Heating alternative for a Passive House
I’m in the early stages of planning my own passive house (PHI) in climate zone 6 near Minneapolis. Current design is about 1,550 SF, so that sets the budget of 4,960 btu peak heat demand…long road ahead! I’m exploring alternatives to minisplit or baseboards for primary heating with good zoning and distribution.
Here’s the hair-brained idea: radiant heat cable run through base trim and underneath the deep window sill finish, call it stone or tile. Kerf cut some vertical trim boards to cover the loop up to the corners of sills keeping everything inboard of drywall for eventual service or replacement. Cables get about 10.4 btu per lf and come in plenty of lengths. This should make it easy to right-size for the peak load in each space/zone and targets the sill which tends to be the lowest surface temp. Invisible, great window CR, temp equalization and hopefully comfort with lower set points. Obvious con is atypical installation, COP of 1 and higher cost than baseboards. But I’m fairly confident I could design and install this myself. Not much to it.
Thoughts or other cons I’m missing? Would this be better suited as a backup for a minisplit all things considered?
Thanks in advance for responses.
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
What about summer cooling and dehumidification?
I can deal without the cooling. Night flushing works most of the time and fans for the rest. Considering just a standalone dehumidifier and really good shading strategies.
Sounds complicated. I would just use resistance floor heat mat under the windows. There is stuff you can staple up in the joist bay from bellow.
This still doesn't give you any cooling. You can probably use something like a Minotair, but the cost is more than an efficient ERV+mini split. If you go with that, any extra heat is pointless.
The heating cable I'm looking into is essentially the same as what comes in mats, advantage of the free cable being size and placement.
I'm not sure I follow. Floor heat mats can be place anywhere, they can be sized for the the load (come 4' to 16' in length), heat is where you want it (down low) and are pretty simple to install.
You have to be careful with some of the floor heat cable, most are have to be covered with thinset to protect the cable and spread the heat.
You're on the money! Finish sills would be on thinset. The base trim-concealed idea I'll admit is more 'out there' but cables can be installed under wood flooring on sleepers per mfr data. In my mind an air gap behind trim is equivalent but perhaps to Charlie's point it may be a code issue. I need to look into that.
It could work, but it would triple your energy consumption compared to a good heat pump so I'm not sure it makes sense.
The other issue would be code/safety. Code for wiring run in walls has provisions for protecting it from accidental damage occurring when someone hangs or picture or decides they want to change trim styles. You'd need to figure out both how you can satisfy the intent of the code and whatever degree of the letter of the code your inspector wants to enforce.
Radiant cove heaters are a popular option that have their standing with respect to code well established if you want straight electric heat.
As Doug McEvers pointed out, you may want cooling and dehumidification. Peak summer humidity in Minneapolis can be pretty severe.
Chiltrix units are oversized for your need, but they have a good turndown ratio and they can deliver very small amounts of heat to individual rooms.
I'd run the numbers as to whether less insulation and hydronic/heat-pump radiant or ducted mini-split makes sense. Or PV solar (either installed onsite or paid for offsite).
IMO, a high-end home should should maximize comfort - which usually means thermostatically controlled hvac zones.
Keep in mind that you need ventilation - which often means ducts anyway.
Regarding Chilltrix, add a buffer tank and there is no issue with over-sizing. It can do per room zoning with heating and cooling. At about 1000 gallons, it would allow daily time shifting for greater efficiency (operate during the warmest part of the day), less environmental impact and possibly reduced electricity cost. It can also produce domestic hot water, a large portion of energy use at passive house levels. Downsides - will be more expensive and complex.
Jon,
Good thoughts. Definitely doing on site PV at some point. With the radiant heat cable I do plan to have thermostats for each zone. Ducted minisplit is my next-best option, but not a lot of extra volume in the design for ducts...other than low volume ERV ductwork.
There is something odd about going to all the effort to build to PH standards and then using a heater with a COP of 1.
Malcolm,
Fair point, and I'm toiling with that myself. I'm learning that electric resistance backup of some sort is a good idea being that even the best minisplits become useless at -17. Not many comparable projects here but I think all have backup heat. So I'm buying two systems. Crunching the numbers on just this one makes it more appealing.
There are a number of mini splits out there that deliver 80% capacity at -15F with above 1COP. Since you'll be running most of the time at warmer temperature, on annual basis you'll still be WELL ahead of resistance only option.
