Frequently Asked Question-28
- By Jayesh Patel
- Published 01/31/2010
Tech Quest
API 610 recommends Minimum size requirement for casing connections other than suction & discharge nozzles based on Discharge nozzle size excluding seal flushing & gauge conections. Why?
What I appreciate most about API Standards, is that these standards are evolved by an organisation of pump-users, with little influence of pump-manufacturers.
Size of casing drain connection is important to ensure that the casing can be drained as fast as possible, saving downtime during maintenance. Size of pressure-gauge connection is important because that automatically dictates dial-size of the gauges and in turn their accuracy and readability.
Once the significance of these sizes is also appreciated, one understands and appreciates the standard also better
What role would moment of inertia of the motor play in damaging, rather, shearing the power screw of a triple screw pump? Salient details are -
1. Pumped liquid - lubricating oil
2. Capacity 160 lpm
3. Rated pressure - 180 bar
4. Drive - 60 kW, 2900 rpm
5. Mounting - Vertically into a 36 m3 oil tank. Minimum submergence recommended to avoid vortex-formation, air entrainment, dry running is 40 mm. The recommendation is not always followed.
6. Constructional features - There is no packed gland or mechanical seal. Pump shaft or the power screw has a balance piston on the driver side, with close clearance. between balance piston and pump casing. Leakage if any, maximum 0.5 lpm, would flow over the pump and drip back into the oil tank.
7. Material strength and diameter of power screw -24mm, 16CrMnS5, with minimum UTS of 570 MPa
To get a clarity for myself, on the role of moment of inertia of the motor, I would paraphrase the question a little differently. "Though a pump requires only 60 kW rnotor, would a 100 kW motor, inadvertently or even wantonly connected, cause damage, rather shearing of the pump shaft?" My logical answer would be "No".
A 100 kW motor would have much higher moment of inertia than a 60 kW motor. If the shaft has suffered seizure, a 100 kW motor has rather, a better chance of overcoming the seizure and make it run than a 60 kW motor.
The root cause of the failure seems to be the seizure of the shaft than excessive moment or torque imparted by the motor.
Standards for motors specify pull-up or starting or locked rotor or breakaway torque to be 150 percent of full-load torque. This testing is typically done by locking the rotor. This actually simulates a seized shaft and demonstrates capacity of the motor to overcome seizure. A motor does try to overcome a seizure. Excessive torque may get imparted in this effort of the motor to overcome the seizure. But the demand for excessive torque comes from a seized shaft. If the shaft is not seized and does not demand excessive torque, the motor will not impart, by its volition, any torque more than what the driven shaft demands. If the driven shaft demands only 10 kw load from a 60 kW motor, the motor would provide just as much. That is what is called as part-load running of the motor. Efficiency of the motor would of course be poor in such part-load running. Motors have power demand of their own, even when running on no load or zero load. So, motors can run all the way from zero load or no load to full load and somewhat beyond full load, which is overload. Basic fact is that motors respond to the demand. They do not impose load on to the driven equipment. They impose load on to the supply system, not on the driven system.
Root causes of the shear of a driven shaft would be misalignment, thermal load, seizure. Possible causes for a shaft of a screw pump to suffer seizure would be dry running.
Incidentally, if both the power screw and the idler screw have same metallurgy and area of course, running in frictional contact, they are susceptible to suffer electrolytic galling and consequently a seizure, more so in dry running.
To prevent dry running in the given installation, it seems that a level controlled interface with the motor's starter would be a good protection. The other check should be on using dissimilar metallurgies to avoid electrolytic galling at surfaces in frictional contact.
Bearings in Pumps
We are a well established company in the production and distribution of various kind of Ball & Roller bearings. Would highly appreciate if could let us know which bearings are used in Pumps production. What is their application and quantity per Pump. We await your reply per return.
Bearings used in pumps are of various types. To list -
1. Monobloc pumps would have no separate pump-bearings, because the pump assembly is on the extended shaft of the motor. So, bearings of motor serve a also as pump-bearings.
2. Pumps coupled to the driver through a coupling will have pump bearings. Number of bearings varies depending upon application.
