Some basics of VFD install

Status
Not open for further replies.
petersonra said:
If you insist on adding a contactor it should be on the load side and interlocked with the drive enable.
Click to expand...
Yes, and that's what we did on VFDs. Not because we insisted on it - it was specified in the RFQ. Non-compliant bids would be filed in the recycling bin. At least they served that useful purpose.
 
kwired said:
I don't think opening the output circuit of the drive is as much of a problem as closing the switch when drive output is active. So aux contact to drive enable is more critical for not letting drive start unless the switch is closed.

Even with aux contact set to open slightly before the main contacts on the disconnect, if you have drive programmed to decelerate the load to a stop/near stop you are going to interrput output current while decelerating anyway.
Click to expand...

I used the term “disable” input on purpose, the difference is that it will immediately turn off the output transistors, no decel profile. Many newer drives now give you this with a jumper across the input, so the circuit must remain closed for the drive to work. Remove that jumper, wire in the aux contact and when the aux opens, there is immediately no output from the transistors.

The danger from opening a switch on the output to the transistors is not always immediate, it is more incremental. When contacts open under load, an arc forms and acts like a capacitor, charging itself up across the gap and increasing the voltage until the distance across the air gap providers enough dielectric strength to stop the current flow. During that fraction of a second, the rate of rise (dv/dt) of the voltage in that circuit will exceed what the transistors are rated to block, so it eventually breaks down the junction layers until they fail. Many drives include protective circuits to help mitigate this, allowing the use of safety contactors on the output side for occasional emergency operations. But really they just delay the inevitable if you keep doing it over and over again as the normal course of action and you will eventually kill the transistors.
 
Just to be sure I'm okay here: I don't plan to have anything between the drive and the motor.

Breaker in MCC --> Disconnect --> Drive --> Motor
[ ^line reactor somewhere here^ ]

The disconnect, drive, and motor will all be a few feet from one another. There will be no regular/routine shutting off of power to the drive. The motors are all roughly 40FLA @ 460V.

As a matter of context: This facility has been running these pump motors at constant line voltage and controlling flow rates with valves. They want now to adjust the motor speed with the drive and a flow meter input.

My continued appreciation.
 
Electromatic said:
As a matter of context: This facility has been running these pump motors at constant line voltage and controlling flow rates with valves. They want now to adjust the motor speed with the drive and a flow meter input.
Click to expand...
Energy saving is the usual reason for doing so.
 
Besoeker said:
Energy saving is the usual reason for doing so.
Click to expand...

For the same flow rate they will use about the same amount of energy.

Does not matter much if the flow rate is controlled by a valve or a VFD.

there might be some additional friction losses with the valve, but it will take a long time to make up the cost of the VFD.
 
petersonra said:
For the same flow rate they will use about the same amount of energy.

Does not matter much if the flow rate is controlled by a valve or a VFD.

there might be some additional friction losses with the valve, but it will take a long time to make up the cost of the VFD.
Click to expand...

pump does not care if loss is dynamic or static, pipe or valve or head
you are correct, total power will be the same, vfd or valve
the valve friction loss is accounted for on the curve, the vfd will also have losses

the best way is to plot on a pump curve
valve at operating point +/- 20%
vfd using affinity law at operating point +/-20%
 
petersonra said:
For the same flow rate they will use about the same amount of energy.

Does not matter much if the flow rate is controlled by a valve or a VFD.

there might be some additional friction losses with the valve, but it will take a long time to make up the cost of the VFD.
Click to expand...
The pressure drop across the valve is wasted energy.
 
