Calculating Heat-Pump Water Heater Payback Period
I am due to replace my water heater soon, so I ran some numbers for a heat pump versus straight electric. I ran these a couple years ago, and ended up with quite different results. One difference this time is I used actual energy usage for the water heater over the last two years, versus projected use based on a typical family. But I also might have screwed something up, so I’m wondering if someone can check my work.
Data:
-1880kWh annual average use with electric water heater (for the analysis I actually did month by month)
-incremental cost of electricity: $0.13 (this is hard to parse exactly, since there are different contributors, some fixed, some tied directly to use but some variable. I took my lowest and highest usage months and compared the energy usage difference against the cost difference to come up with $0.13.
-water heater is inside the conditioned space, so in the summer this gives free cooling, in the winter it steals some heat from the minisplit.
-differential cost of HPWH vs electric replacement: $2200
Assumptions:
-HPWH will be ~3:1 COP
-COP will be de-rated by 34% during Nov-Mar (920kWh total usage)
-COP will be de-rated 17% during Apr and Oct (294kWh total usage)
-COP will be pro-rated by 20% during Jun-Sep (666kWh total usage)
Using this formula, I got an estimated annual savings of 1070kWh, which equates to $139 annual savings. If I add in savings by supplying the dishwasher with water from the HPWH (currently it is fed by cold water and heats the water internally), that might save another $20. $2200/$159= 14 years. If my numbers are correct, I can’t really justify the HPWH, considering it probably won’t last much more than 10 years.
Did I do something wrong? Are our hot water usage numbers really low? I remember previously coming up with a payback period of about 7 years.
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
The usage numbers seem reasonable enough, but do you have a water bill with total usage?
In regards to the $2200 difference - do you have a breakdown between equipment cost and installation cost?
I am on a well, so no water bill. I have an energy monitor that tells me exactly how much energy the hot water tank uses.
The HPWH is $2750, a standard electric is in the ballbark of $500. I apportioned nothing for install, as I'd do that myself.
Oh just based Home Depot I figured it was more like $1000. Can you fit a Drain Water Heat Recovery unit? You'd save nearly as much for like 1/3 the cost with a longer life.
I'm in Canada, where that $1200 water heater is $2457. A small part of that is exchange rate, part of it is that it's a slightly newer model, most of it is just the "because we can" upcharge. I was looking at the 65 gallon version, since it's only 10% more.
I already have a drain heat recovery. It's the biggest one I could fit, and it's only 45%, so it's not saving an equivalent amount as the HPWH. It also only affects showers; not baths, washing machine, kitchen sink or standby losses. Standby losses are about 1/4 of our entire hot water "usage".
Bummer! You didn't happen to be monitoring the heater prior to the DWHR installation did you? Would love to see that data
The drain heat recovery was part of the original build. I would have put one in anyway, but it's actually a code requirement here (one that I expect gets ignored in most builds).
I would argue that any use when you're running air conditioning is effectively free.
Yes, that's true. When comparing the relative cost vs the straight electric, the most I can save of the water heating bill is 100%. If I allow for the fact that the water heating is free, there is still electricity usage for cooling, whether that is done by the minisplit or the HPWH. Even in summer, there's quite a bit of time where we don't need to have air conditioning.
Right, if currently you're spending roughly $250 a year that's the most you could ever save and even that is an 11-year payback.
How many in your household? I have been wondering something similar - whether we should bother putting in a heat pump water heater, or just go straight to a smaller electric tank, given how low our usage is. We'd be switching from gas, so the calculations are different, but we're currently spending about $20-$25 a month on gas, and roughly half of that is the base fee, which would go away if we became an all-electric household. If our usage is similar to yours the monthly cost would be a wash, and honestly, even if it went up fifty percent, another ten bucks a month is not exactly going to make or break us.
We're two adults and two young kids. It's certainly conceivable our usage will go up as the kids get older, but I see that as 5-10 years down the line. Your cost for the water heater is probably much different than mine.
