Thoughts on water heaters and water pipes as GE

[electrofelon]

Ok two for the price of one today. Thoughts wanted on two different cogitations I have had lately:

1) Please discuss using a metallic municipal water pipe as an EGC on utility systems where the primary and secondary neutrals are connected together (or multiple services fed off same transformer). IT just really bugs me that one of the basic goals of the NEC is to stop normal current flowing on things that are not supposed to carry current. In the aforementioned scenario, we have a - usually bare - conductor that would be carrying potentially a significant amount of current continuously. Is the advantage gained from using the pipe as a GE worth this result? When did we get so obsessed with earthing that we decided this was good?

2) Please discuss Water heaters and GFCI protection. I have troubleshot water heaters in the past and found a burned up element. That element was just sitting there in the water energized. Now I know water is a poor conductor, but why is this apparently a non risk? Maybe because people generally wont take a cold shower or bath ?

 

[kwired]

Ok two for the price of one today. Thoughts wanted on two different cogitations I have had lately:

1) Please discuss using a metallic municipal water pipe as an EGC on utility systems where the primary and secondary neutrals are connected together (or multiple services fed off same transformer). IT just really bugs me that one of the basic goals of the NEC is to stop normal current flowing on things that are not supposed to carry current. In the aforementioned scenario, we have a - usually bare - conductor that would be carrying potentially a significant amount of current continuously. Is the advantage gained from using the pipe as a GE worth this result? When did we get so obsessed with earthing that we decided this was good?

2) Please discuss Water heaters and GFCI protection. I have troubleshot water heaters in the past and found a burned up element. That element was just sitting there in the water energized. Now I know water is a poor conductor, but why is this apparently a non risk? Maybe because people generally wont take a cold shower or bath ?

1. If you don't intentionally bond the water pipe it often becomes bonded anyway but via equipment grounding conductors to items in contact with the water piping.

2. To be shocked you have to be exposed to differences in potential. The faulted element may change the potential between the water and earth, but you need to be exposed to both before you will feel anything. That risk is maybe there in some cases but not all. If that water passes through grounded metallic piping then that grounded piping shields you. Grounded tank on the water heater also would shield you.

 

[jwelectric]

I have copper water pipes that are bonded to the service with a #2 copper conductor. When my element burns into why does it not trip the breaker?

 

[electrofelon]

I have copper water pipes that are bonded to the service with a #2 copper conductor. When my element burns into why does it not trip the breaker?

Unless the element faults to the tank when it melts, the water is too poor a conductor to trip the ocpd

 

[mbrooke]

Replying to the first, IT DOES BOTHER ME BIG TIME AS WELL. I guess we think alike.:roll: Now, Utility aside, in my opinion bonding to the water main has its benefits. The metallic water pipes make the best ground rods, especially when part of a municipal system. As a result it forms a redundant mesh, which is best for equal potential and contingencies such as an open service ground (normally the neutral). Also is the added protection that when a live phase comes in contact with a water line there is a near complete guarantee that enough current will flow to clear the fault. And lightning protection to could be argued as enhanced.

However, these are not the only reasons and not the only influence behind the water bonds. And Utility practices make it look awful. In the real world water lines are the utilities biggest neutral, aside from the earth itself and the common neutral on the poles. And of course the phone shield wires as well and Im sure others are being left out. The NEC does try to limit stray current as much a possible (heck one can not use a ground bar as a neutral bar at the service entrance because the current can flow over the enclosure to the bond screw), but keep in mind the NEC is influenced by outside sources as well like the Utility industries. A service is the point where the NESC (usually the code utilities choose to follow) ends and where the NEC begins. The NESC and NEC do contradict themselves frequently, and at the service is where the two fight to blend together with out issue. The NEC has to give in, in this area to get things to work out.

And I think our obsession with water bonds was pushed the most by utilities. By adding a water bond at each service the utilities get a free super neutral and free lightning protection at the customer's expense. Utilites treat ground, earth and neutral as the same thing to save costs. The water bonds save a utility a larger neutral, ground rods, grounding grids, lightning arresters, headaches from broken neutrals, ect and as a result using single bushing transformers/primary wye grounded becomes easy with less worry for them. The primary wye further gives savings for reasons to much for me to go into now.

Now the end result that I am not found of: Current is always dividing between the neutrals and ground/other metal objects. The outcome from this is EMFs, stray voltage/elevated touch potentials and shock/electrocution. Summarizing the 3:

>EMF because of the inability for all current to cancel in a race way/cable/conduit/wire bundle ect because often half if not more of it is going somewhere else. EMF though debatable, could be a health risk. And when one considers the the fact North America has higher cancer rates over other countries where such practices are rare ... well lets just say conspiracy theorists and those in the field of non-ionizing radiation have something to think about. And If it is to be proven a health risk that info might never come out for obvious reasons the utilities could get sued by many.

