Cryogenic effects on start torque

[Cryodude]

I have been told that the avreage start time for a large pump
should be in the range of 4-5 seconds. But thus far have been unable to backup this number with any data available. Is there a standard for start times 0-1800-0-3600 or an average that is shot for. On a Direct On Line start basis.

 

[jraef]

Are you asking what the typical starting time is for any pump, or are you asking about cryogenic pumps? I ask because I have some experience with cryo pumps for LNG transfer. The problem is in the motor. Most have a severely limited rotor heating time meaning that if, on DOL statrting, they are not at full speed within 7 seconds the rotor overheats and swells, destroying the motor. Mind you that 7 second time limit was for the one that I worked on, it may be different for others. However, that was based upon the heating effects at DOL starting. We successfully used VFDs to extend the ramp time out to 15 seconds with no problems, because with a VFD the current is much much lower so the rotor does not heat up.

This isn't an issue on regular pump motors though. The acceleration time limits often have to do with pump lubrication via the fluid pressure created by the pump itself, especially verticals. If you are not at a certain speed fast enough, there is insufficient fluid pressure to lift the rotor or impeller off of the bearings, causing excessive wear. Every pump design that has this issue is going to have a different time requirement.

"Venditori de oleum-vipera non vigere excordis populi"

 

[vanstoja]

Starting time for induction motors is given by
t=Wk^2*delta RPM/308*T
where t=time(sec)
Wk^2=Rotor inertia(lb-ft^2) k=Radius of gyration
T=Accelerating torque (lb-ft)
Accelerating torque is the integrated difference between the motor torque and the load torque determined from torque vs speed curves. The load torque curve is usually a parabola for a pump load from load torque curve intersection with motor torque curve at load slip speed.
Hertl,D.W.(1963), "Get the Right Induction Motor", Power, January, pp.61-63 states a rule of thumb, "have the motor accelerate the load to full speed in 8 seconds or less. If time falls between 8 and 10 seconds, number of times the motor is started during each work period should be established with fair accuracy to determine motor heating."
US military specifications for pump motors generally specified a 5 second maximum starting time sometimes for 70 to 80% of rated voltage to account for startup voltage dips in non-stiff power systems. For cryogenic fluids pumping, the only starting issue may be high pumped fluid viscosity which might raise the load torque curve to a higher level than water pumping thus reducing the accelerating torque, T in the above equation and increasing starting time.

 

[electricpete]

I think it is well covered above.

My quick two cents (being uninformed about cryogenic applications): There is no target for start time unless it is dictated by the process requirements in very rare cases by power supplier. Usually for large motor applications (above NEMA frame) you tell the motor vendor what your pump characteristics are and other relevant starting info including minimum volts, and ask the vendor to give you a motor which can start the load safely without harming itself under worst case starting conditions. They should supply motor thermal damage curve and also motor starting curve... both current vs time. You may also specify some margin between these curves including margin required to squeeze your protection in between them.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.