Single Phase Induction Motor & tripping GFCI

[bbleiler]

We manufacture aeration pumps/fountains. Our latest product line are large units, which use 7.5 & 10HP motors. Because of the weight of the units, about 200-250lbs., the capacitors for the single phase permanent split-capacitor (PSC) motor models are mounted on shore in the control cabinet. The aerators need to be GFCI protected. What we have found is that nuissance tripping of the GFCI occurs when the cable runs get too long. Things are generally fine up to 400' of 4/4 SOOW cable. We just finished testing a 600' length of cable. When the capacitors are in the control cabinet 600' away from the motor the GFCI trips immediately. When the capacitors are mounted inside the motor unit right at the motor connections everything is fine. One thing I have noticed is that when the capacitors are in the control cabinet the line currents show an imbalance, which seems to be proportional to length of the cable, meaning the longer the cable the greater the imbalance. When the caps are mounted in the motor unit the line currents are almost equal. Does anyone understand what is happening here? If it is the capacitive losses in the cable, why would the location of the capacitors matter?

 

[aolalde]

GFCI will operate with ground currents as low as 4 to 6 milliamperes.
Capacitive reactance to ground to reach a current leak of 6 mA, required to operate the GFCI should be such to produce at 60 HZ a capacitive reactance Xc around 36.8 k-Ohms (at 220 V).
The running capacitor installed in the control panel will demand an extra line (600’) to connect the auxiliary motor winding. That cable could provide the extra capacitance lowering the total reactance to ground and triggering the GFCI protection.

 

[bbleiler]

aolalde,
Thanks for the reply, if I understand you correctly you are saying the cap losses are cumulative. The more conductors the greater the losses. That does make sense. Although, one thing I did not mention is that we also have 3 phase units out there that operate on runs longer than 400'. This condition seems to be more related to the 1ph units, but I have no evidence at the moment. Is there anyway to determine a theoretical maximum length of cable, you know maybe the cap losses per foot? If only 2 conductors are current carrying we know 600' works, how long could the cable be before it also trips the GFCI. Also with 3 current carrying conductors we know we are good to 400', but tripping occurs at 600', can the tripping point be calculated? The dielectric constant of EPDM is about 3. The insulation thickness is about .080".

 

[dpc]

The moisture in the conduit will have a big impact as well, so a precise calculation will be difficult. You really need to get the GFCI device closer to the motor to have reliable operation.

 

[DickDV]

I'm outside of my expertise here but, as I recall, the cap on a PSC motor is rather low capacitance. I would think that putting that cap 600 feet from the motor would change the total capacitance enough to affect the second phase creation in the motor.

Just a hunch. Don't have any data.

Anybody care to comment?

 

[waross]

Hi DickDV;
Low capacitance is a relative term.
However, leakage to ground due to cable capacitance will be measured in ma., while the current through the PSC capacitor on a 10 hP motor will be measured in amps. The ratio will probably be more than 2000:1. Also, the PSC capacitor is series connected and the cable capacitance is shunt connected.
The bottom line will be no effect on the phase angle due to cable capacitance.
respectfully

 

[aolalde]

Capacitors in parallel add Ct= C1+C2+...Cn
as oposed to resistors in parallel; 1/Rt = 1/R1 + 1/R2 + ...1/Rn

For a given frequency (f), the larger the capacitance the lower the reactance. Xc = 1/(2*pi*f*C).

Leakage current due to humidity throgh the insulation resistance could add to the problem.

Ask to your wire provider for average capacitance per feet figures for the diferent wire sizes. Mesure the insulation resistance of the specific cables to ground, with a Megger.

 

[bbleiler]

The megger readings for the cable are normally around 1-1.5 megohms when submersed in water, otherwise it is generally infinite.

The capacitors are 3-60uF in parallel for 180uF in series with the aux. winding.

I agree the cable cap losses have nothing to do with the volt-current phase angle, but does the high cable resistance act in conjunction with 180uF of capacitance to somehow give the I imbalance?

