confused about water heater wiring standards

[RustyShackleford]

I'm confused about what seems to be the usual standard for a conventional (storage-tank style) electric water heater: 4500 watt non-simultaneous heating elements, 10 gauge wiring, 30 amp breaker. At 240 volts, that 30 amps translates to 7200 watts, and with the 80% factor for continuous loads, that's leaves 5760 watts. Or, 125% * 4500 watts, divided by 240 volts is about 23 amps. So why isn't the standard heating element 5000 watts (or more) ? I could also ask why the standard breaker size isn't 25 amps, though I guess that's not really a standard size. Maybe it's simply that they want 30amps/10gauge to work even if voltage is only 208v ?

 

[Sierrasparky]

I'm confused about what seems to be the usual standard for a conventional (storage-tank style) electric water heater: 4500 watt non-simultaneous heating elements, 10 gauge wiring, 30 amp breaker. At 240 volts, that 30 amps translates to 7200 watts, and with the 80% factor for continuous loads, that's leaves 5760 watts. Or, 125% * 4500 watts, divided by 240 volts is about 23 amps. So why isn't the standard heating element 5000 watts (or more) ? I could also ask why the standard breaker size isn't 25 amps, though I guess that's not really a standard size. Maybe it's simply that they want 30amps/10gauge to work even if voltage is only 208v ?

Good question.
My thought would be that back in the day 240v water heaters were placed on only 20 amp circuits
as 4500 /240 is actually 18.75. So without the continuous load factor it worked out. Manufactures of tank water heaters have never changed basic design. I would suppose they did not want to make a element that was 5500 watts using the same thread as a 4500 unit.
Just my opinion.

 

[Dennis Alwon]

I have seen water heaters that were rated 5000 watts but I assume it is a standard because the power company's want the minimum draw and still have a reasonable heat time.

Imagine the extra load on the power company if all water heaters were 5000 watt

 

[texie]

I'm confused about what seems to be the usual standard for a conventional (storage-tank style) electric water heater: 4500 watt non-simultaneous heating elements, 10 gauge wiring, 30 amp breaker. At 240 volts, that 30 amps translates to 7200 watts, and with the 80% factor for continuous loads, that's leaves 5760 watts. Or, 125% * 4500 watts, divided by 240 volts is about 23 amps. So why isn't the standard heating element 5000 watts (or more) ? I could also ask why the standard breaker size isn't 25 amps, though I guess that's not really a standard size. Maybe it's simply that they want 30amps/10gauge to work even if voltage is only 208v ?

Don't really know how 4500 watts came to be the most common size. You have a point in that it would cost the same to wire a 240 volt, 4500 or 5000 as they both can use #10 and a 30 amp OCPD. 5000 watt water heaters are available, just not as readily as a 4500. BTW, the only reason you can use a 30 amp OCPD instead of the next standard size of 25 amp on the 240 volt, 4500 watt heater is because of 422.11(E)(3).
Your 208 volt comment may have some validity. Thing is though it is relatively rare to see to see 208 rated water heaters. Sadly most just use a 240 volt heater on the 208 system and wonder why there is not much capacity. Sure makes for long lasting elements though.

 

[RustyShackleford]

It's worth noting that Bradford-White allows a multitude of options for wattage and voltage, on most of their models. From my limited experience, there's not much of a cost surcharge, but significant lead-time penalty (even though it simply means changing out the elements, though I guess you could buy the other elements and change 'em yourself).

Thanks for confirming I'm interpreting the NEC correctly.

 

[kwired]

Imagine the extra load on the power company if all water heaters were 5000 watt

More to it then just being 4500 or 5000 watts. Still takes the same kWhrs to raise 100 gallons of water 60 degrees, you just get it done a little faster with a 5000 watt element then with a 4500 watt element.

 

[RustyShackleford]

More to it then just being 4500 or 5000 watts. Still takes the same kWhrs to raise 100 gallons of water 60 degrees, you just get it done a little faster with a 5000 watt element then with a 4500 watt element.

