More Building Matters
Over the last four years, I’ve written about many aspects of heat pump design and installation. One theme that emerges over and over is the importance of accurate equipment sizing. Correctly sized equipment is matched to the needs of the house; it delivers enough heating and cooling to keep occupants comfortable in extreme weather but is small enough to perform well in milder conditions.
Two-part process
Sizing equipment correctly requires two steps. The first—spelled out in ACCA (Air Conditioning Contractors of America) Manual J—involves determining how much heating and cooling the home needs. These design heating and cooling loads take into account local weather conditions, building dimensions and construction, air leakage, ventilation, distribution losses, shading, and internal gains.
The second step, described in ACCA Manual S, involves comparing design loads to equipment capacities to select appropriate equipment. Because capacity depends on outdoor temperature, we need to look at extended performance data to know how much heating and cooling a heat pump can produce at local design temperatures. Manual S has recently undergone revisions that reflect the modulating abilities of modern variable-capacity heat pumps and the increasing use of heat pumps in Northern climate zones; an online version is available here.
Variable-capacity heat pumps can ramp up or down to match the house’s heating and cooling needs. These needs vary with temperature: less in mild weather, more in extreme hot or cold. When they fall between the heat pump’s maximum and minimum outputs, the system can run continuously rather than cycling on and off. By selecting equipment that spends most of its operating hours in this “Goldilocks range,” we can minimize both cycling and the need for supplemental heat.
The ACCA manuals warn against oversizing. So do heat pump manufacturers, state energy programs, and efficiency advocates…
Weekly Newsletter
Get building science and energy efficiency advice, plus special offers, in your inbox.
This article is only available to GBA Prime Members
Sign up for a free trial and get instant access to this article as well as GBA’s complete library of premium articles and construction details.
Start Free TrialAlready a member? Log in
14 Comments
Thanks for this piece, Jon. It's helpful.
One factor that you didn't touch on is defrost. I live in Juneau, Alaska. Our temperatures are quite mild by Alaska standards - winter temperatures rarely fall below zero-F and often are in the teens and twenties. But we're in a very wet climate - we're in a temperate rain forest - and humidity levels are frequently close to 100%. What this means is that our heat pumps have to defrost frequently, sometimes every 30 minutes.
During defrost the outdoor unit reverses refrigerant flow direction, extracting heat from inside the house to warm the exterior heat exchanger enough to melt the accumulated ice (essentially it goes into air conditioning mode), then switches back into heating mode. The process can take 5-10 minutes.
This cycling is effectively the same as an oversized system, with frequent stop/starts, and a corresponding fluctuation in indoor temperatures. This has to result in a hit on efficiency as well as effective heating capacity, but I haven't been able to find solid testing data quantifying that (can you point me towards any?).
A larger unit provides two possible benefits regarding defrost:
1. It can more quickly bring the home back up to temperature after the defrost cycle.
2. It may delay the onset of defrost because its physically larger heat exchanger area will take longer to frost up. I don't know if that's indeed the case because I'm not sure what methods the manufacturers use to determine when to initiate defrost. Is it a simple humidistat and timer system? Fan motor loading? Refrigerant temperature differential? I'm not smart enough to identify it, and I'm guessing it's proprietary with each manufacturer.
I'd be very interested in your thoughts on this.
Thanks!
Bob D
I believe that a larger area heat exchanger can also extract more heat from the air per gram of frost accumulated. That's certainly true when the air is below 100% humidity--I'd have to study a psychrometric chart carefully to see if it's still true at 100% humidity.
Hi Bob, I don't have a good answer about how sizing impacts defrost. I could see it being an important factor in places like Southeast Alaska.
Older heat pumps typically employ simple defrost algorithms. A common one is based on accumulated run time at below-freezing temperatures. After a fixed run time (which can be set at 30, 60, or 90 minutes--longer for drier climates), the system goes into defrost and continues defrosting until the outdoor coil rises to a target temperature (usually in the 60F range) or 10 minutes, whichever comes first.
