Solar options for infloor heating in Colorado
I am building a new 2,200 sq.ft. home near Durango, CO and already plan on incorporating solar electric with net metering and no storage. I also plan to install an infloor heating system that would utilize solar energy. My question is:
Is it better (i.e. cost effective and efficient) to utilize a solar assist air source heat pump (e.g., Chilltrix), or a solar thermal system to provide the heat for my infloor system? I will have propane as a backup energy source. If the answer is not straightforward, how do I make the decision?
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
You can't really optimize the mechanical systems until you at least calculated the loads. In higher-R houses it can be cheaper to embed wires in the floor as long skinny resistors and run DC power (no inverter, just an MPPT controller to max out the available power). PV panels are cheap & reliable compared to hydronic heat pump or hydronic solar solutions.
This guy did the math, designed some controls, and has been working on fully commercializing his solutions on kickstarter, etc:
Solar thermal has largely gone out of fashion. It was always problematic in that solar energy tends not to be available when you need it, and produces a lot of unusable heat when you don't need it. A grid-tied solar electric system with net metering means you can return energy to the grid when you don't need it and take energy when you do. Historically solar thermal was more cost-effictive, but with the advent of high-COP heat pumps and rapidly declining prices for solar panels that's no longer the case.
If your constraint is limited roof area and you just want the system that pulls the most energy off your roof, they are close. Evacuated tube solar thermal systems will capture 25-30% of the sun's energy. PV panels will capture about 15%, but heat pumps will have coefficients of performance in the 2-4 range, which turns that 15% into 30-60%.
Where you really see the limitations of thermal is when you start thinking about sizing. You'll two things, the historical insolation by month for your area, and the historical heating degree-days. What you'll see is that they pretty much work opposite of each other, so you have to make a choice: do you size the system small so that it provides all your heat in October and April but only a small fraction in January and February, or do you size it big so that it serves your needs in January and February but is way oversized the rest of the year?
Solar electric with grid-tie doesn't have this problem. You also don't need a backup heat source.
Embedding hydronic heating tubes in your concrete floor is not the state of the art that it once was. Solar thermal heating definitely doesn't pencil out any more.
In my experience, embedding electric resistance heaters in either the ceiling or the floors is a bad idea primarily because of the difficulty of service and replacement. I actually own homes with failed resistance heaters in both locations. The easy and cheap solution is effective but not a real repair:
You might say resistance heaters don't fail because there are no moving parts. I would reply the wiring connections have a high failure rate because they actually move a lot due to thermal expansion and contraction.
You would have to value a warm floor quite a bit to risk being an early adopter of a Chilltrix system. Air conditioning with cold floors in Durango CO is kind of silly when you could simply open a window.
As a landlord I love concrete floors for their durability, but as a spec homebuilder I have found they hurt the home's resale value.
Here's one of many fun discussions on the topic:
Consider mini-splits and if you really want warmer floors, put resistance wires in some small areas. Maybe not CO, but in some areas, net metering is at risk. If so, thermal storage (like Chiltrix plus a large water tank) would gain some advantage.
In some cases, a mini-split over-heating a crawlspace can produce low cost radiant floors.
+1 on getting an approximate load (peak and annual) before trying to decide what is most cost effective.