Yeah, but peak load is a big deal for power generation companies. When instantaneous (peak) load increases, they must bring less efficient generation online, increasing cost and/or CO2. That's why peak-hour rate plans are good not just for customers' pocketbooks, but for environment. Energy storage is a huge issue (esp. for renewables) and millions of electric water heaters represent a huge amount of energy storage.

 

[GoldDigger]

Yeah, but peak load is a big deal for power generation companies. When instantaneous (peak) load increases, they must bring less efficient generation online, increasing cost and/or CO2. That's why peak-hour rate plans are good not just for customers' pocketbooks, but for environment. Energy storage is a huge issue (esp. for renewables) and millions of electric water heaters represent a huge amount of energy storage.

But a higher peak load for an individual heater element translates to a higher peak load for POCO only if there is effectively no diversity factor across tens of thousands of water heaters.
There might be a real problem if everyone had an off peak timer on their water heater which turned on the element at the same time of day, or if for some reason everyone decided to take a shower at the same time. (Possible if there were a universal 9 to 5 work schedule and identical commute times maybe?)

There can be really significant (on a local distribution basis) peak effects from electric demand water heaters, but little effect back at the generating plant.

 

[jumper]

Y'all over thinking this, we use #10 wire because most resi and a fair amount of small commercial WHs are wired with NM-B, and a 30A breaker because finding a 25A one can be a PITA and most AHJs are just fine with it, only 25A breakers I will get are for HVAC units.

 

[RustyShackleford]

Y'all over thinking this...

That's for darn sure.

 

[kwired]

Y'all over thinking this, we use #10 wire because most resi and a fair amount of small commercial WHs are wired with NM-B, and a 30A breaker because finding a 25A one can be a PITA and most AHJs are just fine with it, only 25A breakers I will get are for HVAC units.

That's for darn sure.

Not really, if NEC required us to protect it at 125% then we would have to use 25 amp protection devices. But as mentioned earlier by texie - check out 422.11(E)(3). It allows us to protect it at 150% which we can go next standard size - so for 4500 watts @ 240 that would be 30 amps.

Many people probably just figure that since it needs 10 AWG conductor that means they can protect it with 30 amp device - but that isn't exactly why we can protect it at 30 amps.

 

[Dennis Alwon]

Exactly as kwired stated. Water heaters overcurrent protective device can be 150% of the load and then you can use the next size breaker.

Not sure when that rule changed but 20 years ago one of the jurisdictions around here made us use 25 amp breakers

 

[JFletcher]

PoCo's first thing they offer rebates for doing load monitoring/shedding is HVAC. Second is water heaters.

4500W/240V is 18.75A, which because of continuous load cannot* run on a 20A breaker or #12 wire. Sure we could use 5000W on the same wiring/ocp as 4500W. But to be frank, 5000W is fixing a non-problem. You want hotter water, adjust the thermostat. More hot water, get a bigger tank. 5000W is an 11% increase in heating over 4500W, not exactly a huge difference. If you get 20 min of hot water out of your tank before it goes lukewarm, 23 minutes isnt a huge increase if you want 40 or infinite.

*cannot legally now anyway. I've seen at least 2 full size/capacity water heaters wired with 14/2 on a 20A breaker that never had electrical problems; 3' of free air cable to a breaker, heck you could probably use high temp 18ga and not set the place on fire. I have to wonder if the water heater co's keep conservative numbers because of 12/14 Cu installs or water heaters wired with old #10 Al.

 

[kwired]

PoCo's first thing they offer rebates for doing load monitoring/shedding is HVAC. Second is water heaters.

4500W/240V is 18.75A, which because of continuous load cannot* run on a 20A breaker or #12 wire. Sure we could use 5000W on the same wiring/ocp as 4500W. But to be frank, 5000W is fixing a non-problem. You want hotter water, adjust the thermostat. More hot water, get a bigger tank. 5000W is an 11% increase in heating over 4500W, not exactly a huge difference. If you get 20 min of hot water out of your tank before it goes lukewarm, 23 minutes isnt a huge increase if you want 40 or infinite.

