What is a minisplit’s lifespan?
A friend of mine here in Virginia (climate zone 4) heats his 1700 square foot house with a 50 gallon electric resistance hot water heater hooked to radiant coils in his floor. It is a tight, new house with cellulose between 2×6 studs. He generates 90% of the energy he needs for his all-electric house from a 6 kw solar array.
I am helping him build another, similarly-sized and constructed house. He wants to use the same radiant floor approach in the new house because it is very cheap and very durable. He is not interested in central air conditioning, and will just install a window unit if the climate becomes unbearable. He does not want to use a minisplit to because he believes that they are not made to last and that the embodied energy and cost of replacing a minisplit is higher than the energy/cost savings it would yield vs. electric resistance heat.
As I read GBA, I have noticed a high regard for minisplits due to their extreme efficiency. However, I wonder about this longevity question and the embodied energy (and financial cost) of replacing a piece of very high technology when it breaks. Is there a life expectancy for a minisplit, or is it totally dependent on the model, brand, and treatment they receive? Has anyone factored the embodied energy and lifespan of minisplits vs. resistance heaters into their calculations? For a small, efficient house, I would think the savings from a simpler, cheaper resistance heating system could easily offset purchasing additional solar panels to make up the extra electricity needed to run the resistance heat.
This friend of mine is very concerned about climate change and lives a very energy-frugal lifestyle (bikes everywhere, dries clothes on a line, etc.) He is okay with temperature differences between different rooms in his house, and he is okay with sweating in the summer. For someone like him, I think the resistance heat approach makes a lot of sense. But I’d love to hear arguments and evidence to the contrary!
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
Timothy,
Q. "What is a minisplit’s lifespan?"
A. Most experts consider that the lifespan of a minisplit is about the same as that of a split-system central air conditioner, widely reported to be between 15 and 20 years.
Timothy,
The question of minisplits vs. electric resistance + PV has been discussed several times at GBA, including here: Can Solar Electricity Trump a Ductless Minisplit?
What is the lifespan of the electric water heater he's using to heat the floor? How about the lifespan of the hydronic pumps for the radiant floor?
It's not significantly different than the lifespan of a mini-split, and likely to be shorter, ESPECIALLY if potable water is being used in the heating system. An "open system" using potable water for heating is allowed in most states, banned or restricted in others. Some states don't allow a hot water heater to be used solely for space heating, but will allow them for combined heating & potable hot water. Small electric boilers can be pretty rugged, but you're still talking an anticipated lifecycle of a couple decades not five.
How about the lifecycle of that window-shaker AC unit? Again, probably shorter than a decent mini-split.
In Zone 4A it's actually not super expensive to take the full-on PassiveHouse approach, which would bring the heating electricity use down to "who cares" territory, if desired. If the goal is Net Zero Electricity it might be do-able with 2x6/R20 cellulose and a roof-load of PV, but it's usually cheaper on a lifecycle basis to go higher R with somewhat less PV, and using heat pumps. The service life of a high-R building envelope can easily top a century (or at least all but the windows) if done right.
In zone 4A a PTHP can work out pretty well on a higher-R house, at a much lower up-front (and replacement) cost than a mini-split, at less than half the power use of a radiant floor running off the water heater, and he wouldn't need the window shaker AC.
Related comments: note that with his concern about climate change, the calculations can't be done in kWh or $. Water and concrete (a guess) plus a tolerance for some temperature swing means that he has a large storage "battery" - much more so than if using a mini split. Such storage effects time of electricity use which effects atmospheric carbon.
A more apples-to-apples comparison would be a higher cost hydronic heat pump.
Thanks for the responses! I appreciate the feedback.
Regarding lifespans: I should note that he's coming off an experience in a previous house where a brand-new, professionally installed $4,000 heat pump died after 4 years and the installer told him it wasn't worth fixing. Obviously, that's an outlier, but even if his hot water heater and pumps go out on his hydronic system, that's a replacement cost in the hundreds of dollars, not thousands. Same with the window-shaking AC unit - it's an off-the-shelf, homeowner fix for a fraction of the cost of a new minisplit.
The hot water tank is allowed to be a stand-alone system (not tied into potable water), which sounds like a plus. The pipes are not going through a slab, but in the floor joists between an insulated basement and the main floor, so there isn't the same amount of thermal storage that a slab would offer. Still, a hot water tank at least provides a bit of a thermal battery.
Regarding building envelope and PV: We're not in the passive house range, but we are using Zip R6 sheathing, so the total wall R-value will be a bit higher than R-20. And I incorrectly stated that his existing house has 6 kw of pv - it's actually 4.5 kw and producing 90% of his usage for a family of 5 in an all-electric house. (Like I said, he is frugal when it comes to energy!)
Another thing that skews our calculations in favor of a cheaper heating system is that we know how to do PV on our own. The last system I installed came in at $1.60/watt, including paying myself and a few helpers for a day. So even a 6 kw system should be less than $10,000.
Timothy,
If you can install a PV system inexpensively, and you live in a region of the country where the local utility offers a good net-metering contract, and your house has a thermal envelope that is better than average, then the argument in favor of PV + electric resistance space heating starts to make a lot of sense.
Personally, I would favor electric resistance baseboards (or electric resistance wall panels) over a hydronic heating system -- for simplicity of installation and reduced costs.
Even buck a watt solar costs a grand per kw, and every kw of heating efficiency saves on the order of a kw of PV.
