OK, I need some help. I’m looking at several fountains that have 24v ac lights and they are fed through a 120/24v transformer. They are connected to a 50 amp c/b. If I understand the NEC correctly, these lights require gfci protection. I cannot find 12v gfci’s. Any suggestions? There has been an electrocution on this project, prior to me being called in, so I am very cautious.
24v GFCI’s?
- Thread starter morepower
- Start date
- Status
hillbilly1
Senior Member
- Location
- North Georgia mountains
- Occupation
- Owner/electrical contractor
A death on a project gets everybody’s attention! If you need gfi protection, it would be on the primary side of the 120 to 24 volt transformer. The protection would be if the transformer failed, and allowed the higher voltage to the fixtures.
Not according to the NEC. First off, protection on the primary is of no value on the secondary if you have a fault on the secondary. The code requires gfci protection on all light fixtures greater than 15v. The transformer is an isolated transformer.
winnie
Senior Member
- Location
- Springfield, MA, USA
- Occupation
- Electric motor research
I am not sure if anyone makes a strictly 24V GFCI, however I know that GFCI _sensors_ have been available.
For example, the Leviton 8895-E (no longer available) was marketed as a 'high current GFCI'. It looked like a blank face GFCI on the front, with a fat current sense toroid on the back. You would power the unit with an ordinary 120V 15A circuit, and pass the protected circuit conductors through the toroid, and control the circuit using one of a list of approved switching devices. So this GFCI _sensor_ operated at 120V, and controlled a contactor or shunt trip at 120V, but the actual circuit being protected could be anything less than some rather high current (like 50 or 70A). If you wished, you could provide GFCI protection for a 24V 3 phase 50A circuit.
A quick search shows: https://www.automationdirect.com/adc/shopping/catalog/sensors_-z-_encoders/ground_fault_sensors
I am not sure if the above would be considered a proper Class A GFCI for personal protection, you will need to do your own due diligence. However I hope the concept of a ground fault sensor or relay will help you searching out suitable equipment.
-Jon
For example, the Leviton 8895-E (no longer available) was marketed as a 'high current GFCI'. It looked like a blank face GFCI on the front, with a fat current sense toroid on the back. You would power the unit with an ordinary 120V 15A circuit, and pass the protected circuit conductors through the toroid, and control the circuit using one of a list of approved switching devices. So this GFCI _sensor_ operated at 120V, and controlled a contactor or shunt trip at 120V, but the actual circuit being protected could be anything less than some rather high current (like 50 or 70A). If you wished, you could provide GFCI protection for a 24V 3 phase 50A circuit.
A quick search shows: https://www.automationdirect.com/adc/shopping/catalog/sensors_-z-_encoders/ground_fault_sensors
I am not sure if the above would be considered a proper Class A GFCI for personal protection, you will need to do your own due diligence. However I hope the concept of a ground fault sensor or relay will help you searching out suitable equipment.
-Jon
SceneryDriver
Senior Member
- Location
- NJ
- Occupation
- Electrical and Automation Designer
I don't see anything in the documentation that shows them listed as Class A GFCI devices. These would likely work well, but do not carry the proper UL listing for protection of people.
SceneryDriver
hillbilly1
Senior Member
- Location
- North Georgia mountains
- Occupation
- Owner/electrical contractor
If that’s true, then somebody makes them. Have you checked with the manufacture of the lights? This would be an issue that they would have run into before.
winnie
Senior Member
- Location
- Springfield, MA, USA
- Occupation
- Electric motor research
Not very good documentation: https://darwinawards.com/darwin/darwin1999-50.html
According to a more authoritative source, the resistance of the interior of the human body is about 300 Ohm: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2763825/
Just skimming the paper they describe how currents too low to cause heart issues can cause loss of muscle control which could result in drowning, and describe an experiment where 3.05V was sufficient to cause involuntary 90 degree flexation of the knees. So there are documented controlled cases of low voltage inducing effects that could kill someone in the right circumstances.
-Jon
LarryFine
Master Electrician Electric Contractor Richmond VA
- Location
- Henrico County, VA
- Occupation
- Electrical Contractor
I wouldn't want to experiment with a 9v battery applied to cuts on both hands.
- Location
- Chapel Hill, NC
- Occupation
- Retired Electrical Contractor
Are you saying that you have a branch circuit of 50 amps feeding these lights?
Also the transformer would have to be approved for fountains. I don't believe the nec requires gfci for 24v lights
Also the transformer would have to be approved for fountains. I don't believe the nec requires gfci for 24v lights
synchro
Senior Member
- Location
- Chicago, IL
- Occupation
- EE
If the 24V secondary is ungrounded, which I believe it should be, then a GFCI fed by the secondary will not trip on a ground fault at the GFCI output because a closed circuit will not be completed by the fault. It would take faults on both the line and load sides of the GFCI to form a closed circuit path around the GFCI that would cause it to trip.
The relevant code rule is:
680.51(A) Ground-Fault Circuit Interrupter. Luminaires, submersible pumps, and other submersible equipment, unless listed for operation at low voltage contact limit or less and supplied by a transformer or power supply that complies with 680.23(A)(2), shall be protected by a ground-fault circuit interrupter.
680.23(A)(2) Transformers and Power Supplies. Transformers and power supplies used for the supply of underwater luminaires, together with the transformer or power supply enclosure, shall be listed, labeled, and identified for swimming pool and spa use. The transformer or power supply shall incorporate either a transformer of the isolated winding type, with an ungrounded secondary that has a grounded metal barrier between the primary and secondary windings, or one that incorporates an approved system of double insulation between the primary and secondary windings.
