of practical measurements, values for the following: 1. Circumferential velocity C2u using the relation: Pc-Pb-Pm = ttMk = G)pQwRsC2u Where, Pb= Power lost due to secondary flow inside the fluid Pm = Mechanical power loss Rs = Vi (Rin2 - Rex2) C2u: means the increase in circumferential velocity with raduis R dividing stream volume entering an axial impeller. For radial impeller we have, Rs = R2 and Qw = Q + Qd 2. Clu: The circumferential (whirl) velocity at entrance for axial impeller calculated on radius R dividing the stream volume, which can be determined from the relation: i;., : 05 MAR 2002 QwR1Clu = QdCdRd 3. Cd: exit velocity from nozzle, influenced by value of losses taking place inside the control valves, can be determined from the relation: Cd = qW2 Ap / p Where, Ap = pa-p, A more accurate relation, where the value Ap is decreased by an amount Ap2 equivalent to loss in pressure inside the control valves and coefficient of flow <p2 is suggested from actual values in practice, we then get the relation: Cd = q>2 V2/p (Ap - Ap2) The ratio of cp / cp2 is an indication of the control valve system performance. 4. Influence of rotation on centrifugal pumps with a whirl ahead of the impeller can be found by applying the relation: ("a C2U - ui Clu) P = Apt Knowing the value of C]u the theoretical increase m pressure (rotation) can be easily calculated. With the aid of velocity triangles, losses and useful pressures can be determined. frHjv;oru l&i Prooarty Oifice of 0 5 MAR 2002 RECEIVED

Claims (24)

WHAT WE CLAIM IS -25 -

1. A pumping system comprising means defining a pump having an inlet and an outlet, an inlet conduit connected to said inlet of said pump for the inlet flow of fluid to said pump a delivery conduit connected to said outlet of said pump for the outlet flow of fluid from said pump means adapted to bleed at least part of the outlet flow (hereinafter the "bleed flow"), a nozzle unit engaged to said inlet conduit and positioned upstream of said inlet of said pump, means to deliver said bleed flow to said nozzle unit, said nozzle unit including at least one means to inject the bleed flow into the inlet conduit, means capable of increasing the total velocity head of bleed flow to, in use, inject said bleed flow into said inlet flow to said pump at a higher total velocity head than the total velocity head of bleed flow at said means adapted to bleed.

2. A pumping system as claimed in claim 1 wherein said means to deliver is a conduit.

3. A pumping system as claimed in claims 1 or 2 wherein said means to inject is a nozzle.

4. A pumping system as claimed in anyone of claims 2 or 3 wherein said means capable of increasing velocity head is provided by a reduction of total cross-sectional flow area of said means to deliver to thereby increase the velocity pressure head of said bleed fluid flow prior to being injected into said inlet conduit.

5. A pumping system as claimed in claim 3 wherein said means capable of increasing velocity head is provided at said nozzle in the form of a reduced total cross-sectional flow area for bleed flow through said nozzle.

6. A pumping system as claimed in anyone of claims 3 to 5 wherein said nozzle comprises a passage for fluid with an inlet for bleed flow to be received from said means adapted to bleed (via conduit) and an outlet connected for bleed flow delivery to said inlet conduit, wherein said inlet of said nozzle is of greater cross sectional area that said outlet of said nozzle.

7. A pumping system as claimed in claim 6 wherein said passage of said nozzle is gradually tapered over the length of said passage between said nozzle inlet and said nozzle 0 5 m 2032 -26- outlet such that the cross sectional flow area of said outlet is smaller that the cross sectional flow area of said inlet of said nozzle.

8. A pumping system as claimed in claims 6 or 7 wherein the ratio of said cross sectional flow area of said nozzle inlet to the cross sectional area of said to said nozzle outlet is substantially four.

9. A pumping system as claimed in claim anyone of claims 6 to 8 wherein said passage of said nozzle is substantially circular in cross sectional area and the ratio of inlet diameter to outlet diameter is substantially two.

10. A pumping system as claimed in claim 9 wherein the ratio of the length of said passage to the diameter of said nozzle inlet is substantially two.

11. A pumping system as claimed in anyone of claims 3 to 10 wherein said nozzle is engaged with said inlet conduit to, in use inject bleed flow fluid into the inlet conduit through said nozzle outlet at substantially right angles to the main flow direction of fluid in said inlet conduit.

12. A pumping system as claimed in anyone of claims 3 to 11 wherein said nozzle is engaged with said inlet conduit to, in use inject bleed fluid into the inlet conduit through said nozzle outlet to induce a rotational flow to inlet flow of fluid approaching said pump.

13. A pumping system as claimed in anyone of claims 3 to 12 wherein there are two and preferably four nozzles for injecting fluid into said inlet conduit.

14. A pumping system as claimed in claim 13 wherein said nozzles are each positioned the same distance upstream of said pump.

15. A pumping system as claimed in claims 13 or 14 wherein said nozzles are equi-spaced about the perimeter of said inlet conduit.

16. A pumping system as claimed in anyone of claims 1 to 15 wherein said pump is a centrifugal pump.

17. A pumping system as claimed in claim 12 wherein said pump is a centrifugal pump and said rotational flow is co-rotatory with the rotational direction of the pump impeller.

18. A nozzle unit for injecting fluid into a main inlet flow stream of a pump comprising; a conduit section capable of insertion into a main inlet conduit defining the flow boundary to said inlet flow stream of said pump said conduit section comprising an inlet, an outlet a wall defining region between said inlet and outlet to, in use, provide a substantial continuation of the boundary to the main inlet flow | iruL"'.;crtuai P*oport\ Oiiico of N.Z a r uin -27- stream of fluid, and at least one aperture through said wall defining region an injection nozzle for or defined by said at least one aperture to in use deliver a fluid bled from the outlet flow of said pump into said inlet flow stream, said injection nozzle being a conduit having an inlet and an outlet said outlet at said wall defining region wherein

19. A nozzle unit as claimed in claim 18 wherein said conduit of said nozzle is secured to said conduit section such that in use the flow path of injected fluid at said outlet to said conduit is at substantially right angles to the main inlet flow stream of a said pump.

20. A nozzle unit as claimed in claim 18 wherein said wall defining region of said conduit section is defined by a substantially constant diameter bore corresponding to the diameter of said main inlet conduit..

21. A nozzle unit as claimed in anyone of claims 18 to 20 wherein said conduit of said nozzle is secured to said conduit section such that in use the flow path of injected fluid at said outlet to said conduit is at substantially right angles to the main inlet flow stream of a said pump and at a tangent to said boundary of said main inlet conduit.

22. A nozzle unit as claimed in anyone of claims 18 to 21 wherein said inlet of the conduit of said nozzle is of a flow area substantially four times that of the flow area of the outlet to said conduit.

23. A nozzle unit as claimed in anyone of claims 18 to 22 wherein said conduit of said nozzle is in cross section substantially circular.

24. A nozzle unit as claimed in claim 23 wherein said conduit is gradually tapered between said inlet and said outlet over a distance substantially twice that of the diameter of said inlet to said conduit. the flow area of said inlet to said conduit is larger than said outlet to said conduit. H \hbrary\patents\djj\ahb\428548 wpd 0 5 MAR 2002 r p r f i v f n