If he's able to heat the place comfortably with a standard 4.5kw water heater, it's also within range of a 1.25 ton (15,000 BTU/hr) PTHP, or even a 1-ton, which can be replaced for less than a grand, using less than half the annual power of the radiant floor + window shaker, and probably at a lower installed cost. A 1.25 ton name brand PTHP runs ~$800-900 for the unit, another ~$100 for the wall sleeve. A 3/4 tonner isn't a whole lot cheaper, but it's always better to right-size it for maximum comfort & minimum noise. The warranty on them is usually only 5 years, but it's a competitive market (it's a standard hotel room solution manufactured in high volumes/low margins), but they should go at least as long as a water heater or window shaker. They all seem to have scroll compressors now too, making them considerably quieter than the average window unit.
If it's still in the design phase, optimizing the window specifications by their orientation is a cheap way to lower the overall power use and maximize comfort too. When the glass specification changes based on the orientation, it's also better to change the window SIZES, to make it a lot harder for the contractors to screw it up, such as putting the high-gain glass on the west side, raising the peak cooling load, and the low gain glass on the south side cutting the potential wintertime heating offset gains by half or more.
Run the heating and cooling load calculations, and adjust them as the design changes. Better yet, download BeOpt (a US DOE freebie), a purpose-designed tool which makes modeling the impacts of design tweaks and keeping track a lot easier.
Thanks for the calculations, Dana and Jon. We'll look into the PTHP option. Do you need to have special HVAC equipment to install them?
Installing a PTHP is a bit like installing a window air conditioner- no specialized tools required. The refrigeration is completely self-contained, with exterior and interior side coils, compressor, blower, etc all plumbed together and pre-assembled inside the unit, just like a window AC. The primary difference is that it's intalled in a wall sleeve rather than a window, making it more inherently air tight, without taking up valuable window area.
Ignore the BTU/square foot rule of thumb in this installation video, which would GROSSLY oversize it for an occupied-mostly better than code house (by more than 2x!) - use Manual-J or model it with BeOpt instead. But the rest of it will give you the idea of what they look like and how they are installed:
https://www.youtube.com/watch?v=kU1Nlm1yMtg
Unlike mini-splits PTHPs do not modulate with load, but like all heat pumps they need enough load to run cycles that are long enough and at a sufficient duty cycle to provide the best efficiency.
Also unlike mini-splits, below some outside temperature (typically ~25F) they use resistance elements to provide the heat, usually turning off the heat pump while in resistance heating mode. In a zone 4A climate the outdoor design temps are often below that threshold, but the average outdoor temperatures over a winter are higher, so it still operates in the higher efficiency heat pump mode more than 90% of the time. When in resistance heating mode it's the same efficiency as the electric water heater/radiant floor solution, but the other 90% of the time it's more than 2x as efficient, and when it's in the mid-40s or higher outdoors it's more than 3x as efficient.
eg: Amana is claiming 3.4x as efficient as a resistance heater for average performance, operating in heat pump mode all the way down to +24F:
http://www.h-mac.com/product-catalogs/amana/Amana-PTAC-Specifications.pdf
LGs are comparable, some slightly higher efficiency, depending on specific model- search the document for "COP" (= coefficient of performance.
https://www.ajmadison.com/ajmadison/itemdocs/MFL67884604_PTACEPDB_013D09_20130509140143.pdf
There are others.
The real determinants of as-used efficiency isn't the tested COP on the bench (at +47F outside, +70F inside), but rather how much time it spends in resistance heating mode, and the duty cycle, which is determined by sizing it correctly. Don't expect the real performance to use 1/3 the total power, but it should still use less than 1/2, when sized reasonably. A 2x oversized unit with a somewhat higher tested COP isn't going to beat a 1.25x oversized unit with a lower bench tested COP. on power use. For energy nerds interested in how these things are tested can refer to this document:
http://www.ahrinet.org/App_Content/ahri/files/standards%20pdfs/ANSI%20standards%20pdfs/ANSI.AHRI.CSA%20Standard%20310_380-2004.pdf
For a house with doored off areas it may be preferable from a comfort point of view (and to meet code for heating systems) to use 2-3 half tonners rather than a single larger unit. If most of the house is an open floor plan a PTHP sized for the whole house, with (appropriately sized, not oversized) baseboards in the doored off areas than need it to meet code can still be very efficient. Using the PTHP to heat the main space to a higher temp than say a doored-off bedroom can still provide a decent fraction of the heat to the remote room, especially if the door is left open to allow for convective heat transfer.
Thanks, Dana, for all that info on PTHP installation! It's good to know they are possible to install without specialized HVAC equipment.
This PTHP information is intriguing - can they work in cold climates as well? Do you think realistically that people could comfortably live without having a vent in each bedroom?
Thanks,
Noah,
As Dana explained, a PTHP "works" in a cold climate -- it just uses a lot of energy to work. For most models of PTHP, the heat pump stops working when the outdoor temperature drops to about 25 degrees F, and the unit switches to a different mode -- it uses electric resistance for heating.
So if you live somewhere where the outdoor temperature is below 25 degrees for months at a time, a PTHP will be inefficient compared to a cold-climate heat pump (like a minisplit from Fujitsu or Mitsubishi).
PTHPs are designed for hotel rooms. They work best when they are installed in the room that needs to be conditioned. That said, a house with a very good thermal envelope might be able to be heated without installing a PTHP in each room -- especially if the occupants don't mind temperature variations from room to room.
Using Dana's figures (or a couple of $400 window heat pumps) with replacement every 5 years and any reasonable interest rate, I can't make additional solar PV and resistance heat work on a $ basis. The cost effectiveness of an expensive heat pump (like a Chiltrix hydronic) does depend on life spans. But it supports thermal storage, distribution to each room and more efficiency (especially at low outdoor temperatures).