For economical pumps as for agricultural purposes, the pump may have only one anti-friction bearing, the hypothesis being that the throat bush in the stuffing box also acts as a bearing support.
Industrial pumps which have to often run round-the-clock would have a distinct bearing housing with 2 bearings, one at the driving end, i.e. near to the driver and the other nearer to the pump. If axial thrust in the pump is estimated to be significant, e.g. with semi-open impellers, the pump would have an anti-friction thrust bearing, often of the angular contact type and in matched pair.
Large vertical turbine pumps would have tilting pad (Mitchell or Kingsbury type) thrust bearings.
For very large horizontal pumps where anti-friction bearings of large shaft dia are not available from regular product ranges, people may use journals with splash ring etc. Axially split casing type pumps and multi-stage pumps are also called as "between bearings" pumps, meaning the pumps would have two bearings at the two ends.
For bore well submersible pumps with water-filled wet motors, the pump assembly has mainly stage bushes and the motor also has bush bearings, because the motor is filled with water. Oil-filled motors would have anti-friction bearings.
Helical rotor progressive cavity pumps would often have only one bearing, because the pump shaft needs to drive the rotor through a universal joint.
Twin screw pumps may have as many as four bearings, two on each shaft/screw.
A triple screw pump however may have only two bearings, since there is only one driving screw and other screws run as idlers.
So a variety of logic for number of bearings and types of bearings in a pump.
I want to know the steps taken in a pumping system to attain the desired operating point for the system when
(a) the pump is driven by a fixed speed motor and,
(b) when the pump is driven by either a variable speed motor or a turbine.
The operating point is the point of intersection between the H-Q curve of the pump with the H-Q curve of the system. Once a pump is set into a system, this will happen automatically. But if the operating point, which happens automatically is not the 'desired' operating point, one has to modify either the pump curve or the system curve.
There are two ways to modify the pump curve -
1) Change the speed of the pump
2) Change the diameter of the impeller of the pump
3) The system curve can be notified by modifying the system. This is usually done either by changing the setting of the delivery valve or one can change it also by revamping the system by changing the pipe-sizes and/or layout of the piping.
4) If the suction conditions in the system are prone to cause the pump to cavitate, modifying the system to eliminate cavitation will also modify the pump curve from a cavitating condition to non-cavitating condition.
5) For changing the speed of the pump (option 1 above), changing the driver from an electric motor to a turbine will often become changing from a low-speed driver to high-speed driver.
Such change is possible even by using a gearing or pulley mechanism between the pump and the motor. But at increased speed the pump demands higher power input. So, it becomes important to check whether the motor has adequate margin in power. No such caution is needed if "desired" operating point is obtainable by reducing the speed.
For determining the required speed at the "desired" operating point, say
(Q",H")one needs to find the point (Qo, Ho) on the pump curve H = a*Q^2+b*Q+c which also is a point on the parabola through the origin and (Q", H"). The equation of this parabola will be H = k*Q 2, where k = H"/(Q")^2.
Since (Qo, Ho) is to be a point both on
H = k*Q^2 and H, a*Q^2+b*Q+c
to find (Qo, Ho) one needs to solve the quadratic (a-k)*(Qo)^2+b*Qo+c = 0
Actually all the mathematics starts with knowing the values of the co-efficients a, b, c for the pump curve H = a*Q^2+b*Q+c This is not difficult, if one knows three points on the curve, say,(0, Hso), (Q1, H1) and (Q2,H2) and solves simultaneous equations. A simpler way to do this is to plot the pump curve in an Excel spreadsheet and fit a 'trendline', setting also the option for the display of the equation of the polymonial of degree 2.
MOC for Abrasion & Corrosion
I would like to know the diffirence between abrasion and corrosion. What type of M.O.C for impeller and shaft is suitable for abrasion and corrosion?
Also let me know the selection parameters for Impeller and shaft M.O.C.
One commonplace example of understanding the difference is water laden with sand. Sand, per se, is not corrosive, but it is very abrasive. Conversely acid with no entrained solids, clear acid, will not be abrasive but highly corrosive. Sea water will also be corrosive. But corrosion due to sea water is due to its alkalinity whereas corrosiveness of acids is acidic in nature. MOC for corrosion resistance has to take into consideration whether the corrosiveness is acidic or alkaline.