>Stray voltage and touch potentials. Comes from voltage gradients caused by varying resistance across objects with voltage being pushed through them. Think voltage drop across a wire. The problem is real and has been seen before both effecting people and animals: Reports of people being killed and shocked in lakes from corroded neutrals is one. Another and the biggest are complaints from dairy/animal farmers who have encountered the problem all over the the country loosing livestock production.

>This one hits home for some plumbers and electricians: A water bond masks either a failed low voltage neutral or a failed high voltage neutral or both. Examples: A neighbors failed neutral will cause his current to go from his water bond through the water pipes in the street up through your water bond and go back via your intact utility neutral. Another scenario is a high voltage neutral that has completely deteriorated in an underground cable network or failed from say a broken tree limb hitting a pole, all current from the transformer primary then travels through the HV-LV neutral jumper, through your service, water bond and then back to the substation. Both these can go unoticed for years and usually do until work is done on water pipes, water meters, grounding system ect the danger makes itself known, sometimes even killing. Nothing like doing a service change (because the panel was an FPE:roll with the water bond arcing up a storm and then seeing the fire department arrive next door because the neighbors electronics started smoking:happyno:

And on a quick side note, when utilies use the neutral to double as a ground, low level ground fault protection goes out the window. Substation feeder breakers and reclosers no longer can be set to trip out at say 5 amps but instead set to trip at hundreds. Good luck detecting downed lines during storm:lol:

Now onto question number 2.

In my opinion water heaters should be required by code to have GFCI protection for the very reason you state along with dishwashers, clothes washer, ect. Unfortunetly attention has drifted to psedo-science that warants the use of AFCIs I guess it might not be a risk because the water heater is grounded as are the water pipes. But again GFCIs make a world of difference.

Sorry about the long post but I was in thinking mode. Hope my info was on the money.

 

[kwired]

And I think our obsession with water bonds was pushed the most by utilities. By adding a water bond at each service the utilities get a free super neutral and free lightning protection at the customer's expense. Utilites treat ground, earth and neutral as the same thing to save costs. The water bonds save a utility a larger neutral, ground rods, grounding grids, lightning arresters, headaches from broken neutrals, ect and as a result using single bushing transformers/primary wye grounded becomes easy with less worry for them.

How many more ground rods you thing the utility would use without any water piping bonds? I think they would use exactly the same amount.

How often do you see the utility using a smaller grounded conductor or even no grounded conductor at all because they know there is water piping between structures and they can just rely on that? My guess is it doesn't happen.

Utilities do not require bonding water piping NEC does, yes the utility does benefit some because of it at times.

I think we are required to bond water piping with a conductor sized as a grounding electrode conductor because if we don't there is a good chance the piping becomes bonded anyway via some equipment grounding conductor that is likely a smaller size. Maybe it is bonded but not solidly bonded just because it is in contact with something else that is grounded. Now if that piping happens to also have at least 10 feet in the ground it is a natural electrode in a way and if it is not bonded there is a chance that during a lighting event that raises voltage on the electrical grounding system you may have flash over between water pipe and anything bonded to the electrical system anyway.

In my opinion water heaters should be required by code to have GFCI protection for the very reason you state along with dishwashers, clothes washer, ect. Unfortunetly attention has drifted to psedo-science that warants the use of AFCIs I guess it might not be a risk because the water heater is grounded as are the water pipes. But again GFCIs make a world of difference.

Do you have any evidence or statistics to prove people are being electrocuted by failing water heater elements? If there were enough cases of this happening we probably would have had GFCI requirements a long time ago.

Water in the water heater tank typically has some conductivity, but not like metallic objects. Where is current going to flow in the case of a burned out heater element? Through the shorter path with less resistance near the break in the element (which also has 240 volts across it instead of 120) or through water to piping and through water only again to get into the tub that likely has a very high resistance to ground unless maybe it has metallic drain piping. If water piping is non metallic then you have an even higher resistance path between the water heater and the tub.

 

[mbrooke]

How many more ground rods you thing the utility would use without any water piping bonds? I think they would use exactly the same amount.
Id venture to say 2-8 times more if the soil is dry enough because the voltage rise to ground will be a driving factor. Some utilities might luck out though.

How often do you see the utility using a smaller grounded conductor or even no grounded conductor at all because they know there is water piping between structures and they can just rely on that? My guess is it doesn't happen.

