Submerged Switch

[mityeltu]

I have a question concerning a switch that will be submerged during a flood event. I believe I know the answer, but I want unbiased responses, so I'll hold my tongue.

I have a 120VAC control circuit wired with #14 sis cable. This control circuit is powered by a 1kVA 480/120 single phase xfmr. My wiring diagram does not indicate either leg to be grounded. I'm not saying that it is not, I'm just saying I don't know and the diagram doesn't show it.

The circuit is protected by two 10A fuses on the secondary side of the xfmr (1 per leg).

This control circuit has a pushbutton in it that is mounted on the surface of a junction box. The PB is a normally-closed component. This junction box will be submerged during the aforementioned flood. Additional components in the circuit are a motor starter and a TOL for the motor. The switch and the cables to the switch are the only components of this circuit that will see water. Al other components will remain dry.

My question is this, what happens to the circuit when the water enters the jbox?

 

[Jraef]

We can assume that the secondary is not grounded by virtue of having a fuse on each leg of a 120V circuit. You can only fuse the grounded side of a circuit under very specific circumstances. be that as it may, if the box fills with water and makes contact with either of those wires, current will flow unrestriced into the water to ground, which will eventually cause the fuses to clear. In the mean time, anyone in that water is at risk of electric shock.

The better approach, if the control station must be functional during the flooded time, and assuming it is brief, is to use a NEMA/UL "Type 6P" control station and push buttons. That's what they are for.

 

[mityeltu]

That's what I thought. That's not what I was hoping for though.

Since electrons follow the path of least impedance, wouldn't the closed switch provide that path? It's essentially a shirt between the terminals.

 

[gar]

130424-1613 EDT

mityeltu:

Unless you have salt water I doubt you would ever blow 10 A fuses, and possibly salt water would not.

An experiment:

My tap water, fairly high mineral content.
About 1000 ml in a 2000 ml beaker.
#18 rip cord with about 1" stripped from each lead.
The leads spaced about 3" apart and inserted into the water.
123 V 60 Hz applied between the two leads.

Current flow 0.12 A. About 1/100 of that required to blow a 10 A fuse.

If the source is ungrounded, then the impedance from this and the components connected to the source will have a fairly high value to ground (earth). The leakage resistance and capacitance and the distribution between the two source leads to ground is what will determine the impedance to earth and the current that will flow if someone contacts part of the circuitry and earth. A capacitance of 0.1 ufd at 60 Hz has a reactance of about 26,000 ohms. For cable runs that are not too long, 100 ft, that would be a very high value. More like 0.01 ufd or less. Leakages resistance is likely high, possibly megohms, until you put the switch in the water.

Also note 0.12 A at 120 V = 1000 ohms.

These comments do not say that you should dunk this switch in water. They are just to give you some quantitative values. Clearly under the correct conditions this could kill someone. The circuit needs GFCI at the input end of the circuitry meaning at the transformer secondary output.

If I had this problem I would likely use a low voltage electronic system with the switch fully potted (a reed switch or Hall device or variable inductor). You might even use a fiber optic type of switch.

.

 

[ronaldrc]

You didn't say what type of starter, if its a conventional magnetic starter that would be a kill switch and if it was submerged in water it probably would not drop out the coil even if some one was dumb enough to push it.


You said the secondary of the transformer was not grounded so as to being shocked you would have to be between or very close to the contacts when the button was pushed to be shocked.

Since the normally closed push button is the only thing in the circuit that is in the water. The resistance between the secondary and ground would be several megaohms. Not that big of a shock hazard.

That said do not let any body use it when submerged in water.

 

[mityeltu]

My concern is really not the safety of personnel at this point. The postulated event is a design basis flood. The building where the switch is located will be mostly submerged and vacant (or the ones foolish enough to remain after the warnings will be deceased from the flood and not electrocution - I'm not making light, just making an explanation). The real concern is ensuring the motor continues to run even if the switch is submerged.

The switch is not a kill switch as there is a seal-in contact (limit switch) on a flow control valve that will keep the circuit energized even if the switch were opened.