Abrasion is also of two types. Abrasion due to fly ash in power stations will be from fine particles moving too close to the surfaces and abrading the surfaces. Abrasion due to sand particles or coal particles will be due to the particles hitting hard on the surface and bouncing back and hitting repeatedly. This is rather erosion than abrasion. So nature of abrasive wear depends upon the angle of incidence of the particles w.r.t. the surface. Usually hard surfaces would take abrasive wear better and resilient surfaces such as elastomer-linings would take the erosive wear better. But this is too much of a thumb rule. One needs to study the wear patterns and refer to the data available in handbooks.
How to determine Minimum stable continuous flow & Minimum thermal continuous flow. Is there any standard which explains about these parameters & the determination in detail
Minimum stable continuous flow is to be read on such H-Q curve which is unstable. In unstable characteristics, Hmax is greater than Hso (Head at shut-off). In such case, Minimum stable continuous flow will be where a horizontal thru' Hso will intersect the H-Q curve of the pump.
Minimal thermal continuous flow is that flow, when the liquid will experience churning caused by internal re-circulation. This happens because, the cross-sections of the hydraulic passages prove to be too large for the amount of flow to be carried. The designer designs the passages ideally for the design flow. At flows less than the dsign flow, the passages are not ideal. This is also one reason for the drop in efficiency at flows different from design flow. The effect becomes accentuated at flows less than Minimum thermal continuous flow. Churning of the liquid causes temperature of the liquid to also rise. This in turn raises the vapour pressure of the liquid. In turn the available NPSH gets affected. By all these considerations the curve for NPSHr v/s a becomes uncertain. So, manufacturers show NPSHr curve only ahead from Minimum thermal continuous flow.
Obviously both Minimum stable continuous flow and Minimum thermal continuous flow are to be recommended by the manufacturer and cannot be obtained from standards.
In API-610 one finds a mentibon of continuously rising characteristics, that means a stable characteristics, i.e. where Hmax is only at shut-off. To be more mathematically correct, for a stable characteristics, the point of maxima is not in the first quadrant.
API 610 recommends Minimum size requirement for casing connections other than suction & discharge nozzles based on Discharge nozzle size excluding seal flushing & gauge conections. Why?
What I appreciate most about API Standards, is that these standards are evolved by an organisation of pump-users, with little influence of pump-manufacturers.
Size of casing drain connection is important to ensure that the casing can be drained as fast as possible, saving downtime during maintenance. Size of pressure-gauge connection is important because that automatically dictates dial-size of the gauges and in turn their accuracy and readability.
Once the significance of these sizes is also appreciated, one understands and appreciates the standard also better
What role would moment of inertia of the motor play in damaging, rather, shearing the power screw of a triple screw pump? Salient details are -
1. Pumped liquid - lubricating oil
2. Capacity 160 lpm
3. Rated pressure - 180 bar
4. Drive - 60 kW, 2900 rpm
5. Mounting - Vertically into a 36 m3 oil tank. Minimum submergence recommended to avoid vortex-formation, air entrainment, dry running is 40 mm. The recommendation is not always followed.
6. Constructional features - There is no packed gland or mechanical seal. Pump shaft or the power screw has a balance piston on the driver side, with close clearance. between balance piston and pump casing. Leakage if any, maximum 0.5 lpm, would flow over the pump and drip back into the oil tank.
7. Material strength and diameter of power screw -24mm, 16CrMnS5, with minimum UTS of 570 MPa
To get a clarity for myself, on the role of moment of inertia of the motor, I would paraphrase the question a little differently. "Though a pump requires only 60 kW rnotor, would a 100 kW motor, inadvertently or even wantonly connected, cause damage, rather shearing of the pump shaft?" My logical answer would be "No".
A 100 kW motor would have much higher moment of inertia than a 60 kW motor. If the shaft has suffered seizure, a 100 kW motor has rather, a better chance of overcoming the seizure and make it run than a 60 kW motor.
The root cause of the failure seems to be the seizure of the shaft than excessive moment or torque imparted by the motor.
