We’re indulging in a little smugness. It feels good to be right:) What follows is an explanation of the HVAC system in our new house, and the success that comes from doing some actual engineering design, as opposed to blindly applying ill-fitting and poor standard practice. The TL;DR version is that you really should run Manual J calcs on any building, including residential buildings, to determine the appropriate HVAC specs for that specific site, or you risk ending up with a really bad design that costs more and doesn’t work as well.
Basically, we were right, and everyone else was wrong, and this post is a garrulous “I told you so”. (I can’t imagine this interests very many people, consider yourself warned).
For the HVAC system at our new house, Kevin specified 2 one-ton Fujitsu ducted mini split heat pump units, one for each floor. We were told by several people, including the HVAC subcontractor, that the system would not work as designed for two reasons: 1) it wasn’t enough capacity for the size of the house, and 2) the specified ducts were too large, and therefore the air would move too slowly to ensure proper mixing. Basically, the heat pumps would not be large enough to adequately heat and cool the house, and there would be hotter and colder spots within the house. He told them to install it anyway, and we’ll see who’s right after the system is running.
The HVAC system at the new house has only been completed enough to be turned on for about the last week or so. The timing worked out well, however, as this recent cold snap with temperatures down into the single digits (unusual for our part of NC), has provided an excellent test to see if the heat pumps can keep the house warm.
For comparison’s sake, the heat pump at our current house is more conventional. We bought the smallest unit available to us at the time in 2006, a 1.5 ton heat pump rated at 13 SEER (a measure of efficiency) with back up heat strips. This provides heat and air conditioning to about 850 sq ft of conventionally constructed space. (Yes, I live in a small house. Try it, it saves money). During this cold snap, the air-source heat pump cannot make enough heat to keep the house at set point, so the back up heat strips have been running (which is expensive and inefficient). This is why people who live further north have mostly not chosen heat pumps in the past, as heat pumps were unable to make heat in very cold weather.
Enter the modern mini-split heat pump: The ducted mini-splits specified for our new house are a special cold rated version (slightly more expensive) rated at 21 SEER (so, a lot more efficient) and claim to make heat down to sub 0°F temperatures (which must involve magic somehow). (There is actually a third mini split for the room over the garage. Its a regular (non-ducted) type, and its rated at a remarkable 33 SEER). There are no back up heat strips in any of the units. Kevin specified 2 one-ton units, to heat a bit over 2500 Sq ft. He also specified larger than typical ducts, as we are both easily annoyed by the noise of air moving through the ducts. Smaller ducts costs less, but make increasingly more noise as the duct gets smaller. We don’t want to listen to our heating and cooling system, we just want it to be comfortable in the house. This will work for us because when you have such a well insulated and tight (non-drafty) envelope, proper mixing is not an issue.
Fast forward to today: We don’t often get temps as cold as the single digits, so we were motivated to go over to the house in the morning in sub 10°F weather to see what the temperature was inside the house. (It was between 7° and 10°F when we arrived, cold enough for a good test). To make it even more challenging: since the units were just turned on last week, we discovered only the downstairs unit is working. The HVAC guy hasn’t been back to fix the upstairs one yet.
When we walked in the door, the house was warm. We walked around with the thermal camera (which we have verified accurately reports temperature), and the entire first floor was within 1°F of set point. (To be fair, we only have it set to 60°F since we aren’t living there, and don’t see the point of heating an empty house any warmer than that). The upstairs (which, remember, is not heated right now because that unit isn’t working) was uniformly at 54°F. We pointed the thermal camera at corners, walls, floors, ceilings, and the temperature did not vary more than 2 degrees, better than our current house. (On the first floor, the air coming out of the vents was in the mid 70s).
Bottom Line: The single one-ton unit, half of the installed capacity, is keeping our 2500 sq ft house at 60°F (at set point) in single digit weather. And its not even running at capacity (it was not running wide open all of the time). Pure heat pump, no resistance heat strips. And there were no hot or cold spots in the house.
Boom. Mic drop. He told you so.
The general contractor and HVAC contractor specified 4 tons of capacity. They were wrong. Our system cost $8000 less than their recommendation.
So, why? Why could we specify half the capacity and larger ducts and still have a system that works when the professionals said it wouldn’t?
Well, there is this thing called engineering design. It involves doing calculations (you know, math) that model the system to determine the heating and cooling loads. (We learned this in school!).
Ok, I’ll stop being snarky for a moment. There are several reasons why we’re right and everyone else is wrong. (ok seriously, I’ll stop).
First, professionals do not perform HVAC calculations for residential houses as a rule. (Look up “Manual J load calculations”). HVAC contractors generally don’t even know how to do them (they could, its not rocket science, just tedious work). The calculations are done by the manufacturer’s engineers, who create general guidelines for installation in typical residential applications. They reduce all sorts of variables about house size, wall thicknesses, insulation types, window area, glass types, doors, basements, floors, ceilings, room sizes, etc, based on the “average” home, to two basic numbers: You need “x” many tons of capacity for every “y” number of square feet. As you can imagine, this often leads to sub-optimal results. This is part of why so many residential houses have such poorly designed and implemented HVAC systems; how well the system works in your house depends on how closely your particular house is built to the assumptions made in those calculations.
The problem for us is that our house is not conventionally constructed. We have SIP walls and roof, and better than average windows. The “average” HVAC calculations make all sorts of assumptions that do not apply to our particular house design. By doing the Manual J calcs ourselves, it was obvious that 4 tons was far too much capacity, despite what the guidelines our HVAC contractor was using said to do.
Also, HVAC contractors are incentivized to sell larger systems. They make more money, and no one complains about oversized systems (despite the higher costs). (Well, except for oversize air conditioning systems, which cool spaces too quickly to allow the system to pull the moisture out of the air, creating humidity problems). If they under-size it, people complain.
In general, modern residential houses are not designed to have a particularly good envelope (Roof, walls, windows, doors, floors, foundation), or to work with solar gain, or other site features. The philosophy seems to be: pick a floor plan that may or may not be suited to the site itself, build the house as cheaply as possible, and then overcome its poor design with a large, expensive HVAC system that makes the house more expensive to own.
I think I’ve mentioned before, we intentionally made the trade-off on House 2.0 to spend more on insulation and windows, so we could spend less on the HVAC mechanical system. Not only are the monthly bills lower, but the total cost of ownership is lower as well. All HVAC systems need maintenance, and eventually wear out and break, and need to be replaced. Smaller HVAC systems are cheaper to replace, and SIPS insulation is forever. Its just smarter home design.
As most know, reality does not always conform to engineer’s designs, so we are always pleased when it does. In this case, we were right, which means we will have a home that is more comfortable, efficient, and less expensive over time. Engineering 1, standard Practice 0.