Besoeker said:
The pressure drop across the valve is wasted energy.
Click to expand...

if sized properly moot

gpm = Cv sqrt(diff press)
Cv is the valve constant unique to the valve
usually given in 10% increments of stroke
size for the middle gpm and you can from 0 to 150% with accuracy
maybe 5 psi drop or 10 ft head at the operating point

in process control valves are preferred
control valve 'bible'
depends on media, temp, desired accuracy, turn down, etc
 
Besoeker said:
The pressure drop across the valve is wasted energy.
Click to expand...
Yes, but may take a while to recover cost of lost energy when compared to investment cost to install the drive and apparently pressure/flow monitoring components to feed a speed reference to the drive. That also would assume no additional maintenance is created from this.
 
assume 500 gpm 100 ft head 50% pump eff
hyd hp = 25.3 hp
use a 30 hp or 84% fl

vfd losses 3.5% or 0.88 hp
likely more since to get a control band you may run at 80-90% speed

if valve is size for 3 psi drop at 50% position and 500 gpm
7' head or 7% of the operating range

hp req for valve dp = 500 x 7 x 8.34 /(60 x 550) = 0.88 hp

as the valve moves off center point drop will increase with flow, basically offsetting
for the vfd as pump speed slows eff decreases

valve accuracy is typicaly better
pid 4-20 mA to I/P xdcr 3-15 psi to pilot positioner on pneumatic actuator

the choice valve vs vfd depends on many factors but usually favors a valve
 
Last edited:
Ingenieur said:
assume 500 gpm 100 ft head 50% pump eff
hyd hp = 25.3 hp
use a 30 hp or 84% fl

vfd losses 3.5% or 0.88 hp
likely more since to get a control band you may run at 80-90% speed

if valve is size for 3 psi drop at 50% position and 500 gpm
7' head or 7% of the operating range

hp req for valve dp = 500 x 7 x 8.34 /(60 x 550) = 0.88 hp

as the valve moves off center point drop will increase with flow, basically offsetting
for the vfd as pump speed slows eff decreases

valve accuracy is typicaly better
pid 4-20 mA to I/P xdcr 3-15 psi to pilot positioner on pneumatic actuator

the choice valve vs vfd depends on many factors but usually favors a valve
Click to expand...

cost wise these days a VFD is often more cost effective than a good control valve. a control valve might provide slightly better control but it almost never really matters. close enough is good enough most of the time.

having said all that, I have run into cases where they put a VFD on the pump and install a control valve too. too much money I guess.
 
petersonra said:
cost wise these days a VFD is often more cost effective than a good control valve. a control valve might provide slightly better control but it almost never really matters. close enough is good enough most of the time.

having said all that, I have run into cases where they put a VFD on the pump and install a control valve too. too much money I guess.
Click to expand...

you usually have a pump set serving multiple individually controlled loads/coils
this can't be controlled by a vfd

I have never seen a hw or cw loop vfd controlled
pressure is controlled by a small by-pass valve

there are plenty of cheap valves out there (belimo)

about the only application I (and most others) used vfds is sewage
debris fouls valves, although I have used a rubber lined pinch valves
but a vfd is better in this app
energy and capital costs are generally a wash
https://www.redvalve.com/products/control-pinch-valves
 
I would use a AB Powerflex 400 configured with a disconnect, full feature bypass and line reactor, then its power in to motor out. Easy install.

You can feed it from the MCC bucket without changing anything. You can leave the controls to start and stop the way they are already setup.

Program to drive to start on power up.

Then you need to decide on how to control the speed. It can be as simple as keypad speed control, or differential pressure control.
 
petersonra said:
cost wise these days a VFD is often more cost effective than a good control valve. a control valve might provide slightly better control but it almost never really matters. close enough is good enough most of the time.
Click to expand...
It matters in terms of energy savings. That alone justifies their cost.
 
Ingenieur said:
energy and capital costs are generally a wash
Click to expand...
No so, at least in my experience.
An example.
In one cement works, Blue Circle in this particular case, they couldn't run either of the Precip and Preheater fans in bypass without exceeding the plant max demand and incurring swingeing financial penalties.

I've also come across a good many cases, particularly in pumping stations where the VSD energy savings paid for the capital cost in about one year. Design life expectancy was 20 years.
 
Status
Not open for further replies.
Top