I wanted the forum's opinion on this one. I've been digging into the differences between a soft start vs a VFD, when it comes to starting a motor. From my understanding a soft start limits the output voltage going to the motor, which then significantly reduces torque. While on variable torque loads this may be fine, but on constant torque loads this can pose a challenge, because more torque is needed. A VFD varies the output voltage going to the motor, as well as the frequency during start-up, which allows reduced inrush current while still providing more torque. We're installing centrifuges which we don't plan on varying the speed on once it reaches full RPM. Historically, I've used soft-starts for anything I'm not wanting to vary the speed on. On some centrifuges I recall that the start times were pretty high (2x-3x FLA for 5-7 minutes, or something similar). I would think this kind of start profile is harmful to motors. Would a VFD be a better option to help start constant torque loads such as centrifuges? From everything I'm gathering the inrush current during a VFD ramp-up hardly goes above full-load amps, is this correct?
VFD vs Soft Start
- Thread starter adamscb
- Start date
- Status
Russs57
Senior Member
- Location
- Miami, Florida, USA
- Occupation
- Maintenance Engineer
It does sound like the VFD would be the better solution. The following might be worth a read.
- Location
- San Francisco Bay Area, CA, USA
- Occupation
- Electrical Engineer
TYPICALLY, centrifuge motors are selected KNOWING the starting duty issue; i.e. the motor nameplate HP is actually lower than what it really would be for another application, or the centrifuge mfr puts on a standard motor that is 2-3x what it needs to be for the continuous load. So before making your decision, you should check with the centrifuge mfr. I have started many many centrifuges with Soft Starters and although I have had to over size the soft starters as if the starting current were the continuous current, I have never had an issue with the motors.
That said, the VFD is a great choice if you don't know one way or the other, it will just cost more and varying the speed may be pointless. The VFD can be programmed to Current Limit at the motor nameplate FLA and it will hold there however long it takes to accelerate the centrifuge with just 100% of the motor FLT. Mind you, that can be a lot longer than you might think... I did a 500HP gold separator centrifuge once with a VFD (because they wanted braking too), at100% FLA it took almost an hour for it to get to full speed.
And as to braking, that's sometimes a benefit worth looking into. If this is some sort of batch operation where you load a batch into it, run it, then empty and start over, having a "regenerative" drive can increase your throughput by cutting down the wait time between batches.
That said, the VFD is a great choice if you don't know one way or the other, it will just cost more and varying the speed may be pointless. The VFD can be programmed to Current Limit at the motor nameplate FLA and it will hold there however long it takes to accelerate the centrifuge with just 100% of the motor FLT. Mind you, that can be a lot longer than you might think... I did a 500HP gold separator centrifuge once with a VFD (because they wanted braking too), at100% FLA it took almost an hour for it to get to full speed.
And as to braking, that's sometimes a benefit worth looking into. If this is some sort of batch operation where you load a batch into it, run it, then empty and start over, having a "regenerative" drive can increase your throughput by cutting down the wait time between batches.
Is it true let's say for the centrifuge I mentioned in my example - if you were to put a VFD on it, it would take less than the 5-7 minutes to get up to full speed, while drawing less than FLA, because the VFD can supply 100% torque?
I just redid a centrifuge drive a couple years ago and I do soft starts all the time.
The thing to understand is that the idea of “variable” and “constant” torque grossly oversimplifies things. For instance a conveyor which is “constant torque” pulls about 140% of its running torque until it accelerates to speed then barely changes torque even with massive load changes after that…most of the torque is overcoming idler friction. Pumps and fans have roughly a quadratic curve (torque is proportional to the square of speed).
In terms of technologies yes soft starts cut the voltage to control starting. Generally you can choose voltage ramping, torque limiting, or current limiting. With a “constant torque” load such as a conveyor usually you use constant current (current limiting) mode and limit it to say 400-500% current. How high you set it is based on starting tone. Increase/decrease as needed to get reasonable start times but “reasonable” is certainly under 39 seconds and preferably under 15-2@ seconds. The other modes are similar…just use a different controlled variable. With centrifugal fans and pumps we might be able yo start at 250% of FLA using current limiting but might just use voltage ramping for a smoother startup when starting times are short. With a soft start we need to keep starting times short of the motor overheats. If you look at motor torque/speed curves effectively a soft start just reduces the curve.