The Low Voltage Contact Limit is: 15 volts (RMS) for sinusoidal ac, 30 volts for continuous dc .
If you can't find a 24VAC GFCI, you may need to change the transformer and lights to 15VAC or less using a transformer identified for pool/spa use. If you could find a listed pool/spa power supply that is 24VDC (fixed, not just rectified pulsating DC), the lights may not care about the DC source. Could be an excuse for a new LED retrofit.
680.51(A) Ground-Fault Circuit Interrupter. Luminaires, submersible pumps, and other submersible equipment, unless listed for operation at low voltage contact limit or less and supplied by a transformer or power supply that complies with 680.23(A)(2), shall be protected by a ground-fault circuit interrupter.
680.23(A)(2) Transformers and Power Supplies. Transformers and power supplies used for the supply of underwater luminaires, together with the transformer or power supply enclosure, shall be listed, labeled, and identified for swimming pool and spa use. The transformer or power supply shall incorporate either a transformer of the isolated winding type, with an ungrounded secondary that has a grounded metal barrier between the primary and secondary windings, or one that incorporates an approved system of double insulation between the primary and secondary windings.
The Low Voltage Contact Limit is: 15 volts (RMS) for sinusoidal ac, 30 volts for continuous dc .
If you can't find a 24VAC GFCI, you may need to change the transformer and lights to 15VAC or less using a transformer identified for pool/spa use. If you could find a listed pool/spa power supply that is 24VDC (fixed, not just rectified pulsating DC), the lights may not care about the DC source. Could be an excuse for a new LED retrofit.
Eddie702
Licensed Electrician
- Location
- Western Massachusetts
- Occupation
- Electrician
As @hillbilly1 mentioned the equipment mfg. must run into this on a daily basis. Through it back in their lap.
Yea, as I read it, the NEC requires gfi protection for lighting and other equipment operation above 15v.
These lights are LEDs. The transformers are part of a control panel that feeds pumps as well as lights, obviously on separate circuits. You state the low voltage contact limit is 15v, ac, but I don’t see that as a limitation. I feel it can be higher, as in this case where it is 24v.
Also, to change the lights would be very expensive in this application. It may come to that though.
Unfortunately, this would be a moot issue. The NEC requires gfi protection for anything above 15v in a fountain.
I asked for credible evidence of a fatality. That’s why OSHA and CSA set the AC limit at 50 V. Roberts (Canadian) found some very questionable claims from China and as a result I think one province set a lower limit but not the Canadian feds and none of the western provinces. I have heard one guy in a mining operation with a pacemaker and some crazy circumstances died at 48 VAC in a US mine but I have never found the report. Supposedly even the mine inspector said it was a crazy unlikely event.
As to the timing question below 8 milliseconds the amount of current doesn’t matter...it’s just to short of a pulse to induce fibrillation. Above 5 seconds if it hasn’t happened yet it won’t. There is a time aspect to it so we can allow higher currents if we just trip out quickly. The 3-5 mA stuff is based on not allowing anything more than feeling a tingle where in substation design for instance we don’t care if it hurts as long as you still walk away,
The let go threshold is around 20-30 mA, fibrillation 100 mA, and burning flesh at around 1 A. Generally 1000 ohms is assumed in IEEE standards. IEC uses 650 ohms but it’s voltage dependent and if you pay attention to that last part it’s up to 1000 ohms right at the 100 mA threshold. So if we take 1000 ohms x 0.1 A we get 100 V. So OSHA, CSA, 70E, etc., are already super conservative at 50 VAC. This is based on wet skin.
A GFCI should just work. It’s looking at very small current differences (under 5 mA). Only issue is the power circuit to run the electronics and the relay would have to be modified to run on a lower voltage. Maybe some kind of universal one that works on say 15 to 50’VAC or DC if one was built. Seems like it’s just simpler to use a DC power supply so you can use the higher limit.
The stupid part is that NEC and OSHA still don’t realize that a GFCI can’t prevent a shock.
As to the timing question below 8 milliseconds the amount of current doesn’t matter...it’s just to short of a pulse to induce fibrillation. Above 5 seconds if it hasn’t happened yet it won’t. There is a time aspect to it so we can allow higher currents if we just trip out quickly. The 3-5 mA stuff is based on not allowing anything more than feeling a tingle where in substation design for instance we don’t care if it hurts as long as you still walk away,
The let go threshold is around 20-30 mA, fibrillation 100 mA, and burning flesh at around 1 A. Generally 1000 ohms is assumed in IEEE standards. IEC uses 650 ohms but it’s voltage dependent and if you pay attention to that last part it’s up to 1000 ohms right at the 100 mA threshold. So if we take 1000 ohms x 0.1 A we get 100 V. So OSHA, CSA, 70E, etc., are already super conservative at 50 VAC. This is based on wet skin.
A GFCI should just work. It’s looking at very small current differences (under 5 mA). Only issue is the power circuit to run the electronics and the relay would have to be modified to run on a lower voltage. Maybe some kind of universal one that works on say 15 to 50’VAC or DC if one was built. Seems like it’s just simpler to use a DC power supply so you can use the higher limit.
The stupid part is that NEC and OSHA still don’t realize that a GFCI can’t prevent a shock.
- Status