It does happen everywhere all the time in terms of a reduced size MGN. Often a 3 phase trunk line will be 500MCM lets say but the neutral is only 4/0 or 2/0. 50% of the phases is often the norm. Even where the phase and MGN are equal in size theoretically a customer load a few poles down from the transformer bank will make it undersized. Consider a typical install of 2/0 phase and 2/0 MGN, 3 phase wye wye bank, 25kva each. 2/0 XLPE quadplex is run from the bank to several poles down then turning sideways to the customer; main line btw. What do most utilities do? Splice the MGN to the 2/0 quadplex messenger then after dropping the phases the messenger is re-spliced back into the MGN on to the next bank/set of poles. Any customer neutral current will reduce the capacity for the HV neutral return. Look at any single phase overhead latteral on a street feeding resi, does the LV have one neutral and the HV have another? Chances no. Look also at the size of the MGN compared to the phases, 90% of utilities 80% of the time will have a reduced size MGN over the phases. As for no MGN conductor at all: its not allowed by the NESC, earth can not carry load alone. Now, of course one can argue a balanced 3 phase load wouldn't need a full size MGN especially if the imbalance is a few amps, yes. But in the real world there is imbalance and lots; a redundant water bond systems will always do something to reduce that load which opens up the option of a smaller MGN without consequences to the utility.

Think 80amp 7.2kv single phase lateral feeding a subdivision on a hot day, with rods and water bonds that neutral current will now be 40 maby even 30amps, I see no issue in a reduced sized MGN from a thermal percpective, heck go full size and you reduce the voltage drop to the customers.






Utilities do not require bonding water piping NEC does, yes the utility does benefit some because of it at times.

The NEC does require it but why? Because utilities and other large high profit powers have an influence in the decision making logic of the code making panel, if you catch my drift. AFCIs got in the code the same way.

I think we are required to bond water piping with a conductor sized as a grounding electrode conductor because if we don't there is a good chance the piping becomes bonded anyway via some equipment grounding conductor that is likely a smaller size. Maybe it is bonded but not solidly bonded just because it is in contact with something else that is grounded. Now if that piping happens to also have at least 10 feet in the ground it is a natural electrode in a way and if it is not bonded there is a chance that during a lighting event that raises voltage on the electrical grounding system you may have flash over between water pipe and anything bonded to the electrical system anyway.

I agree.




Do you have any evidence or statistics to prove people are being electrocuted by failing water heater elements? If there were enough cases of this happening we probably would have had GFCI requirements a long time ago.

No, like I said its probablly because of the fact grounding reduces the potential for electricuiton. But, I have never heard of a GFCI failing a water loving appliance in terms of saftey.

Water in the water heater tank typically has some conductivity, but not like metallic objects. Where is current going to flow in the case of a burned out heater element? Through the shorter path with less resistance near the break in the element (which also has 240 volts across it instead of 120) or through water to piping and through water only again to get into the tub that likely has a very high resistance to ground unless maybe it has metallic drain piping. If water piping is non metallic then you have an even higher resistance path between the water heater and the tub.

Electricity takes all paths of resistance, if it took the least every time you turn on your stove the lights in your house would go out. Keep in mind its also where the element fails. If it failed in the center the resistance is about equal beween sides and assuming the wire is not brocken the potential will near zero. Even if it broke in the center Im thinking current cancellation would reduce the garadiant... but theory aside you are correct conductivity is small and in my opinion grounding does take the few amps leaking into the water.

 

[kwired]

I hate replying to quotes within quotes, but here goes:

Originally Posted by kwired

How many more ground rods you thing the utility would use without any water piping bonds? I think they would use exactly the same amount.
Id venture to say 2-8 times more if the soil is dry enough because the voltage rise to ground will be a driving factor. Some utilities might luck out though.

I still think they would drive one rod at every pole or structure like they usually do now. In fact most work I do is in rural areas with no municipal water piping and nothing is really done any different than within municipal water areas. One ground rod per pole or other structure for POCO and any electrodes connected to premises wiring is installed to NEC not to POCO specs.

How often do you see the utility using a smaller grounded conductor or even no grounded conductor at all because they know there is water piping between structures and they can just rely on that? My guess is it doesn't happen.

It does happen everywhere all the time in terms of a reduced size MGN. Often a 3 phase trunk line will be 500MCM lets say but the neutral is only 4/0 or 2/0. 50% of the phases is often the norm. Even where the phase and MGN are equal in size theoretically a customer load a few poles down from the transformer bank will make it undersized. Consider a typical install of 2/0 phase and 2/0 MGN, 3 phase wye wye bank, 25kva each. 2/0 XLPE quadplex is run from the bank to several poles down then turning sideways to the customer; main line btw. What do most utilities do? Splice the MGN to the 2/0 quadplex messenger then after dropping the phases the messenger is re-spliced back into the MGN on to the next bank/set of poles. Any customer neutral current will reduce the capacity for the HV neutral return. Look at any single phase overhead latteral on a street feeding resi, does the LV have one neutral and the HV have another? Chances no. Look also at the size of the MGN compared to the phases, 90% of utilities 80% of the time will have a reduced size MGN over the phases. As for no MGN conductor at all: its not allowed by the NESC, earth can not carry load alone. Now, of course one can argue a balanced 3 phase load wouldn't need a full size MGN especially if the imbalance is a few amps, yes. But in the real world there is imbalance and lots; a redundant water bond systems will always do something to reduce that load which opens up the option of a smaller MGN without consequences to the utility.

