On Saturday, December 15, 2018 at 12:18:01 PM UTC-5, danny burstein wrote:
In trader_4 writes:

Continuous is defined in the code as an expected load over 3 hours.
If the code didn't address water heaters separately and essentially
treat his 20 gal WH as a continuous load, I would argue that it isn't
a continuous load at all.

au contrare. If the water heater is starting
with a full tank of cold water, hmm, let's see...

20 gallons (yeah, I'll stick with US units)
weighs about 160 pounds.

If we take a value of, oh, fifty degrees F for
the incoming cold water, and want to raise
it to 125F, that's a 75 degree change

160 pounds [enter], 75 [times] = 12,000 BTUs.

conversion of watts - BTU is about 3.5

12,000 [enter], 3.5 [divide] = 3,400 watt-hours

He's got a 2,000 watt heater element, so that's
about one and 3/4 hours.




So yeah, if I were designing specs (or insalling
the wiring, etc.) for a water heater, then
I'd do the derating to 80 percent.


You can do whatever you want, but code says a continuous load is 3 hours plus. That is the case for continuous loads, but they spec water heaters separately, which I previously cited.
I'd sleep well knowing the physics and using a 20 amp breaker, regardless of the code. Even applying the 80 percent rule, a 20 amp circuit is good to 16 amp and this is 16.6 and that's if it runs continuously. BFD.





(Of course, that's a non issue since Code does,
in fact, have a specific entry for water heaters.
Thanks for that pointer. It had washed out of
my memory).

--
__________________________________________________ ___
Knowledge may be power, but communications is the key

[to foil spammers, my address has been double rot-13 encoded]