In contrast with a VFD you run the motor on the mostly linear torque/speed curve that you see all the way to the right beyond the peak torque point. By varying frequency this region can be moved to almost any speed.
With a centrifuge you have a massive amount of inertia. You could certainly accelerate in 30 seconds but you’d need a huge motor that would be almost idle the rest of the time. But that would be wasteful. Better would be to apply a much smaller accelerating torque over several minutes so the motor is reasonably sized. This is the realm of the VFD.
At speed there are a couple things to consider.The first is in a continuous scroll centrifuge you also have the scroll or “back” drive. This runs almost the entire tune as a brake so the scroll runs slower than the bowl speed. You can honestly just connect the DC bus of the bowl and back drives and just let it regenerate all that energy back onto the DC bus and feed the bowl drive. This is highly efficient,
The other consideration is if you don’t have a back drive the bowl drive is about 95% efficient. To capture this you can run the centrifuge with a bypass contactor. Accelerate with the VFD. Once up to speed simply close the bypass contactor snd this eliminates the drive losses. Since it becomes AC only if there is a back drive the drive needs to be able to regenerate AC because the bowl VFD DC bus is shut down.
So yes in theory you can use a soft start but realistically the motor would be grossly oversized. I’ve never seen a non-VFD drive for a reason.
The thing to understand is that the idea of “variable” and “constant” torque grossly oversimplifies things. For instance a conveyor which is “constant torque” pulls about 140% of its running torque until it accelerates to speed then barely changes torque even with massive load changes after that…most of the torque is overcoming idler friction. Pumps and fans have roughly a quadratic curve (torque is proportional to the square of speed).
In terms of technologies yes soft starts cut the voltage to control starting. Generally you can choose voltage ramping, torque limiting, or current limiting. With a “constant torque” load such as a conveyor usually you use constant current (current limiting) mode and limit it to say 400-500% current. How high you set it is based on starting tone. Increase/decrease as needed to get reasonable start times but “reasonable” is certainly under 39 seconds and preferably under 15-2@ seconds. The other modes are similar…just use a different controlled variable. With centrifugal fans and pumps we might be able yo start at 250% of FLA using current limiting but might just use voltage ramping for a smoother startup when starting times are short. With a soft start we need to keep starting times short of the motor overheats. If you look at motor torque/speed curves effectively a soft start just reduces the curve.
In contrast with a VFD you run the motor on the mostly linear torque/speed curve that you see all the way to the right beyond the peak torque point. By varying frequency this region can be moved to almost any speed.
With a centrifuge you have a massive amount of inertia. You could certainly accelerate in 30 seconds but you’d need a huge motor that would be almost idle the rest of the time. But that would be wasteful. Better would be to apply a much smaller accelerating torque over several minutes so the motor is reasonably sized. This is the realm of the VFD.
At speed there are a couple things to consider.The first is in a continuous scroll centrifuge you also have the scroll or “back” drive. This runs almost the entire tune as a brake so the scroll runs slower than the bowl speed. You can honestly just connect the DC bus of the bowl and back drives and just let it regenerate all that energy back onto the DC bus and feed the bowl drive. This is highly efficient,
The other consideration is if you don’t have a back drive the bowl drive is about 95% efficient. To capture this you can run the centrifuge with a bypass contactor. Accelerate with the VFD. Once up to speed simply close the bypass contactor snd this eliminates the drive losses. Since it becomes AC only if there is a back drive the drive needs to be able to regenerate AC because the bowl VFD DC bus is shut down.
So yes in theory you can use a soft start but realistically the motor would be grossly oversized. I’ve never seen a non-VFD drive for a reason.
GeorgeB
ElectroHydraulics engineer (retired)
- Location
- Greenville SC
- Occupation
- Retired
- Status