Tank electricity and grounding probe thread

[jestep]

I figured I'd post one of these because I see a lot of related questions and recently a full tank crash or two due to shorted equipment. Will touch on grounding probes as they carry real risks which are rarely addressed by the people selling or promoting their usage.

Basics...

Saltwater is extremely conductive. It's nowhere near copper, but in something the size of a fish tank, you can assume it's close to the same as touching a bare wire.

Electricity is extremely dangerous, current flow being the harmful factor. You get current flow when there is a difference in voltage between 2 points that are connected by conducting material. It takes a minimum of 30V to be able to generate .1A of current flow in a human due to our skin and internal resistance. It takes .1A through a human heart to stop it.

Lastly, voltage in a tank does not mean there's current in a tank. Current flow is what causes damage in a tank and anything else. You must have a path to ground, another shorted piece of equipment, or a sizable difference in voltage, for current to exist in the tank.

Electricity and livestock

Electricity is very efficient at killing fish and even more so at killing coral and and inverts. They are made of mostly water and even though they are less conductive than the surrounding seawater (hence why ocean fish doesn't taste salty), they still get the crap shocked out of them. For coral, the electricity destroys the tissue, and even small voltages can cause enough damage that they can't repair it.

Getting electricity in a tank

First is natural induction that occurs due to spinning pumps. Since the water is a conductor, the pump essentially acts as a transformer. You test any tank on the AC setting of a multi-meter, you will typically see 1 - 10V. This is induced voltage, and is generally not harmful to coral. However, there is enough anecdotal evidence to suggest that induced voltage can be bad. This is most likely because there are a ton of devices inducing voltage causing minor but still damaging current flow between areas of the tank.

Second is when you have an actual short into the tank. In essence, one or both of the wires connected to a piece of equipment is directly in contact with the tank water. Or, your light fixture fell in or something similarly catastrophic This will cause a multimeter to read the full line voltage of the circuit, which is typically 110 - 120v, with some larger tanks using 220v or 440v equipment. When you have a full short into a tank, you should consider it to be extremely dangerous. 110v has no problem killing a human if it goes through the right organs. I hear the ol' "it's only 110v" all the time, which is so ridiculously ignorant it's amazing. Additionally, getting shocked by 110 is painful, and 220 or 440 can cause external and internal tissue damage which can lead to external infection or sepsis if internal tissue is destroyed.

Voltage in a tank

It's a debate whether voltage in a tank damages tank inhabitants. Since voltage in a tank is not indicative of current flow, there's no reason to assume that just because you are measuring voltage in a tank it's causing any damage. Despite some suggestions I've seen you cannot measure current flow in a tank using any normal means. Current flow must be measured inline unless you have some special equipment which still wouldn't work on a fishtank.

Grounding probes

A grounding probe is a wire connected to the round ground hole of a GFCI outlet and the other end of the wire goes in the tank. They're typically made of titanium so they don't corrode or introduce copper into a tank. Grounding probes are designed to protect a human from killing themselves by sticking their hand in a shorted tank. However, they are a double edged sword with risks to livestock.

Proper use

All of your in-tank equipment must be plugged in some way to the same GFCI outlet as the probe. If anything in the tank develops a short, the probe provides a ground path through the GFCI outlet, causing the GFCI outlet to trip, thus de-energizing all of the equipment. Since you can't know what equipment shorted, everything must be turned off.

You could conceivably get around this by plugging each piece of equipment into their own GFCI. Since the increased current flow would only exist on the GFCI with the shorted equipment, the correct GFCI would trip. I've never seen anyone try this, but it should work.

The probe should be placed in the sump to minimize current flow caused by induced voltage and to minimize the damage from the instantaneous current spike when a short develops.

Pros

Any current from shorted equipment in the tank will be shorted to ground saving you from getting shocked.

If there is some condition that would allow current flow in the tank when something shorts, it would also be alleviated using a grounding probe.

Depending on specific grounding probe setup

If a probe is placed in the correct location (such as a sump with pumps), they can short induced voltage to ground without tripping the GFCI. However, if there is anything between the voltage source and the probe, it will be shocked. Also, if you have large pumps, it's possible they would induce enough voltage to trip the GFCI without there actually being a short.

Cons

You get a short and instead of nothing happening, everything turns off! If you're on vacation or even away for the weekend, tank is toast, even though there may be no danger to the livestock.

If you have a pump that induces a lot of voltage in the tank, you will get current flow between the pump and the probe. You don't want this taking place in the DT. You also don't want a pump tripping your GFCI for no reason.

