Swimming Pool Lighting Safety
8th August 2007
Hi all -
I was talking with some students some basics about lighting safety the other day (avoid water, etc.). One student asked what specific lights, if any, would be used on a shoot where the lights were near -- but not in -- a body of water, say a swimming pool.
Though my immediate answer was "same lights as usual, but used with extreme care and safety", it got me thinking. I haven't been on a set where this was done, and I wanted some pro opinions.
Searching through the forums the best I could find (and good information it is) was stuff on lights in water and underwater filming, both found here:
There are a few companies that specialize in underwater stuff; IE HYDROFLEX and such. When we have to light near a large body of water, use all the safety and precautionary tricks. Be sure to put all distro and connections up on swamp boxes.
If you have to place a light near water MAKE sure you have a GFI on the lamp fixture. Personally if I know it's going to be a particularly dangerous situation, I rent some shock blocks, basically large GFIs that take bates into them. Also you may want to think about anchoring your stands...can get tedious. If something goes into the water, NEVER go in after it until the leg has been removed from the distro box. Also turning off your genie isn't a bad idea either. There is a good chapter in the Harry Box SLT handbook that has a lot more info in it.
Gaffer // Best Boy Electric
Los Angeles, CA
>>>Though my immediate answer was "same lights as usual, but used >>with extreme care and safety",
Yes the same lighting units but with extra precaution. I suggest using GFI circuits for any lights near a pool. They can be quite expensive but in the relative scheme of things a very good investment.
For larger lights you'll need another approach but for under 20amps, Hubbell makes an in-line 20 amp GFCI. I bought six of these for use on a low-budget shoot. It has male and female short pig tails coming out of an all weather type unit, I think model number is GFP3C20M.
They also make 15amp and others, in the in-line style. I think I paid around $75USD each. The only problem w/ the 20amp model is it has the 20 amp style (|-) plug and I had to
replace the male (or was it the female?) end with a standard 20amp plug and rewire plug myself. Did not have to replace both plugs, just don't recall which. 15 amp versions much cheaper. Harder to find 20amp versions.
Also you want to plug these in-line units directly to source, i.e., not at the end of a stinger run. Stinger run, if any, should come out of GFCI.
Obviously, it would be better to place the whole electrical system on a main GFCI, but this gets expensive. For smaller projects or a few lights, the Hubbell in-line is fairly cost effective.
John Roche, gaffer
Well to add my two cents . . .
I suppose that I will just reiterate one of our fellows who said the for bigger units, there is an item made in California called the "shockblock". Now what a shock block is, it is for the most part the same thing as a household GFCI, but in a form of a larger breaker that once the load is shorted will kill the breaker that is inline( typically designed for high amperage but has as well units for low amperage lights). It is a Film oriented item that at least in Los Angeles most rental house carry.
Another item I believe not mentioned(and I'm not 100% on the subject so I apologize if I am wrong), would be the use of DC power over AC power if available. Although it is not 100% safer, from what I understand there is a less of a risk of electrocution from a true DC output ( i.e. not a converted DC output from an AC generator).
In two I guess I have a question for you all.... How much safer is DC than AC in or around water?
LA,CA slt 728
Don't even start that DC in water one again.
Even at low leakage currents DC can cause electroporosis damage of the body cells and still has the potential to stop the heart if a high current passes through the body. What passes off as DC may not be perfectly smooth anyway, and it's the current transitions and ripples that cause muscle (heart) contraction in the first place.
The way forward is to subdivide circuits down into a large multiple of individually GFI (RCD) protected circuits to allow quick localised tripping and avoid the problems associated with cumulative leakage where one large GFI is used to protect a large number of lights.
Keep the GFI's well away from the water too!
Clive Mitchell wrote:
>>>Don't even start that DC in water one again.
>> The way forward is to subdivide circuits down into a large multiple of >>individually GFI (RCD) protected circuits to allow quick localised >>tripping and avoid the problems associated with cumulative leakage >>where one large GFI is used to protect a large number of lights.
>>Keep the GFI's well away from the water too!
What Clive said. Also keep someone standing by the source of power -- generator, mains to disconnect all power. If it's a night shoot, work out some emergency lighting so that you're not left in total darkness in an emergency.