The starter is a size 1 magnetic starter.

I really appreciate all the input.

 

[JoeStillman]

Didn't Square D used to have a line of fiber optic devices for this kind of thing?

 

[gar]

130425-1041 EDT

mityeltu:

When flooded this switch is closed and will perform no useful function.

Thus, the question is "whether the wires being in the water will cause a malfunction of the control circuit?".

In turn if these wires are directly shorted to earth or any reasonable external voltage source relative to earth, then does the control circuit malfunction. I believe that malfunction here would mean cause the motor to stop. Reasonable voltage means a voltage that does not cause any insulation voltage breakdown.

If the secondary of the transformer and none of the circuitry is grounded, then grounding the wires to the submerged pushbutton, even with a direct wire short, would not have an effect on the operation of the circuit.

If water pressure causes the contacts to open, then you have a problem. An above water level shorting switch across the NC pushbutton might be useful. This might be a NO key switch. The logic of what was done would be dependent various factors. Also could be just two terminals on a terminal strip that are manually shorted during a flood.

.

 

[mityeltu]

This makes sense, but let me ask another question, if I may.

With the switch submerged, is there a danger of losing power on the circuit due to the grounded metal enclosure? In other words, is it likely that the AC power will short in some way to the enclosure through the water and blow the secondary side protective fuses?

I read the reply that stated the impedance of the water will be on the order of megaohms, so I believe that all will be well and the leakage current will be minimal, but I would like to make certain.

 

[ronaldrc]

Sorry I did not make that statement clearer.

When I said megaohms I was refering to the resistance of a nongrounded secondary in regards to the earth ground, not the resistance of the water.

And if the secondary wasn't grounded this would be like having a 120 volt battery with a load and having a loop somewhere in this circuit down in the water so this would not be a hazard to someone in the water unless they where in the water and broke this or opened this wire and where between it becoming the conductor. I hope that made sense?


Oh,if they did push that button while under the water well then it would be a kill switch. By the way what is that push button for?



Ronald

 

[gar]

130425-1350 EDT

mityltu:

If you look at the post where I ran the experiment with two exposed wires in my tap water you would see a current of 0.12 A or about 1000 ohms between the wires. This resistance is greatly dependent on the water conductivity (distilled vs most tap water vs sea water), and the contact area of the electrodes.

With river water and not much contact area to the water, and we believe the circuit is ungrounded it should be very difficult to blow 10 A fuses. Your leakage impedance is probably much above 1000 ohms and the leakage is likely to be the dominate current limiting factor.

You can put the switch inside a fiberglass box and thus force the leakage impedance to be much higher in this area of the circuit.

.

 

[mityeltu]

I have an update, the switch is NOT normally closed as I was told. It is normally OPEN and is used to START a strainer motor.

So the question now is, presumably, what will happen when the switch gets submerged? Will the switch short out and start the motor or will the swith geound and blow the protective fuses? Or maybe NOTHING will happen?

I apologize for the confusion, I was basing my questions on information I was being given verbally by a mechanical engineer. After looking at the drawings, the picture is a bit clearer.

 

[GoldDigger]

I have an update, the switch is NOT normally closed as I was told. It is normally OPEN and is used to START a strainer motor.

So the question now is, presumably, what will happen when the switch gets submerged? Will the switch short out and start the motor or will the swith geound and blow the protective fuses? Or maybe NOTHING will happen?

I apologize for the confusion, I was basing my questions on information I was being given verbally by a mechanical engineer. After looking at the drawings, the picture is a bit clearer.

All of the above are possible depending on exactly what sort of circuit voltages and current are appearing at the switch terminals.
If the switch is waterproof and the wiring to it, including terminations, is waterproof, nothing will happen.

Do you want the switch to work while submerged?

 

[eHunter]

Is the switch and associated enclosure, cabling, placement, installation methods and usage as proposed by the OP suitable for the intended application and permitted by code?

 

[ronaldrc]

I have an update, the switch is NOT normally closed as I was told. It is normally OPEN and is used to START a strainer motor.