Your seasonal derating numbers need to be justified with some real data, not a WAG. A 34% hit for the winter season seems way high.
A UEF of 3.1 is not the same as a COP of 3.1. At higher room temps than the 65-70F band used in a UEF test the COP is higher, and when the room temps are <<65F the COP will be lower. But unless the average room temp during the wintertime use colder than 55F it's not likely to be a 34% hit.
Unless you have a 10 year contract for 13 cent electricity it's silly to presume that the future price (in inflation adjusted dollars) will be the same as last month's price.
Good point on future pricing, though if I tried to take that into account, I'd have to contradict your other advice about taking a WAG at what pricing might be like for the next 10 years.
To clarify, I wasn't assuming the COP would go down due to to lower ambient temps, I was just trying to account for the fact that heat taken away by the water heater has to be replaced. This lowers the amount of energy being saved. Maybe 34% was too big a number, but I don't think it's that far off. If the makeup heat was straight electric, the HPWH wouldn't be saving anything in the heating season (100% de-rating in savings). I assumed a COP of 3:1 for the minisplit, which is how I arrived at 34% energy saved by the water heater would go to running the minisplit to recover that heat.
>" I was just trying to account for the fact that heat taken away by the water heater has to be replaced. This lowers the amount of energy being saved. Maybe 34% was too big a number, but I don't think it's that far off. If the makeup heat was straight electric, the HPWH wouldn't be saving anything in the heating season (100% de-rating in savings). I assumed a COP of 3:1 for the minisplit, which is how I arrived at 34% energy saved by the water heater would go to running the minisplit to recover that heat"
That model is only accurate if the HPWH is located in fully heated conditioned space. If it's in an unheated (even insulated) basement it will drop the basement temp a few degrees, lowering the overall heat loss out of the house.
The NEEP coalition of utility partners in the Pacific Northwest have modeled and third-party tested this extensively (mostly with EF 2.something earlier HPWH models). Some of that stuff is probably still online, but would have to be adjusted to reflect the much higher COP of current offerings.
Even circa 2010 HSPF 10 mini-splits tested a seasonal COP 3 even at the warm edge of zone 6B (NEEP has that online somewhere too. Current cold climate mini-split models run at a COP north of 3 even at +17F outside, unless undersized and running full-on max at that temp. eg:
https://ashp.neep.org/#!/product/34607
https://ashp.neep.org/#!/product/25334
Most US climate zone 5 regions have an binned hourly average in January of +25 or better:
https://weatherspark.com/m/14091/1/Average-Weather-in-January-in-Chicago-Illinois-United-States#Sections-Temperature
https://weatherspark.com/m/21245/1/Average-Weather-in-January-in-Rochester-New-York-United-States#Sections-Temperature
https://weatherspark.com/m/3709/1/Average-Weather-in-January-in-Denver-Colorado-United-States#Sections-Temperature
https://weatherspark.com/m/26197/1/Average-Weather-in-January-in-Boston-Massachusetts-United-States#Sections-Temperature
Interpolating the mid-modulation efficiencies at +17F & +47F. at roughly +32F (the average hourly January temp in Boston) the COP will be about 4 when correctly sized for a 0-10F outside design temp.
So only in the cold half of zone 6 or into zone 7 should the winter derating assume a COP as low as 3 for the mini-split, even if the HPWH is in fully heated space. (If you're talking multi-splits, yes it might be that low.)
"That model is only accurate if the HPWH is located in fully heated conditioned space." It would be, we have no basement.
Regarding the minisplit efficiency, a seasonal average COP of 3 seems about right for mine. I have a Fujitsu 9RLS3. This thing looks great if you just look at HSPF. However, it seems to have gotten that good HSPF number mostly from the 47degF performance (COP of 4.5-5.3) . That's probably great for a code built house. It almost never runs at that temperature in my house. No heating is required when it's that temperature outside; body heat, waste electrical and passive solar gain takes care of it. It's about 3.5 at 17F (at the estimated load), but it falls off a cliff at temperatures below that. But even if I changed my average COP for the minisplit to 3.5 or 4, it doesn't move the needle very much in terms of potential $ savings.