Cold climate heat pumps employ more sophisticated algorithms. For example, Mitsubishi uses a variable defrost interval based on how long the heat pump took to defrost on its previous cycle. It's described in detail in this application note: https://www.mylinkdrive.com/viewPdf?srcUrl=http://s3.amazonaws.com/enter.mehvac.com/DAMRoot/Original/10006\Application_Note_1041_ME_-_Defrost_Cycles_in_MP_Systems_-_120121.pdf
I also know that Mitsubishi's design tool (Diamond System Builder) applies a derate factor to account for capacity loss due to defrost. I haven't explored it in detail.
Is there a simpler inexpensive and more user-friendly software for calculating heating & cooling loads besides learning the engineer-focused Manual-J system? As an architect and energy/HVAC nerd I would including load sizing in my services if I could do it myself without having to coordinate with a 3rd party. Less accurate is better then no sizing at all or even leaving it up to installers. I have used CoolCalc with some success but I run into problems when the home is not just one empty box (one floor without interior partitions) and I imagine there must be a software that falls somewhere in between cool calc and manual J in terms of cost to use, simplicity and accuracy.
I’m not sure what you mean by this, coolcalc does room by room manual j calcs.
I really like Right-Suite for load calculations. The drag-and-draw graphic interface makes it easy to model buildings with more complicated geometry. It's not quite as intuitive as CoolCalc at first, but I've found that with practice it ends up being quicker to use, especially for room-by-room loads. Every year when it comes time to renew my license, I bristle a bit at the $500 cost, but when it's spread out over a year's worth of billable projects it ends up being quite reasonable.
Being a hydronics oriented engineer, I ask why not add a buffer tank for the cool water and size it for 60 minutes of run time on a design day? Then the heat pump can run (or Defrost) while moving BTU to the fan coil? I found Loop CAD to be very easy to use for room by room calculations. But it does not do any Manual D calculations for duct design.
That's what air-to-water heat pumps do.
One of the fundamental advantages of air-to-water is sizing isn't really an issue. You can put an air handler in each room. You can also get air handlers as small as 3000 BTU/hr, which are quite well suited to small rooms.
I don’t see why, at least in theory, an oversized variable speed heat pump could not have its capacity capped in cooling mode. The author seems to assume the heat pump software is dumb.
There's a minimum speed the compressor can spin at, electric motors don't produce much torque at low RPM. So there's a minimum output when the compressor is running. If the demand for cooling is lower than that minimum then the compressor has to cycle on and off.
When the compressor cycles, dehumidification suffers. The way an air handler dehumidifies is that warm air flows over a cold coil and dew forms on the coil. When enough has accumulated it drips off and is disposed of. At the beginning of each cycle it takes a few minutes for enough dew to accumulate to start dripping, and at the end of each cycle any dew still on the coil evaporates and is returned to the interior air. If cycles are short little or even no humidity removal can happen.
I own a 3 ton variable speed heat pump. In heating mode at low load it ramps down to 50% or less of its maximum capacity.
Are you saying that in cooling mode it has to run at 100%? It seems to me that ramping it down to match the load would avoid short cycling and humidity problems. What am I missing?
Variable-speed heat pumps will modulate in both heating and cooling modes. The modulation helps, but only so much. If your system ramps down to ~1.5 tons at low speed, it will short cycle more than a smaller system that can ramp down further. Allison Bailes has a good discussion of this issue here: https://www.greenbuildingadvisor.com/article/can-you-oversize-a-mini-split-heat-pump
Thank you. Certainly it makes sense to me that heat pumps need to be sized based on a load calculation. Variable speediness won’t make up for gross oversizing. I get that.
What if your cooling load is 25% of maximum? Or 10%?
Customarily HVAC is sized so that at the 99th percentile temperature the cooling load is 100% of capacity. Think about that, and think of the shape of a normal distribution. The vast majority of the time loads are going to be a small fraction of capacity.
Log in or become a member to post a comment.
Sign up Log in