Standards for motors specify pull-up or starting or locked rotor or breakaway torque to be 150 percent of full-load torque. This testing is typically done by locking the rotor. This actually simulates a seized shaft and demonstrates capacity of the motor to overcome seizure. A motor does try to overcome a seizure. Excessive torque may get imparted in this effort of the motor to overcome the seizure. But the demand for excessive torque comes from a seized shaft. If the shaft is not seized and does not demand excessive torque, the motor will not impart, by its volition, any torque more than what the driven shaft demands. If the driven shaft demands only 10 kw load from a 60 kW motor, the motor would provide just as much. That is what is called as part-load running of the motor. Efficiency of the motor would of course be poor in such part-load running. Motors have power demand of their own, even when running on no load or zero load. So, motors can run all the way from zero load or no load to full load and somewhat beyond full load, which is overload. Basic fact is that motors respond to the demand. They do not impose load on to the driven equipment. They impose load on to the supply system, not on the driven system.
Root causes of the shear of a driven shaft would be misalignment, thermal load, seizure. Possible causes for a shaft of a screw pump to suffer seizure would be dry running.
Incidentally, if both the power screw and the idler screw have same metallurgy and area of course, running in frictional contact, they are susceptible to suffer electrolytic galling and consequently a seizure, more so in dry running.
To prevent dry running in the given installation, it seems that a level controlled interface with the motor's starter would be a good protection. The other check should be on using dissimilar metallurgies to avoid electrolytic galling at surfaces in frictional contact.
Bearings in Pumps
We are a well established company in the production and distribution of various kind of Ball & Roller bearings. Would highly appreciate if could let us know which bearings are used in Pumps production. What is their application and quantity per Pump. We await your reply per return.
Bearings used in pumps are of various types. To list -
1. Monobloc pumps would have no separate pump-bearings, because the pump assembly is on the extended shaft of the motor. So, bearings of motor serve a also as pump-bearings.
2. Pumps coupled to the driver through a coupling will have pump bearings. Number of bearings varies depending upon application.
For economical pumps as for agricultural purposes, the pump may have only one anti-friction bearing, the hypothesis being that the throat bush in the stuffing box also acts as a bearing support.
Industrial pumps which have to often run round-the-clock would have a distinct bearing housing with 2 bearings, one at the driving end, i.e. near to the driver and the other nearer to the pump. If axial thrust in the pump is estimated to be significant, e.g. with semi-open impellers, the pump would have an anti-friction thrust bearing, often of the angular contact type and in matched pair.
Large vertical turbine pumps would have tilting pad (Mitchell or Kingsbury type) thrust bearings.
For very large horizontal pumps where anti-friction bearings of large shaft dia are not available from regular product ranges, people may use journals with splash ring etc. Axially split casing type pumps and multi-stage pumps are also called as "between bearings" pumps, meaning the pumps would have two bearings at the two ends.
For bore well submersible pumps with water-filled wet motors, the pump assembly has mainly stage bushes and the motor also has bush bearings, because the motor is filled with water. Oil-filled motors would have anti-friction bearings.
Helical rotor progressive cavity pumps would often have only one bearing, because the pump shaft needs to drive the rotor through a universal joint.
Twin screw pumps may have as many as four bearings, two on each shaft/screw.
A triple screw pump however may have only two bearings, since there is only one driving screw and other screws run as idlers.
So a variety of logic for number of bearings and types of bearings in a pump.
I want to know the steps taken in a pumping system to attain the desired operating point for the system when
(a) the pump is driven by a fixed speed motor and,
(b) when the pump is driven by either a variable speed motor or a turbine.
The operating point is the point of intersection between the H-Q curve of the pump with the H-Q curve of the system. Once a pump is set into a system, this will happen automatically. But if the operating point, which happens automatically is not the 'desired' operating point, one has to modify either the pump curve or the system curve.
There are two ways to modify the pump curve -
1) Change the speed of the pump
2) Change the diameter of the impeller of the pump
3) The system curve can be notified by modifying the system. This is usually done either by changing the setting of the delivery valve or one can change it also by revamping the system by changing the pipe-sizes and/or layout of the piping.