So the question now is, presumably, what will happen when the switch gets submerged? Will the switch short out and start the motor or will the swith geound and blow the protective fuses? Or maybe NOTHING will happen?

I apologize for the confusion, I was basing my questions on information I was being given verbally by a mechanical engineer. After looking at the drawings, the picture is a bit clearer.

Thanks for answering my question:

The only way you would know for sure is bench test the relay or coil under the conditions of being submerged.

I personelly don't tink it would start since it takes a lot more voltage to pull a contactor in than to drop it out.

More and likely nothing will happen other than a corroded push button.

I stand behind what I said about the shock hazard you would almost have to be near the push button before it would be a shock hazard. But don't ever gamble with Electricity.

Ronald

 

[mgookin]

Your installation needs to be such that water rising to base flood elevation will not cause damage or a hazard. You need a water tight installation or you need to move the switch.

Electricity follows all paths - not the path of least resistance.

 

[mityeltu]

First, I understand the concerns. However, this is not a problem of safety of personnel at this point. The elevation is 698'. This is the 2nd floor of the ERCW building for a nuc. My primary, and really at this point my only, concern is ensuring that the equipment under investigation continues to operate after the switch is submerged and as long as the motor to which it is attached is NOT submerged. Eventually, the motor will be submerged and will cease to perform its function.

There will be no personnel in the facility (the ERCW building) at this time. This is not a "code" issue as this event will be a design basis flood event. The installation as it stands is up to code, but the building is "supposed" to remain dry. This is being done as a justification for the NRC.

After further investigation I think this has become a non-issue because of one reason. This is an UNgrounded system. The switch for this circuit is on one leg only of an ungrounded CPT. If ONE leg of this circuit is grounded, it will not impact the operation of the circuit. This would be similar to any other 120VAC grounded system. One leg of the 120V XFMR grounded. Just like at your house - neutral line connected to ground bus. This is not a problem unless you think I'm missing something.

If so, please, chime in. I want all my bases covered.

 

[gar]

130428-1024 EDT

mityeltu :

Unless you have highly conductive water the current flow between the two exposed terminals of the switch will be insignificant. This I demonstrated in my earlier experiment.

My suggestion: Use an AB switch with with a power reed contact. Put this in a fiberglass PB box. Bring the control wire in thru a compression fitting. Cover the switch terminals with a silicone sealant. Check with AB on how much water pressure the reed can tolerate. Check on how much water pressure would operate the pushbutton. Also consider a rotary knob, spring return if necessary, in place of the pushbutton actuator.

If this looks good, then make a sample and test it.

.

 

[mgookin]

... The elevation is 698'. ...

As a then green plans examiner from Florida attending FEMA's Emergency Management Institute up in Maryland in the 1990's I asked "What are those 3 digit numbers on the flood maps?" Everyone laughed and said in unison "Someone's from Florida." The highest base flood elevation in my home town was (is?) 15' AMSL. I was shocked to learn they have floods in the mountains! They said the worst flooding occurs after a big fire or a volcanic eruption because there's nothing there to slow down or absorb the rainfall. Just thought it would be fun to throw this out there.

 

[ptonsparky]

We can assume that the secondary is not grounded by virtue of having a fuse on each leg of a 120V circuit. You can only fuse the grounded side of a circuit under very specific circumstances. be that as it may, if the box fills with water and makes contact with either of those wires, current will flow unrestriced into the water to ground, which will eventually cause the fuses to clear. In the mean time, anyone in that water is at risk of electric shock.

The better approach, if the control station must be functional during the flooded time, and assuming it is brief, is to use a NEMA/UL "Type 6P" control station and push buttons. That's what they are for.

Why, how, would an ungrounded secondary of a transformer have any current flow to ground or even any reason to flow to ground? Granted the first fault would make it a grounded system, but the fault(s) to ground is within an inch or so of each other and would in no way involve any other path. The current flow will be between the open contacts if it is N/O. The worst that will happen is the pump will start.