If you do develop a short with a grounding probe, you can create a staggering amount of current because the water has very little resistance. Another reason why a probe in a DT is a bad idea.

Why stuff dies when you measure voltage in the tank?

There may not inherently be current flow in a tank when something is shorted into it. But, if things die, there was a path to ground or some path to allow current flow in the tank.

There's a lot of anecdotal reports of induced voltage causing livestock loss. HLLE has been documented to be caused by electricity, and tissue recession in coral can be caused by electricity.

Again, there has to be current flow to cause damage. I'm not discounting that these situation probably had loss due to electricity in the tank. There's likely more to the story in any tank that's experiencing loss due to electricity. But the bottom line is that current flow is causing the damage and not the simple fact that voltage exists in the tank.

[George Monnat Jr]

Great writeup. It's inline with a post on RC.

[jestep]

Here's a great post on the topic on RC.

I hadn't seen that one. Really good breakdown of every possibility.

[FarmerTy]

Appreciate the write-up Jestep. Really clarifies the issue.

[SoggyDollar]

Great stuff. The moral of the story is... Keep a voltage probe in the sump and plugged into the same GFCI plug as all other potentially shocking equipment. Right?

Do the average powerheads and return pumps really produce enough induction to cause a harmful shock?

[FarmerTy]

I'm in the no GFCI and no grounding probe corner. Maybe standing in the corner by myself but I would rather be mad at a pump or whatever frying my corals instead of mad at myself for trying to do too much and having the GFCI trip and me not around to notice it and all my livestock and coral die for no reason other than because it tripped. I would feel like I outsmarted myself in that scenario and wouldn't forgive myself for that.

If a pump fried everything... I'd be hella mad at the pump but then would just rebuild. If the GFCI tripped, i would feel defeated. Nothing is worst than feeling self-defeated.

[jestep]

Great stuff. The moral of the story is... Keep a voltage probe in the sump and plugged into the same GFCI plug as all other potentially shocking equipment. Right?

Do the average powerheads and return pumps really produce enough induction to cause a harmful shock?

A GFCI is the most important for personal safety. As for the pumps, most people never have problems. If you use isolated pumps like vortechs, or your pumps are mainly in the sump, the potential for problems goes down a lot. There are enough stories that it's obvious that some people have problems. But I do believe that scientifically these must include factors that normal tanks don't experience, otherwise we would all have problems.

If a pump fried everything... I'd be hella mad at the pump but then would just rebuild. If the GFCI tripped, i would feel defeated. Nothing is worst than feeling self-defeated.

I would definitely check with a multimeter occasionally I've been shocked on a non-gfci outlet which coincidentally had a broken heater attached to it. My arm was all the way in the tank. I grounded my shoulder to the new metal light fixture. Not something I would ever want to repeat especially being shoulder deep in a tank when it happens.

[Richard L]

Thanks Jamie that is very helpful. The thing I haven't wrapped my head around yet is why does it matter that the probe is plugged into the same GFCI as the equipment? I thought a ground is a ground so electricity will take take its path of least resistance to whatever ground is available. And isn't the ground for a GFCI plug essentially grounded to the same grounding point as every other outlet in most of our houses? I'm pretty sure that nay electricity travelling thru any of my ground circuits in my house eventually ends up at a copper stake driven into the earth near my master breaker panel.

[George Monnat Jr]

Thanks Jamie that is very helpful. The thing I haven't wrapped my head around yet is why does it matter that the probe is plugged into the same GFCI as the equipment? I thought a ground is a ground so electricity will take take its path of least resistance to whatever ground is available. And isn't the ground for a GFCI plug essentially grounded to the same grounding point as every other outlet in most of our houses? I'm pretty sure that nay electricity travelling thru any of my ground circuits in my house eventually ends up at a copper stake driven into the earth near my master breaker panel.

The GFCI senses an imbalance in current flow, so if the current is going from a device on one GFCI to ground through a second GFCI, one or both of them could trip. I have two outlets on either side of my DT, and I saw this problem a lot using two GFCI (one each outlet) with a ground probe. They tripped quite often, so I got rid of them and the probe. With multiple outlets/GFCI you can't control "crossover" through the ground probe.

[jestep]

You're right, I miss worded that. The equipment must be in a GFCI outlet to trip (breaker will trip as well without a GFCI), but ground is common. As long as the probe provides a path to ground, the current spike would be seen in whatever GFCI the faulting equipment is plugged into.