Think 80amp 7.2kv single phase lateral feeding a subdivision on a hot day, with rods and water bonds that neutral current will now be 40 maby even 30amps, I see no issue in a reduced sized MGN from a thermal percpective, heck go full size and you reduce the voltage drop to the customers.

I understand the theory behind smaller neutral where the load is not there. My question was intended to ask how often do you see a smaller neutral run than what would otherwise be run simply because they know there is a water pipe that will carry extra current? I don't think it really ever happens, or at least it should not happen.

Utilities do not require bonding water piping NEC does, yes the utility does benefit some because of it at times.

The NEC does require it but why? Because utilities and other large high profit powers have an influence in the decision making logic of the code making panel, if you catch my drift. AFCIs got in the code the same way

.

The practice of bonding water piping has been in code long before big business were commonly known as greedy and would possibly push for such a thing. And in those days privately owned power companies were not really around either.

Do you have any evidence or statistics to prove people are being electrocuted by failing water heater elements? If there were enough cases of this happening we probably would have had GFCI requirements a long time ago.

No, like I said its probablly because of the fact grounding reduces the potential for electricuiton. But, I have never heard of a GFCI failing a water loving appliance in terms of saftey.

I will agree that it can't hurt anything, but there statistically has not been any significant number of cases AFAIK to support the need to GFCI protect a typical water heater either, and like you said, if there was much risk the GFCI manufacturers would be on it like flies on stink.

Electricity takes all paths of resistance, if it took the least every time you turn on your stove the lights in your house would go out. Keep in mind its also where the element fails. If it failed in the center the resistance is about equal beween sides and assuming the wire is not brocken the potential will near zero. Even if it broke in the center Im thinking current cancellation would reduce the garadiant... but theory aside you are correct conductivity is small and in my opinion grounding does take the few amps leaking into the water.

I am aware that electricity takes all paths, there is likely current flowing through at many different times, but usually at such low levels you can not feel it. When you become lower resistance than some other path is when you will notice it, or if the source is stout enough that voltage does not drop significant enough when there is a lower resistance also in parallel to the path through you - which happens with premises wiring all the time, but the so called "stray voltages" is where impedance of paths makes a big difference in how much flows where and what it takes to shunt current nearly completely around something else.

 

[electrofelon]

Regarding #1, What I was getting at is, what are you thoughts on a non metallic coupling right where the pipe enters the structure? On the negative side we loose a grounding electrode, but on the positive side we dont have the stray current/current on stuff issue and danger to people working on the pipes. Which is better?

Regarding #2, I am not saying water heaters should have GFCI protection as I have never heard of it being an issue, I am just trying to understand the physics why its apparently a big risk in pools and spas but not water heaters. IT seems like the arguments made so far could be said of pools and spas also.

(heck one can not use a ground bar as a neutral bar at the service entrance because the current can flow over the enclosure to the bond screw)

I thought only the GEC had to land on the neutral buss/bar but grounded conductors can mix and match, am I mistaken?

 

[kwired]

Regarding #1, What I was getting at is, what are you thoughts on a non metallic coupling right where the pipe enters the structure? On the negative side we loose a grounding electrode, but on the positive side we dont have the stray current/current on stuff issue and danger to people working on the pipes. Which is better?

Regarding #2, I am not saying water heaters should have GFCI protection as I have never heard of it being an issue, I am just trying to understand the physics why its apparently a big risk in pools and spas but not water heaters. IT seems like the arguments made so far could be said of pools and spas also.



I thought only the GEC had to land on the neutral buss/bar but grounded conductors can mix and match, am I mistaken?

As it is worded, if there were an isolating fitting at the water entrance, you would have to bond around it anyway, plus if there is 10 feet or more of buried metallic water piping it is also worded that you must use it as an electrode.

If you had an isolating fitting you would still potentially have voltage between each side of that fitting.

With swimming pools we bond everything to prevent having voltage gradient zones small enough for users to expose themselves to voltages. Stray currents very well may flow through pool components but everything within reach of users has been bonded together to keep potential between those items the same.

With the water heater and a burned out element, not enough current leaves the unit via the water to develop dangerous voltages away from the tank.