*cannot legally now anyway. I've seen at least 2 full size/capacity water heaters wired with 14/2 on a 20A breaker that never had electrical problems; 3' of free air cable to a breaker, heck you could probably use high temp 18ga and not set the place on fire. I have to wonder if the water heater co's keep conservative numbers because of 12/14 Cu installs or water heaters wired with old #10 Al.

Most of those don't run a long enough cycle to trip breaker or overhead conductor insulation is why they do last, would be my guess.

 

[JFletcher]

Most of those don't run a long enough cycle to trip breaker or overhead conductor insulation is why they do last, would be my guess.

One of those 14ga/ 20A OCPD wired heaters was at my grandmother's house. She wanted an 80gal tank, in 1970, and got it (she hated running out of hot water, ever). About three years ago, I saw it was wired with old cloth NM, and 14ga at that. I can assure you that when the heating elements calcified, that heater ran a LOT. For a long time. Years perhaps (or perhaps the thermostat was equally calcified). Nevertheless, I changed out that ungrounded, undersized cloth NM for #10 NM and put it on a 30A breaker.

There was all of 4' of wiring between that heater and the panel. Free air. Aside from the fact that it doesnt meet today's code, there was no physical damage to it from overheating. I wouldnt want to put 4 or 5kw on #14nm, but it worked.

I remember a post here years ago re: a man who was pulling 60+A thru #12 THHN with no problems. The catch? Free air in Northern Alaska, where ambient was ~-40*F.

NEC doesnt cover extraordinary circumstances, and rightly so. Conservative is the way to go.

 

[kwired]

One of those 14ga/ 20A OCPD wired heaters was at my grandmother's house. She wanted an 80gal tank, in 1970, and got it (she hated running out of hot water, ever). About three years ago, I saw it was wired with old cloth NM, and 14ga at that. I can assure you that when the heating elements calcified, that heater ran a LOT. For a long time. Years perhaps (or perhaps the thermostat was equally calcified). Nevertheless, I changed out that ungrounded, undersized cloth NM for #10 NM and put it on a 30A breaker.

There was all of 4' of wiring between that heater and the panel. Free air. Aside from the fact that it doesnt meet today's code, there was no physical damage to it from overheating. I wouldnt want to put 4 or 5kw on #14nm, but it worked.

I remember a post here years ago re: a man who was pulling 60+A thru #12 THHN with no problems. The catch? Free air in Northern Alaska, where ambient was ~-40*F.

NEC doesnt cover extraordinary circumstances, and rightly so. Conservative is the way to go.

Sorry but when the element gets buried in sediment - it usually burns out because it can't get rid of heat fast enough, and the thermostat on a typical electric water heater is on external side of tank.

 

[RustyShackleford]

Most of those don't run a long enough cycle to trip breaker or overhead conductor insulation is why they do last, would be my guess.

Well, as far as the 125% higher OCPD requirement for continuous loads, doesn't Article 100 define "continuous" as "expected to run for 3 hours or more" ? For typical storage-tank water heaters, the thing is going to easily heat up a full tank of cold water in less time than that. I guess you could leave the faucet running, but ... "expected".

We discussed this a bit, and also with tankless, at:

http://forums.mikeholt.com/showthread.php?t=178316&highlight=

 

[iwire]

There is no way I am going to believe the selection of 4,500 watts by manufacturers had anything at all to do with helping the power companies.

 

[iwire]

Well, as far as the 125% higher OCPD requirement for continuous loads, doesn't Article 100 define "continuous" as "expected to run for 3 hours or more" ? For typical storage-tank water heaters, the thing is going to easily heat up a full tank of cold water in less time than that. I guess you could leave the faucet running, but ... "expected".

We discussed this a bit, and also with tankless, at:

http://forums.mikeholt.com/showthread.php?t=178316&highlight=

For storage water heaters the defintion is irrelevant due to this

422.13 Storage-Type Water Heaters. A fixed storage-type water heater that has a capacity of 450 L (120 gal) or less shall be considered a continuous load for the purposes of sizing branch circuits.

 

[kwired]

For storage water heaters the defintion is irrelevant due to this

There are some other items that are similar - they want 125% conductor size whether the load is truly continuous or not. Motors is one of them (in most instances).