earth fault protection 3

[parlitu]

I have a pump, protected by one eart-fault relay and a breaker. The earth fault relay is set to trip the electrical supply if I have a un-balance current of 4A, for a time longer than 0.5s. I measure the starting time (the time until the current goes from inrush to nominal) 30 days ago and was 0.35s.The pump has to assure a pressure of 15 bar. To protect the pipe and forward components, in derivation I have a relief valve that opens when the pressure is higher than 19 bar. As I notice, this valve does not work properly, seems to open at a lower value that I did not measure. Sometimes when I start the pump, the earth fault relay trips the pump. Seems that the pump starting time is higher than 0.5s. It is possible that due to valve wrong operation, the starting time of the pump to be longer?

 

[7anoter4]

Welcome parlitule
First of all I don't think that the Earth Fault Relay has to act if no insulation damage is involved [also if relay setting is suitable, let say 10-20% of full load current].
Nevertheless, if the discharge valve of the pump is closed the discharge pressure will increase and the load torque will increase also up to 19 bar when the relief valve will open .So, the load torque will grow up to 19/15 =127% from the nominal.
I suppose the pump is activated by an asynchron [induction] motor.
tp=GD^2*nm/375/Tacc start time
Tacc =TM-Tload where TM is the average motor torque and Tload is the pump average torque.
Tacc =net acceleration torque [will be less than necessary, so the start time will be longer].
If the breaker overload is set to 120% the breaker may trip as the motor current may reach 130% and more.
An other phenomenon may be water hammer effect which may increase suddenly the pressure and the load torque up to relief valve action. See attachment. Hope is useful.
For more details try at Eng-Typ Pump engineering Forum, I think.
Best Regards

 

[parlitu]

Thank you for advise 7anoter4. Is true that earth fault relay should not act if i don't have a insulation problem. But when the motor is start, the current are not at the same value on all three phases. For exemple Ir=75A, Is=69A, It=71A. Because all 3 wires goes through the toroidal curent transformer, its sum is not 0, as it should be. That's why, I have a temporisation for earth fault relay.

 

[slavag]

Hi Parlitu.
E/F relay connected to the toroid isn't sensetive to the unbalance current at the start-up time.
What is a size of toroid? What is a size of motor and voltage.
Please check power cable shield connection, it must connected trough toroid and back via toroid to grounding bus.
We discuss about it bfore few mounth, I try found this thread and will attached it.
Good Luck.
Slava

 

[slavag]

See attached, maybe it's your problem
thread238-205428
Regards.
Slava

 

[7anoter4]

Hi parlitu
Thank you for the vote [star].However I am not sure that I won it honestly since I did not answer to your question yet.
Don't be afraid by the elevated discussion of Slava, Gepman and dpc. There are experts on the protection field. So I try to explain in a few words what their problem is. With all respect, I am not sure that is really your problem. As I think, from the few data you fetched up, the actuating motor is an induction motor of approx. 37-40 kw 400 V.So no shield is on this LV cable. I think also the setting of 4 A for this motor is very low.
7-8 A will be better.
What Slava said is that if the cable is shielded, the end would be grounded through a jumper [or three grounding jumpers, one per phase] then the jumper must pass through the homopolar CT in order to canceled the current return through the shield.
In this case as dpc emphased, if no cable phase is damaged[grounded] the sum of phase current would be still 0 then the total net electrical current passing through the CT should be also 0 and the shield conveying a shortcircuit current from a remote substation undetected will continue to flow through the shield for a long time[damaging the shield].I think in this case[a cable shield grounded at both ends in remote stations] we need an other CT with no grounding jumper passing through.See the attachment.I hope is more simple to understand.
Best Regards

 

[7anoter4]

Sorry the second jpg didn't appear

 

[slavag]

7Another4. As usually, you are great!!!! Thank you.
Regards.
Slava

 

[waross]

Is this a centrifugal pump?
If the relief valve sticks open it will be more of an issue than if it sticks closed.
The power drawn by a centrifugal pump decreases when the output is throttled. Some large pumps have a discharge shut off valve to reduce starting times and currents. When the pump is up to speed the discharge valve is opened.
If this is not a centrifugal pump, disregard this information.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[7anoter4]

You are right Waross. We don’t know what kind of pump is. It may be any kind of rotary pumps [positive displacement or other] too. There are a lot of centrifugal pump types, also. Here is a citation from an article published by Pacific Liquid& Air System –"Centrifugal Pump Theory":
"Different pumps, although designed for similar head and capacity can vary widely in the shape of their Characteristic Curves. For instance, if two pumps are designed for 200 GPM at 100 ft TDH, one may develop a shut off head of 110 ft while other may develop a shut off head of 135 ft. The first pump is said to have a flat curve while the second is said to have a steep curve."
In my approx. calculation I took 2" pipe and 15bar [approx.450 ft head] so I got 40 kw.
If would be 6" pipe I should get 600 kW [then medium voltage cable, shielded of sure].The viscosity is other important factor. Nothing is sure.
Best Regards

 

[electrolitic]

Hi parlitu
Your 11 of march reply has a misunderstanding about toroidal ground sensor. If you do not have ground leakage the sum of flux will be always zero even if you have a single phase running. It is a vectorial sum of current.
Using the residual CT connection different degrees of CT saturation can cause erratic relay operation. Using toroidal ground sensor, don't.Please check capacitive leakage to ground if you have long cabling. Correct shield connection it was already explained by Slavag.

 

[parlitu]

The pump is a centrifugal one. The motor is low voltage supplied. The cables from breaker to pump motor are around 100 meters. The cables passing through toroidal CT core are not enclosed in a metalic sheath. Good point slavag and 7another4, if a sheath was involved, it must be compensated by threading-back the sheath grounding through the CT core. The ampere-turns produced by the return conductor and the ampere-turns produced by the sheath current will be cancelled and the net flux linking the core will be produced only by the three phase currents.
Interestin point waross. I will check this possibility too.
Thank you all.