Electrical hazards around a pool, hot tub or spa can lead to multiple deaths or injuries. This occurs when an individual becomes incapacitated by stray current in the water and one or more persons jump in or reach out to save the victim, resulting in multiple electrocutions or serious shocks. In May 2002, a 14-year-old girl from Arlington, Texas, was electrocuted when wiring problems in an apartment swimming pool's underwater lights charged the water with electricity. A 16-year-old boy was seriously shocked when he jumped in the pool to try to save the young girl. Another teenager used a fibreglass shepherd's hook (a non-conductive device) to pull both victims from the water.
To put things in perspective, the CPSC reports that there have been 60 deaths and nearly 50 serious shocks reported over the past 13 years involving electrical hazards in and around swimming pools, while they estimate that more than 500 drowning's occur each year in swimming pools (approximately three hundred being children under 6).
IA 600 DP
>>>The way forward is to subdivide circuits down into a large multiple of >>individually GFI (RCD) protected circuits to allow quick localised >>tripping and avoid the problems associated with cumulative leakage >>where one large GFI is used to protect a large number of lights.
Just to further clarify...the GFI's should be mated with the individual lighting units. So, in the US, a 20A GFI will work best with a unit using 10 - 20 amps of load and a 15A GFI will work best with a load of 15A or less. A 220v 100A GFI breaker would be best for a 12K and not 5x 20A lamps.
Be sure to take into consideration the continuous amperage and striking amperage for HMI's as they generally draw a bit more than a tungsten load.
A question I have is... how do the 2 pole, 220/240v, GFI breakers work if there is no neutral used? The 120v GFI's monitor the load between the neutral and the hot leg and shut off if there is a discrepancy. Does the 2 pole 220/240v GFI's just monitor the difference between the 2 hot legs of the load?
Hope this explains it..
A GFI usually measures the current going through and coming back from the light. If there is no leakage to ground the current to and from the light should be exactly the same. If a fault with the light occurs to ground or if current leaks away to ground when the light falls into water, there will be a discrepancy between what's going out and coming back. and will create a magnetic field (the magnetic field is normally balanced out as the outgoing and incoming current both generate a magnetic field inverted of each other.
Best to look at the picture displayed here :
RCD = a residual current device which is another name for GFI
Any electrical instrument or cable has some leakage to ground due to the age and molecular changes over time (insulation gets thinner, brittle etc.) and most RCD's have a threshold (otherwise they'd trip as a false alarm too often, causing people to bypass them in the long run) of a few mA before they trip
An RCD or GFI does not actually need a groundwire, other than passing through for the load and grounding/shielding the metal housing, as long as the load and generator is earthed with the groundpins
Werner Van Peppen
>>... how do the 2 pole, 220/240v, GFI breakers work if there is no neutral >>used? The 120v GFI's monitor the load between the neutral and the hot >>leg and shut off if there is a discrepancy. Does the 2 pole 220/240v >>GFI's just monitor the difference between the 2 hot legs of the load?
My understanding of US 220v is that it is a "two phase" supply. i.e. two actives, each 180 degrees out of phase to the other, balanced at 110v above earth.
As such a Residual Current Device [GFI] will work just as well as and exactly the same as a single phase [110v] unit - as you say, by detecting an imbalance between the two actives.
Note that RCD’s DO NOT measure earth current to function.
Perth, Western Australia.
One other thing that may be surprising... the best way to re-start the heart of a shock victim is to use a defibrillator. These aren't just for ambulances and hospitals any more. You can BUY a fully automatic defibrillator for around 1000 bucks. Units like Philips Heartstart will guide the user with spoken prompts and pictures on the back of the electrode pads show their correct placement. When applied the unit will automatically monitor the hearts electrical state and decide whether it's possible to make a recovery attempt, and upon operation of the confirmation button will attempt to defibrillate the heart at an exact calculated point in the detected heart activity.
Every film crew should have one for either electrocution or other heart problem scenario's. They're not big, but they're clever.
GFI's (RCD's) monitor the current flowing out and compare it to the current flowing back in. This is as simple as a few wire turns in opposing directions on a toroidal core. When an imbalance is detected it means current is finding an alternative route back and the imbalance in the core generates a trip signal.
The actual load current rating should be suitable for the light connected, but the actual tripping is based on a threshold of leakage current which may be 10mA or 30mA depending on local regulations.
The generators can also be fitted with much heavier devices with a higher trip threshold, but should also be programmed with a slight delay to allow localised GFI's to clear the fault first if possible.