Backup heat for a place that insulated is not needed. You can easily heat the place with a couple of plug in heaters.
Sounds like I need to do more research! Is this Mitsubishi or Fujitsu?
Pretty much all hyper heats, some are better than others. Will take some digging for the full engineering specs that show what they are. LGred, Midea Premier, Mitsubishi MUZ-FH, Fujitsu ASU RLS3Y.
Much appreciated!
Comfort aside, the only reason to insulate to PH levels is to reduce energy consumption on heating and cooling. If the source of that energy can be three times as efficiently derived (and okay, this is a bit of a simplification), you could have achieved the same results with a third of the insulation.
I think a good compromise between built-in resistance heaters and portable ones are these: https://www.convectair.ca/en/products/120v-plugin/apero They can be demounted if not needed, or left in place permanently.
Malcolm,
Due respect, but that's not a simplification, that's bad math. Decreasing insulation increases assembly U factors on an exponential curve toward greater heat loss. Again, I do follow your point on the COP though and ultimately I may arrive at the ducted mini for that reason if it's a fit for the design.
Jason,
Sorry, I'm not following you. If you double the R-value of a wall you don't half the heat loss?
Crap. The bad math is mine. 1/3, 3x, but oversimplified as you said.
I still always give passive insulation precedence to active mechanical heating or improved COP. Active systems require maintenance and have a much shorter lifespan, and improved heat pump COP falls when temp deltas increase. They do that plenty here.
I agree entirely. It leads to a much more robust house.
Deleted
Your statement references "heating with good zoning and distribution". I wonder if you realize that in a super tight house, you actually don't have to be quite as concerned about zoning and distribution, since there is very little heat loss in any specific areas. And distribution is typically taken care of by the normal flow of air. This isn't to say that you can install one point source of heat when the house is divided into many small distinct, closed off areas and have the exact same temperature everywhere, but in practice it seems to work out pretty close to that. We've built a number of 1500 SF houses with two or three non ducted minisplits, resulting in one or more of those being left off and the all the rooms being equally comfortable.
On the other hand, if you can figure out a way to install radiant cable legally and safely and can still meet the (potentially) higher loads of the lower COP with solar, for a good deal less cost, why not? It would be a good experiment provided you share your results. But with warmer, more humid summers becoming the new norm, we greatly appreciate the comfort effects of the "DRY" cycle on our Mitsubishi wall units!
Bob,
When you say "distribution is typically taken care of by the normal flow of air" I assume you mean by the ERV. I doubt that will actually have sufficient airflow to equalize temps at reasonable ventilation rates. I also suspect lower temps in some areas will lead to the family fiddling with thermostats.. which increases the load.
I live in zone 6 Maine in a pretty good house. Sole heat is a pair of minsplits. You really don't need a back up system. Buy a couple of cheap electric heaters if you're worried. The only time we could use a backup is when the power goes out, which is often. But of course without power, most heaters don't work. If your load is that low ( which is astonishing), you could go for days without heat and be fine.
After a few years, we broke down and got a generator that runs almost everything when the power is out.
Solar hot water for hot water and heat, using baseboards, old cast iron radiators or radiant slab???
Why not try solar hot water? We're in an experimental period; try everything & it will sort itself out! (assuming cost isn't the primary driver)
Based on calculations, closed door rooms won't be warm enough without independent heat. Open door rooms may be OK (3F max drop) with passive-house loads in Minnesota.
Jason,
Most of the points you need advice on have already been commented on. I will suggest that the wandering wires to provide heat are, as you note yourself, hare brained. Aside from the likely rejection by code officials, you don't really need to be putting heat at window locations. This thinking is a holdover from another era.
If you are really looking to go all PHIUS or near to it, then the windows you should be looking at will have U values in the .15 range. Mine are .15 fixed, .19 operable and my design temp is minus 10F. Walls are R32 whole wall, roofs R55. I have taken measurements of floor and wall temps at the windows edge returns and directly below the windows. I found less than 1 degree difference. The "cold waterfall" effect even under the 70" high picture windows is not perceptible. I admit I don't spend a lot of time on my window sills as the cats regard my intrusions unfavorably. Suffice to say it is not a rational metric to design a heating system around.