 

[dpc]

The megger readings for the cable tell you the dc resistance, not the 60 Hz impedance. With a wet cable, there is a lot of shunt capacitance in the cable. This can create leakage that does not return in the neutral, causing the GFCI breaker to trip. This is a well-known issue and is the reason why GFCI protection is normally placed at a receptacle rather than in a panelboard breaker. The longer the feeder, the greater the leakage. Determining a maximum length using theoretical means will probably not be terribly accurate, due to all the physical variables involved. Empirical testing is probably better, but even then, I wouldn't push the envelope.

Any reason you can't put a GFCI breaker out at the aerator?

 

[mc5w]

Almost every permanent split capacitor and capacitor start capacitor run motor that I have seen has less current in the auxiliary winding that is connected in series with the capacitor than for the winding that is connected across the line.

A long cable will add some capacitance to the 180 uF in the control cabinet. Some of the capacitance will be in parallel with the 180 uF capacitor and will change the phase angle of the auxiliary winding. Some of the other capacitance will be to ground and essentially creates a T network that may or may not affect phase angle. The effect of the capacitance to ground will also be dependent upon system grounding method. Corner grounding and solidly ground wye systems and 4-wire delta systems produce different effects. Ungrounded places system capacitance is in series with auxiliary conductor capacitance to ground and ungrounded systems have a very high rate of motor damage due to static electricity buildup during rainstorms. An ungrounded system tends to reenact Benjamin Franklin's kite experiment and lightning arrestors for 480 volts ungrounded have a clamping voltage that is 200 to 800 volts above the 30 minute voltage withstand rating of a 480 volt motor. This is a bit how a Cleveland, Ohio police officer's bullet resistant vest did not work because the bad guy got lucky and sent a bullet up through the armpit. However, one way to get rid of static electricity on an ungrounded system is to use a ground detector that creates a direct current path through the primary windings of voltage monitoring transformers. Another way to combat static electricity particularly on the load side of a 480 volt motor controller is to connect 1 megOhm 5 watt resistors from each phase conductor to ground. What you can get are 240,00 Ohm 3 watt metal film resistors that have a peak voltage rating of 750 volts. You would use series strings of 4 of these per phase on 480 volts, 5 per phase on 600 volts, and 3 per phase on 240 volts.

If a motor is turned off for significant periods the load side of the motor controller needs static bleeder resistor regardless of system grounding method except 120 volts solidly grounded.

An alternative to conventional GFCI is to use a combination ground fault/ground check unit that does continuous monitoring of the equipment ground. The ground check circuit allow the ground fault relay to have a setting of 0.1 amps to 10 amps which will tolerate the leakage current of a 4,160 volt motor. One place that peddles these is

http://www.bender.org

.

The usual ground check circuit has a diode at the load end of the circuit that connects the ground check condutor to the machine frame using fasteners and lugs that are independent of the equipment ground. The diode is operated with the anode connected to the machine frame so that in high vibration applications a 50 amp diode can be used for mechanical ruggedness. The ground check relay uses either a 12 volt transformer winding or alternate connection to the +12 volt and -12 volt analog power to do 2 different continuity checks. The positive ground continuity check checks the integrity of the equipment ground. The application of negative ground power to the ground check conductor checks for ground faults in the ground check conductor.

If the power system is solidly grounded or ungrounded you will also need to put a fused telephone protector block in the ground check conductor to protect the ground check relay frommthe voltage drop in the equipment ground that will occur curing a ground fault. A ground fault/ground check relay as is is approved for resistance grounded systems where the first ground fault is limited to on the order of 1 to 20 amps. The level of ground fault current on a resistance grounded system will not create enough voltage drop in the equipment grounding conductor to affect the ground check relay.

You will need type G-GC cable for this application. This contains 3 or 4 large power wires and the same number of smaller wires in the case of the unshielded version. One of the small wires is used for the ground check circuit and the others are paralleled to make u0p the equipment ground.

Mike Cole

 

[Alanator]

If memory serves me right, the 3-phase GFCI units are rated as ground fault protection for equipment. The trip levels may be higher (~30 mA??). The single-phase units are probably rated as ground fault protection for personnel (4-6 mA trip level). A look at the data sheets might clear things up.

 

[armandh]

it the tripping when turning off

pesky triping when turning off from counter EMF kick back can be cured or vastly reduced with a few extra MOVs