>"However, it seems to have gotten that good HSPF number mostly from the 47degF performance (COP of 4.5-5.3) . That's probably great for a code built house. It almost never runs at that temperature in my house. No heating is required when it's that temperature outside; body heat, waste electrical and passive solar gain takes care of it."
With shoulder season heating loads as low as yours the additional load of the HPWH improves the mini-split's efficiency, but it's still a modest increase in power use, no doubt.
What is your mean January temp, approximately?
>"But even if I changed my average COP for the minisplit to 3.5 or 4, it doesn't move the needle very much in terms of potential $ savings."
Getting subsidy or higher electricity price inflation might tip the balance, but for sure it's never going to be a real money-maker for low use situations. With use as low as yours the lifecycle is likely to exceed the 10 year warranty period, possibly by 2x. (The same would be true of a plain old electric tank.)
In my area the levelized cost of rooftop PV is now dipping below 13 cents/kwh (even without factoring in tax incentives, etc) and there doesn't seem to be a real bottom to PV costs. Even if grid prices rise it's likely that PV + battery will be able to beat that within the next decade.
Mean January temp here is about -7C.
I also see prices around $1500 for a HPWH. And since the guarantee is 10 years, my guess is a 14 year average life.
On the other hand, I shower at 1 GPM and some hot water comes from a POU heater.
It's all a good reminder to not accept conventional advice - run the numbers for your case.
The 10 year warranty doesn't inspire any confidence in me. Considering that only covers parts, I'm sure they are counting on the fact that most people will balk at upwards of $400 for labour, $100-150 of which would be just for them to show up and diagnose the problem.
My electricity is a bit more at $0.14 and I couldn't get any reasonable ROI for a HPWH. The big issue is the silly cost of these up here. Once the borders open up and can pick it up from the US for 1/2 the cost, it can be made to work.
I tried to make this idea make sense a couple of years ago, and couldn't. One problem is that the big price difference only exists for the 50 gal size. Right now it's $1229, whereas it was $999 when I looked a couple of years ago. The cheapest one at HD Canada is $2457, but it's not exactly a fair comparison because the $1229US one is a discontinued (or at least soon to be discontinued) model. The apples to apples model in the US is $1514 ($2014C). But let's say you don't care about that detail. $1229US is about $1635C. That's an $800 savings, which is a lot. But if I factor in gas, what my time is worth, the fact that I'd have no warranty, and finally that it's not really the size I want, then it's far from a slam dunk. It's marginally better, at best. If you're within half an hour from the border, or going over there for another reason already, that changes a lot. If you want bigger than 50gal, the price difference pretty much goes away. 65gal is $1950US, here it's $2726C; almost the same. For 80gal, it's actually cheaper here by a couple of dollars. Three years ago I was waiting for the prices to start falling. That really hasn't happened. The prices in the US have gone up, making the Canadian prices look not as bad by comparison, but that doesn't help the bottom line of whether it's worth it or not. Right now the only things that might sway me are the likelihood of electricity prices going up, and likelihood of our usage going up. Considering either of those, nothing makes more sense than doing nothing until the current water heater fails. At least from an economic standpoint. For obvious reasons, I'd rather replace the water heater before it fails.
The larger HPWH don't make any financial sense in any of the markets. I don't understand why they charge so much for one.
The lowest cost option is get the 40gal unit, plumb it in before your existing tank and set the HPWH to run a bit hotter than your electric unit. This does increase the standby loses (a blanket on the tank will help) but most of your hot water demand will be supplied by the HPWH.
As for the life of the existing tank, the lowest cost thing to do is replace the anode rod. Not fun the first time around but much easier the next time once you have teflon tape on the threads. My repurposed electric tank that I use as an indirect is coming up on 20 years old and 3rd anode rod.