The mini-split proponents are quite right about efficiency and comfort in humid summer locations. I would suggest however, that if you go this path, you will need to closely analyze the head choices and dial down capacity of the unit choices due to your very low loads. It may prove very tricky to pull enough moisture from the air without over cooling. Malcolm is quite right to note that spending on efficient mini-splits in a slightly less over insulated home will provide better comfort and bang for your buck.
An independent dehumidifier may seem like a good choice, but you must account for the heat that you put back into the house. Maybe someone has produced a dehumidifier that puts the heat outside, but I think those are called window air conditioners. Maybe you can adapt a Penguin type room conditioner? Would it be worth the effort?
The room to room variations of temperature we have experienced in my house are greatest during the summer, as we do not have AC. We do "flush" the house every evening courtesy of a 35-40 degree swing in our daily temperatures. The 9% relative humidity means we can put up with the 10-12 degrees of daily upward temp gain comfortably until the nightly temp shift.
Our local dry environment is quite different than your own, so I must draw on past experiences with high humidity locales lived in and visits to MN. Flushing the house with cooler night air might be feasible depending on you own needs, but do ask yourself if the others in your home really share the same enthusiasm for suffering. It was apparently north of 90F at 10PM in Minneapolis a few weeks back when I got called by a family member about how to diagnose a failed AC unit. They were not very happy with the night air.
Maybe they exaggerated, but my prior years in Illinois has provided real time knowledge of how muggy the house felt the whole day after using the whole house fan. In trying to avoid the costly to run window AC units we were just miserable. Spousal sense won out and the AC went on for the long muggy summers. I will also note that the closets in Illinois were always a bit dicey in the humidity even with AC since they were essentially closed off.
Mr. Irving comments about room layout are very relevant to any design choices you make. My own home has been referred to as a rabbit warren, but the comment came from a fan of open concept which neither my wife or I like. Still, the ability of heat to move from where the wood stove is, is very hampered by our layout. Ducting that I planned for helps, but by and large we would be staying in the living room during an extended power outage. Try to read more about how Mr. Holladay feels about magic air arrows.
Based on several now invalid cost projections, we went with cove heaters in all the rooms. All with their own thermostat. It works, it is not a great COP, and we don't get to have AC. For our environment and desires (no more fan noise) it is very comfortable and resalable. A mini split may be in our ever warming future, but for now we are good. Do consider the future and how salable the house might be.
You have a lot to look at and decide on. I would suggest you read widely and deeply here at GBA. Listen to the many voices. A highly efficient home is the goal, how to get there without being too far off the map will keep you and your family happier. Plus, have you considered how the mortgage people will react to your proposing new and untested methods? It may be a good thing to check that out before investing huge amounts of time in a home no one will underwrite.
> you don't really need to be putting heat at window locations.
Using this, it's easy to find some window heights/distances that still create a downdraft comfort problem and marginal radiant comfort. Room side Low-E creates a problem even with shorter windows.
Some continuous heat below windows can be useful.
Jon,
Good resource to have! Thanks for that.
Comfort is a very squishy topic. My recomendation is to aim towards more comfort whenever making any design decisions. Nobody ever complains about that.
For example, my own well insulated home with good (not passive house) triple pane windows, if I turn off the floor heat, the area by my largish windows get uncomfortable on those cold clear winter nights. A bit of heat there is simple to add, doesn't cost much to run and improves comfort even if it is not used 75% of the time.
Similar approach to air conditioning bedrooms. In a well insulated house, even on cooler nights your losses are very small, the heat gain from two people sleeping with doors closed can easily increase the room temperature 4F to 6F, which is the opposite of what you want. So for comfort, always have an active means of cooling bedrooms. This is why I always recommend a ducted mini split.
These are small design changes that are easy to do up front but expensive to fix down the road.
That's a good example of low load increasing (vs the often claimed reducing) room-to-room differentials.
But how can a ducted mini split accommodate a significant variation in ratio between bedroom and other room heat loads? Smart vents? And even that won't cool the bedroom when other rooms need heating.
A ducted unit will provide additional ACH in the a closed bedroom. It will be a bit warmer than the living space, but nowhere near as warm as no cooling.
You can also crank up the cooling in the living space at night time to get the bedrooms colder. Since it would be off peak power and colder outside temps, there is very little energy or financial cost to this.