4) If the suction conditions in the system are prone to cause the pump to cavitate, modifying the system to eliminate cavitation will also modify the pump curve from a cavitating condition to non-cavitating condition.
5) For changing the speed of the pump (option 1 above), changing the driver from an electric motor to a turbine will often become changing from a low-speed driver to high-speed driver.
Such change is possible even by using a gearing or pulley mechanism between the pump and the motor. But at increased speed the pump demands higher power input. So, it becomes important to check whether the motor has adequate margin in power. No such caution is needed if "desired" operating point is obtainable by reducing the speed.
For determining the required speed at the "desired" operating point, say
(Q",H")one needs to find the point (Qo, Ho) on the pump curve H = a*Q^2+b*Q+c which also is a point on the parabola through the origin and (Q", H"). The equation of this parabola will be H = k*Q 2, where k = H"/(Q")^2.
Since (Qo, Ho) is to be a point both on
H = k*Q^2 and H, a*Q^2+b*Q+c
to find (Qo, Ho) one needs to solve the quadratic (a-k)*(Qo)^2+b*Qo+c = 0
Actually all the mathematics starts with knowing the values of the co-efficients a, b, c for the pump curve H = a*Q^2+b*Q+c This is not difficult, if one knows three points on the curve, say,(0, Hso), (Q1, H1) and (Q2,H2) and solves simultaneous equations. A simpler way to do this is to plot the pump curve in an Excel spreadsheet and fit a 'trendline', setting also the option for the display of the equation of the polymonial of degree 2.
MOC for Abrasion & Corrosion
I would like to know the diffirence between abrasion and corrosion. What type of M.O.C for impeller and shaft is suitable for abrasion and corrosion?
Also let me know the selection parameters for Impeller and shaft M.O.C.
One commonplace example of understanding the difference is water laden with sand. Sand, per se, is not corrosive, but it is very abrasive. Conversely acid with no entrained solids, clear acid, will not be abrasive but highly corrosive. Sea water will also be corrosive. But corrosion due to sea water is due to its alkalinity whereas corrosiveness of acids is acidic in nature. MOC for corrosion resistance has to take into consideration whether the corrosiveness is acidic or alkaline.
Abrasion is also of two types. Abrasion due to fly ash in power stations will be from fine particles moving too close to the surfaces and abrading the surfaces. Abrasion due to sand particles or coal particles will be due to the particles hitting hard on the surface and bouncing back and hitting repeatedly. This is rather erosion than abrasion. So nature of abrasive wear depends upon the angle of incidence of the particles w.r.t. the surface. Usually hard surfaces would take abrasive wear better and resilient surfaces such as elastomer-linings would take the erosive wear better. But this is too much of a thumb rule. One needs to study the wear patterns and refer to the data available in handbooks.
How to determine Minimum stable continuous flow & Minimum thermal continuous flow. Is there any standard which explains about these parameters & the determination in detail
Minimum stable continuous flow is to be read on such H-Q curve which is unstable. In unstable characteristics, Hmax is greater than Hso (Head at shut-off). In such case, Minimum stable continuous flow will be where a horizontal thru' Hso will intersect the H-Q curve of the pump.
Minimal thermal continuous flow is that flow, when the liquid will experience churning caused by internal re-circulation. This happens because, the cross-sections of the hydraulic passages prove to be too large for the amount of flow to be carried. The designer designs the passages ideally for the design flow. At flows less than the dsign flow, the passages are not ideal. This is also one reason for the drop in efficiency at flows different from design flow. The effect becomes accentuated at flows less than Minimum thermal continuous flow. Churning of the liquid causes temperature of the liquid to also rise. This in turn raises the vapour pressure of the liquid. In turn the available NPSH gets affected. By all these considerations the curve for NPSHr v/s a becomes uncertain. So, manufacturers show NPSHr curve only ahead from Minimum thermal continuous flow.
Obviously both Minimum stable continuous flow and Minimum thermal continuous flow are to be recommended by the manufacturer and cannot be obtained from standards.
In API-610 one finds a mentibon of continuously rising characteristics, that means a stable characteristics, i.e. where Hmax is only at shut-off. To be more mathematically correct, for a stable characteristics, the point of maxima is not in the first quadrant.