> coming up on 20 years old and 3rd anode rod.
Maybe switch to a powered anode rod, available here for < $100.
“Akos
The larger HPWH don't make any financial sense in any of the markets. I don't understand why they charge so much for one.”
When I look at the sales literature for the Rheem line of HPWH, to my eye it seems like all the units have the same compressor and produce the same number of BTUs per hour regardless of the tank size. Note the smallest one 40 gal? Seems to have a different fan and maybe coils so it can live on a smaller diameter tank.
Walta
Interestingly, the $1229, soon-to-be discontinued water heater at Home Depot went up to $1399 just two weeks later.
I installed a Rheem 50 gallon about 2 1/2 years ago replacing a 19 yo electric. My price at the time was $1299 including tax. A 50 gal straight electric was about $450. I'm in CZ2, central Texas and the heater is in the garage. In the summer (which is 9 months of the year) garage temps will be in the mid to upper 80's, which increases the heaters efficiency. I also got a $850 city rebate, making the heater pay for itself before it was installed. Install of a new heater, heat pump or regular I figured would be a wash, as the old unit wasn't even close to being up to code. Also took the opportunity to add a moisture sensor that automatically shuts off the water to the heater if a leak is detected. So far has worked flawlessly. Unfortunately don't have a way to figure out individual electrical usage by component.
They essentially gave you the HPWH for the same price as straight electric. It's paid back the second you turn it on. I wish we had rebates like that, or any for that matter.
I am in a similar situation. I would like to invest in a HPWH, but hard to justify at $3k. Our usage is a bit more than yours, but ROI is still hard to justify (we are at 6.5kwh daily average). The new Canada Greener Homes Grant that recently came out might help some. I believe you can get $1000 back for HPWH install. I think it has to be installed by a professional, which for a DIY may cancel out the $1000 grant. I tried to get a quote from a local rheem dealer but he has never installed one and didn't want to start now. There is also the required home energy assessment before and after. It's another expense and time if you aren't looking to get one anyways ($500 back through the grant for the assessment but that didn't cover the entire fee for me). I've already maxed out the 5K return you can get through the grant for heat pump/air sealing/insulation improvements so I am hoping for another oppourtunity to come up. I did listen in on a few presentations from some GTA municipalities that are hoping to launch similar grant programs as part of their climate change planning. Fingers crossed these take off across the province and compliment the federal grant program. Thanks for sharing your analysis.
Yes, that rebate program is kind of ridiculous. Seems more like a way to subsidize energy evaluators than anything else. There's a pretty significant risk of getting robbed. You have to pay for an evaluation, and they they tell you what you should do. If you don't do at least one of those things, then you don't even get the rebate against the evaluation. So for example, let's say I pay $1000 for an evaluation. They decide not to recommend the heat pump water heater I want, but instead recommend external insulation that I don't need. Now I'm out $1000 for literally nothing. Why would they do this, you may ask? Well, it goes back to my suspicion that the entire thing is just a way to get people to pay for home energy evaluations. If they actually cared about giving people rebates, there wouldn't be strings attached like this. Either let people to decide on their own what they need, or offer the evaluations at no charge, or a nominal charge of $50-100 to discourage tire kickers.
Another interesting question to ask is which one is better HPWH or PV.
I think at the current cost PV, the incremental cost of panels to cover the energy difference for a HPWH would be less. Plus no compressor noise, filter cleaning or maintenance.
Good point. However, it rests on your ability to connect to the grid. Once you add batteries, charge controller and an automatic switching system, the price quickly becomes impractical.
There are some DIY kits for connecting PV directly to tank hot water heaters - serving as a proof of concept. Perhaps this should become mainstream.
The ones I've seen just use resistive heating, which is less efficient than a thermal collector. PV solar really shines when it is connected to a heat pump. Unfortunately running a motor requires a lot cleaner power than running a resistive heater.
Maybe, but this topic is a good reminder that sometimes less energy efficient but less expensive upfront wins.