Using the windows I have, I could not find a combination that failed to meet the default comfort criteria on the calculator, using an outdoor temp of -25F. I found changing the height of the window and sill had a pretty modest effect. But they do let you change the comfort criteria, so I guess in that sense it's pretty easy to just tweak those to create the problem. I couldn't see any information on how the defaults were chosen, or any guidance on changing them.
Actually there is a split-system dehumidifier! https://www.santa-fe-products.com/product/ultra-aire-sd12/
Interesting, although it would much less expensive to install a Daikin Quaternity.
The split Snata-fe does not solve the low load humidity issues. It still provides sensible cooling although a much better SHR.
You can get pretty close to that with a standard mini split by reducing the air flow. I've been doing this for a while now and it lets me get the house down to 40%rh while keeping it around 78f. The warmer low humidity air feels much more comfortable.
Jason: referencing your comments in #23 - "lower temps in some areas " - seems to indicate that you suspect some wall/roof areas will have greater heat loss than other areas; but this is exactly what a Passive House is designed to avoid. In a Passive House, and to a very slightly lesser extent in the "Pretty Good House" standard that Stephen Sheehy used, and what I also use, the heat loss is similar in all areas and surfaces of the building, so there are not areas that have greater heat loss than other areas. This concept is understandably hard to wrap your head around, but it is exactly why the methods we champion work.
I would strongly recommend that you try to find homeowners in your area that have these houses and talk to them; I think it would help you to understand how the concept works in real life.
Both Stephen and I live in climate zone 6, and there are dozens of builders in Vermont, Maine and NH building similar homes, using non-ducted minisplits with no issues. It may very well be colder in Minneapolis, but has been said here, that might require a small backup electric resistance heater or a wood stove, but in general a simple non-ducted minisplit should be ideal for your house year round. When we are in construction mode, and the house is at last partially insulated, one 1500 watt resistance heater will usually bring a house up to 80 in the middle of winter; all the subs are routinely amazed, but that's what the house is designed to do.
Bob,
With regard for my comment, I wasn't concerned with envelope transmission losses. I'm concerned with the redistribution and tempering of spaces of differential solar gains, stack effect, and internal gains. Spaces may be separated by sound-insulated partitions, floors and weatherstripped closed doors. Not a lot of heat will move through jumper/transfer ducts even with the ERV moving 100 cfm. For that reason and others I've mentioned, I'm leaning away from point source heating though I'm aware it can work.
I believe Tim Eian planned to use just a ductless heat pump on his home here. It was still under construction when I toured it. Another project I saw under construction in Afton ultimately went with a zoned hydronic system (imported, uber expensive but very good COP). Meanwhile Katrin Klingenberg used electric baseboards on her own passive house, the first in North America.
The solution depends on the design, the project goals and budget.
You're in a similar position as we were before building our passive home in zone 6. The shell package builder convinced us that because our heating demand would be so low that in-floor heat from pure electric source made the most sense. This turned out not to be the case. We installed a heat pump on the main floor after the first year, and this provides all the heat for the main floor (open concept, ~1500 ft^2), and 80-90% of the second floor heat (three bedrooms and a bathroom, ~1000 ft^2). The small shortfall on the second floor is made up by a 600W cove heater in each bedroom on separate thermostats. They rarely run, and never at more than 60% capacity. We also have 3 cove heaters on the main floor, which only get used as emergency back-up. The Fujitsu heat pump not only has always managed to keep up with demand, it's never been pushed near its output limit. It's saved us a lot on heating costs, despite passive-level insulation, windows and air sealing.
In this case, the choice of water heater was as important as space heating:
http://flatrockpassivehouse.blogspot.com/2017/01/heating-water-in-passive-house.html
Thanks to Cody for spotting the Santa Fe dehumidifier for those people stuck in tricky load vs. humidity situations. I would imagine that it would be very useful in helping out many with forced air furnace installations in PGH builds, where the AC function is hampered by the low loads for cooling, but still needing enough blower time to draw down humidity.
And Trevor, I came to a different conclusion trying to use the buttons on the Payette site which has been recommended before. Using the new page example of .35U windows 7' tall in a 12' high wall rated R15 at 10F with 3' window sills, I can see why Predicted Percentage Dissatisfied exceeded 20% even 3' away. Those are really big A cold windows.
