Actual cost to heat water

In March I concluded that a timed water heater 12 hours On/Off saved 6.25 KwH per month. See Does a water heater timer save, a case study

I have also calculated that my cost to heat water is 1.3606 cents per gallon

When I found myself with an extra water monitor after I upgraded to a Flume water monitor. I placed the extra monitor at the water heater and bumped the gallons used against Sense data to verify these conclusions. I have found the actual cost is 2.06222 cents per gallon. Obviously, the cost to keep the standing water hot is driving this. This does support a tankless heater in my situation over the long haul as I would DIY installation. The average annual savings $74.13

To be fair I have very low water consumption, averaging 23.3 gallon per day per person.

Here is the formula to calculate to cost to heat water:
It takes .000293 Kwh to raise 1 pound of water 1° F
A gallon of water weighs 8.34 lbs.
Average temp tap water for me is 72 degrees
Water heater set at 120 degrees so 48 degrees of rise
Efficiency of heater .95 (95%) For 2015 and later models .90 pre 2015
Ave price of electricity .116 KwH
Formula: .000293 X 8.34 X 48 X .90 X .116 = $.013606 per gallon


I really love when you share these insights. Keep them coming!

And feel free to cross-post them over at

There is no comparison with a tankless system compared to a storage tank. Vendors are even helping people move away from the storage tank. They don’t however consider solar as a possible solution. If you have a relatively new storage tank you can do a lot of different options. One is to disconnect the wiring connecting the top element with the bottom and add a DC element to the bottom ($65). This way you don’t need the expense of an inverter and can connect directly to the element. Some have had to buy an additional relay adding another $50 but it’s not a bad solution. The top element is still AC powered so you never are without hot water.
I am running old school solar thermal but a PV-to-storage is the better option. Not as clean as a tankless particularly a gas tankless, but if you are running an electric tankless it is a viable option.

1 Like

Interesting! I was just doing a related calculation last weekend:

We turned on an electric water heater that we only use when we have visitors. The room was at ~20 C and I assumed the heater would go to 40 C or so. Then the energy needed is:

19.9 gallons * 3.8 liters / gallon * 1 kcal/ C x 20 C = 1512 kCal = 1.75 kWh

With Sense I measured the actual power use as 1950 W for 1:03, so 2 kWh total.

Our electricity costs 11.7 c / kWh, so the cost per gallon is:

2 kWh * 11.7 c / kWh / 19.9 gallons = 1.17 c / gallon = $0.0117 per gallon

Which is quite close to your number of $0.0136

Of course this is the cost to heat the water once, then if you don’t use the water, the heat keeps coming on every 6 hrs or so to keep the water hot.

I think that depends upon your perspective.

Existing solar-thermal is a lot more efficient (literal efficiency) than existing PV in terms or raw solar irradiance watts-to-hot-water conversion. To a certain extent solar thermal also seems like simpler technology. Much of the world was using solar thermal for hot water long before PV was an option. Solar thermal (hot water) is also fundamentally an energy store so as isolated systems solar thermal has another advantage.

Ever taken a heavy garbage bag and filled it with water and hung it from a tree in the mid-day sun? Grab a fork for a low tech shower!

I can see the rationale behind using both systems.

That said, the idea of using PV to feed DC to an existing tank element is genius and although PV is less efficient than solar thermal, having PV-only has some clear advantages … especially, perhaps, in scaling the industry more rapidly.

Are you guys factoring in Utility delivery costs? My electricity cost (delivered) was 23.5c/kWh in January and 24.3c/kWh (really!) in May. If we’re comparing costs with PV, delivered cost seems like a fairer comparison.

Oh, dreaming of PV in NYC!