What did surprise me was whatever metrics they extrapolate from would still predict 15% to be dissatisfied sitting 3' from my .15U 5' x 4' windows with 1.5' sill heights. I do think that more than 5% additional subjects would be happy sitting in front of my windows.
While I am only a balding sample of one, I find that having breakfast or dinner in front of three such windows in the corner dining nook at less than 2' distance is quite comfortable down to 10F and well below. Holding the same outdated belief that heat needs to be placed under windows, I had the electrician install two 240V tap points for baseboard heaters. My reasoning was based on having only one 450W cove heater facing the dining area mounted on the wall 9' up and 10' away from the window walls. I have not bothered with baseboard heaters.
For this reason, based on real experience, I suggested in an earlier comment that a better than PGH home would likely have windows in the .15U range making the idea of heating window areas unnecessary. I would add that I have at no time experienced cold drafts off of even the much larger windows in the living room. I also have never seen significant temperature variation below the windows or along the perimeter walls. Maybe my rim detailing is the secret, I am not sure.
The Payette company is focused on advising or designing large institutional buildings that (based on promotional pictures) are full glass facades and meant to be occupied by largely dense human gatherings. Not very related to parameters found in most residential homes. I did dig into a few papers presented in support of their methodology, which only furthered my perception that the comfort suggestions they make are more relevant to large public buildings.
For perspective, the opening splash on the site touts how they plan to reduce total building energy usage from 370 to 108 E.U by 2030. An EU seems to be 1000BTU per s.f. per year. This divides down to a drop from 42BTU per s.f to 11.4BTU per s.f PER HOUR! You can see why they are dealing with another whole level of building design. Also the tremendous energy waste in public buildings.
For comparison, my 100% electric home uses 2BTU per s.f. per hour per year. The water heater is a big chunk of that followed by the heating load which is about 1.3 BTU per s.f. per degree day. Of course I don't have to deal with hundreds of people coming and going, requiring commercial lighting levels, etc. That is also why I think trying to use the Payette comfort guide is not terribly useful.
> I think trying to use the Payette comfort guide is not terribly useful
Interesting that you conclude this when it's results are completely consistent with the comfort experience you report above. Not that a n of 1 should be used to conclude anything in terms of general comfort advice. I don't find the ad hominem arguments persuasive.
Do note the large effect of room air temp on PPD - air temp boost is an alternative to localized heating (although they don't seem to account for discomfort from being too hot).
to Jason and further discussion,
Akos, in your comment #39 you suggest the Santa Fe unit would not solve low load humidity issues, but would provide a better SHR. I am very weak on my AC understanding and had to spend time getting some definitions sorted out.
It seems that a "better" SHR (sensible heat ratio) might mean a low value like 0.3 which biases water removal over sensible heat (temperature) reduction. Without knowing what SHR means it is somewhat unclear as to why you feel the Sante Fe unit can be successfully imitated by changes in the run profile of a minisplit head.
Sante Fe's UA-SD12 literature does not mention an SHR value, instead they use pints per Kwh and some other info. The sensible rating for one unit is 4300 BTU. The biggest surprise to me was realizing the unit is rather like an in-duct mini-split head. Perhaps they have tweaked the coil surface area and temperatures to extract moisture at maximum rate.
In larger A coil set ups in forced air systems, I have read materials that suggest potential humidity problems can arise due to capacity oversizing or mis-adjusted blower speeds. The problem seems to be SHR values falling closer to 1. In my limited understanding, this means too little loiter time for the air passing over the coil results in less water being extracted. There seem to be variant ways to get poor results in AC systems relative to humidity control.
From other threads I have followed over the last couple of years, I have picked up the notion that due to smaller air cfm and smaller coil sizes, that it can be tricky to balance noise from the fan against desired output. This seems to come up when single head or two head installations encounter problems due to room isolation or humidity overloads.
In a low sensible load home ( at near PHIUS insulation level), with tight ACH numbers, it would be beneficial to many to understand how to make the standard minisplit head able to manage the humidity build up from cooking, breathing, plants, etc.
The original start of this thread was about satisfying a very small heat demand, but soon went into the issue of cooling demand and how best to meet that. Having now seen the Sante Fe unit and understanding it is a split system, would you not agree that running something like this unit in the summer months would simultaneously satisfy both sensible heat load and humidity reduction effectively for Jason's very low load house?
40% RH at 78F seems on the high side of comfortable parameters, though I may have a distorted perception now that I have left the midwest to an area where it is 12%RH. I think it would be instructive for followers on this thread to expand on your comment and help define the underlying concepts you are using to make your minisplit head perform well in what I am guessing is a more humid environment than mine.
I am truly not saying this to be snarky. I think many on GBA express frustration with the settings on their minisplit systems and often resort to hacks to defeat the programming built into the control units. I often wonder if the control units programming is suffering from a source bias; the parameters might represent average Japanese weather and preferences.
An educational case to analyze is when it's 75F and 80% humidity outside.
We can calculate some of the items missing for the SD12.
The unit is rated for 184pints, so 7.6pints/hour. Condensing a pint of water requires 1066BTU of cooling, so the unit provides 8100btu of latent removal. With 4300btu sensible rating, its SHR is 4300/(4300-8100) is 0.34 which is excellent, you probably will never get a mini split near that.
The way I've gotten mine to better SHR is by reducing the air flow. Since I have wall mounts, I just blocked the intake with a couple of books till the air flow dropped to about 1/2 the unit's rated value. This was mostly for comfort (very bad idea to put a wall mount on a multi split in a bedroom) but did have the unintended consequence of improving the SHR of the unit. I haven't measure it, but the outlet temperature is significantly colder than before, so it is definitely removing more water. Don't know how this will effect the life of the unit, so far it has been running like this for about 3 years now.
With a ducted unit this is generally easier as you can program in the external static pressure into the unit. By playing around with that setting, you can reduce the air flow to get better moisture removal. All mini splits have a coil temperature sensor and won't frost up if you reduce the airflow a bit. Don't know how it will effect the life of the unit, the setup is pretty close to a clogged coil or filter so the units are design to handle it.
Even with these mods, I doubt you can get a well sealed house humidity under control, you would still need a bit of extra dehumidification or extra sensible load. For example, in my case, I have some large west facing windows and the solar gain adds a fair bit of sensible load that increases the runtime of my mini split significantly in the afternoons. Without that, I would have a hard time maintaining a low RH.
If you do need dehumidification, I'm not sure the SD12 is the best fit. It is very expensive and it can still over cool a low load place. I think it is great fit for a larger house in warmer climates where it can replace an ERV+dehumidifer.
A stand alone dehumidifer is much cheaper, controls only RH and the extra bit of sensible load from it can be rejected by an efficient mini split for very little extra cost. Unlike the SD12, the mini split can also heat your house with very good COP, which is where most of the operating cost is in northern climates.
As for my 78F 40%rh setpoint. Along with a 75F wintertime setting, that is a compromise one must make in marriage. Even when walking in from outside with similar temperature but much higher RH, the house feels fresh and cool.
If you want a code approved hidden resistance heat consider electric ceiling radiant heat. The heating wires are applied to the ceiling and hidden in a light texture.
Walta
> you can program in the external static pressure into the unit
But note that during the afternoon (high load), the SHR is fine (so better to use normal static pressure) and at night (low load), you need lower static pressure to reduce flow. So lots of playing around. A fixed low setting will have a negative effect on efficiency and capacity.
> A stand alone dehumidifer...
Agreed, this might be a better choice.
Check the chiltrix and Artic heat pump. They are a good fit for our climate zone...I am in St. Paul. They can do DHW, heating, cooling, and dehumidification. Warmboard is a great product I have it my house.
I want to thank everyone again for the great insights and discussion. Based on feedback, ultimately I think I will land on using the electric radiant as a backup system, installed only underneath the window sill finish. I was reluctant to go with the ducted minisplit for reasons of sound and simplicity, but I don't want to run the risk of finding the radiant system inadequate. I'll find a home for the ducts...begrudgingly. Exactly which ducted system to choose is a post for another day.
Best regards to all,
j
Jason,
At the risk of sounding like a zealot or a shill for my heating choice, may I once again politely suggest that you are not going to be well served by worrying over chilly window sills. If you are going to select low U value triple pane units then the chill factor from the window sills is not going to affect the room significantly. Electing to create heated window sills will likely be expensive and maybe entail special detailing to satisfy your local inspector or electrician.
I live in a similarly cold climate (I have experienced Minneapolis in the winter) and heat my entire home with radiant cove heaters. The electricians I used for my build lead me to the choice and they assured me that they installed many as booster units in the high mountain country. Try looking up Silverton, CO winter temps for reference. With a few caveats that I have explained on other posts, they have proven to be an excellent choice for my circumstances.
I will freely confess that I would have greater energy efficiency overall if I had gone with mini-splits (and also have gained future AC capacity), but for my location and build year the mini option was crazy expensive and I didn't need AC. My pending next build will almost certainly have main heat from mini-splits and backup with radiant cove heaters.
The radiant cove heaters are very inexpensive to purchase and simple to wire in. Even though they run at a COP of 1, you will probably find they run very little time overall. I suspect that you will find the mini keeping up most of the year. When it is extremely cold, your mini will labor to make heat at better than 1+ COP and if a resistance feature is included, then the COP is the same as the cove or less since the fan will be a negative. If you have distribution issues with the minis or ducting, the cove heaters can be a nice way to balance a room if the occupant wants more heat than can be practically directed with the mini system.
Additionally, they are pretty impervious to failure as the heating element is not running all that hot. The cove heaters can be felt if close to them, but they are not at all like a common bath heater with fan. I carry a 12x24 kitchen/breakfast corner with 50% of the wall area being glass on two heaters (450W and 600W) set 9' up on the wall facing the windows. Maybe the stone window sills do get cooler than the room, but stone always feels colder than a room due to conductivity of the material. The heaters do not run that much according to the display.
I happily have meals in the nook with three large windows behind and beside me whether the temp is 85 or -15. The windows are Alpen's with a U of 0.15 and 25%SHGC which certainly helps with the comfort levels. If you choose windows with similar U values you will be similarly happy. Only plants and cats sit on the window sills. The cats don't mind the sills as long as there are birds outside to stalk. The plants don't share with me.
Look up Radiant Systems Inc as a start. I used their products and have been happy. There are others that make very similar looking items. For placement purposes think of cove heaters as somewhat directional light bars of different brightness. Check your individual room loads for how much your worst hourly loss rate is and select for that. Our NE corner bedroom of 200+ sq. ft. if carried on two units which do not cycle on very much most of the year. If it gets down to -10 you can hear them "tink" a little bit when heating. For the most part they are dead silent, which is the main reason my wife liked the whole idea to start with.
I think that embedding heating elements into a window sill will result in a lot of warming where least needed. The wall will warm up as much as the sill. The cove heaters will warm objects (not air) in the room which ultimately produces a more even feeling. I will probably irritate Walta by saying I lived with ceiling radiant panels in the 60's and hated them. Lots more to that story, but I have as usual gone on too long already.
Roger,
The cove heaters seem comparable in my mind to the radiant sills in most regards, but I still like the idea of targeting the coldest surfaces in the room, which will be the sills even with excellent windows with very low installed U factor and any finish material of choice.
I have no doubt the cove heaters are easy to install not at all unlike baseboards, but they're not as elegant a solution with regard for visual appeal or matching the system size to the peak load. The visual appeal is just my opinion. I'd be surprised if a code official takes issue with radiant heat cable under stone sills but I've won harder battles with higher stakes.
Jason,
Not to persist to annoyance, but while you might convince your code person and even the electrician to allow heated window sills, it might be a good idea to consider how your insurance provider might respond to a failure of your proposed heating idea. There will be no manufacturer's certification like a UL or whatever to fall back on. Would they seek to avoid coverage if the installation is not "normal"? I realize that heated tile floors are virtually identical materials, so I am just speculating, though insurance companies are not notoriously honorable.
The visual aspect of a cove heater is a personal choice. My wife accepted them over the wall cassettes for mini-splits. The lack of fan noise and vastly lower installed cost clinched it for her. As I have noted before, my wall temperature range is barely a degree from interior to exterior surfaces.
I do get a few degree difference as I measure into the deep window surrounds. (The walls are over 12" total thickness) I expected the lower temps, as the window, good as it is, is still about R6.5 in an R32 wall. The window perimeter on all four sides shows the same cooling effect, which is why I would suggest that chasing the sills alone is not going to generate a productive return on the complicated effort proposed. That said, your house - your choice.
And really the cats don't seem put off in the winter.