Hacker News Reader

Comment by Nition | Hacker News Reader

This is a wonderful thread, thanks for sharing it here.

> "The kicker is, that the smallest I have been able to find is four ton (48,000 BTU/hr). So this means that in order to proceed with the project, I'll have to build my own heat pump. In HVAC, bigger is not better... just slightly smaller than big enough is best, economically speaking."

Is he saying here that a heat pump that's too large for a space is less efficient at heating, or is he just saying that a bigger one costs more to buy?

If it's the former, can anyone here tell me why that is? It's hard to Google since it gets mixed up with other info like "heat pumps are less efficient at very low temperatures" etc.

I would have thought they'd end up the same. The only reason I can think of is that a bigger one will maybe turn on and off more often to maintain temperature than a small one that's on all the time struggling to keep up.

Edit: Thanks all for the excellent responses.

View on HN

The term of art is "short cycle". There's also the related "turndown ratio" which is just minimum_output/maximum_output. The discussion at [1] isn't bad, but [2] has a little diagram showing the cycling (in this case as a benefit of variable-rate inverters, but also as a warning).

[1] https://www.greenbuildingadvisor.com/question/what-is-minisp...

[2] https://carbonswitch.com/heat-pump-sizing-guide/

I agree it’s hard to find reliable information in this - ASHPs and GSHPs are very hyped right now so a lot of the information that comes up in searches is quite low level or low quality. For example I’ve been told that a heat pump is at peak efficiency when it is sized for a duty cycle of 50% but asking why exactly this figure hasn’t yielded good answers.

As I understand it, it comes down to a couple of things.

Heat pumps are much more efficient when providing lower temperature flows, their efficiency drops off significantly where the delta between the heat source (air or ground) and their flow temperate is higher.

When an oversized heat pump cycles it is going to quickly reach a higher temperate and shut off where as a correctly sized unit will cycle for longer. So I guess simply it comes down to having the heat pump operate in the most efficient range for that property for longer vs. short spikes at temperature where it is less efficient.

Most heat pumps have a variable speed motors so they can modulate their output to match a desired flow temperature but this only operates within a range relative to the size of the unit. So if the unit isn’t correctly sized for the required heat output to be in its range then it has no choice but to cycle inefficiently.

A couple of sources I found helpful:

https://assets.publishing.service.gov.uk/government/uploads/...

Most of the consumer advice is very low level - this Heat Geeks site seems better that most, but it still lacks enough detail for me to get into the numbers: https://www.heatgeek.com/3-steps-to-maximise-your-heat-pump-...

Thanks. Yeah, especially difficult when you read multiple different claims and none have a source. They probably heard it themselves from someone else.

I'm quite curious because I have an old-ish heat pump that's a little small for the space and I've been thinking of getting a larger one. But if it'll be less efficient then maybe I won't. The one I have does perfectly fine except on a few of the coldest winter mornings.

My 7kw unit can only possibly draw 900ish watt.

800 watt for the refrigerant pump, and about 100 watt combined for the two fans.

This means even when I'd like it a little warmer on the coldest days, making the heat pump use more power isn't an option.

A larger unit would have a bigger refrigerant pump and larger fans. This would result in us inadvertently leaving the unit on a higher heat setting more often.

Also, a larger unit will be turning a larger pump and larger fans even when smaller would achieve the same outcomes. Mechanic losses.

These two factors, human habit & mechanical losses, are, as I understand it, why you're generally better off having more small systems than one big system.

We have three split systems, in addition to the 7kW unit at one end of the house, there are two 2.4kW units, one I'm each of the bedrooms we use. When one or more units aren't required they can be turned off. Another advantage of this setup is that it's extremely unlikely all three systems will breakdown simultaneously.

Another issue with the larger units is they tend to be ducted, where the ducting is in a poorly insulated roof space, so there's additional heat losses there, and additional losses due to energy required to push air through the ducts.

Are you sure? Every ~7kW unit I’ve seen pulls about 2.6kW electrical.

Agree that multiple smaller split systems is the way to go though.

Your first point is a bit silly: get a smart thermostat and it’ll take care of that for you.

Your second argument is silly: a larger pump will surely take more energy to run but will run for less time. It’ll also wear out slower as it doesn’t have to work as hard unless you have drafts and it has to cycle on and off all the time. Generally having more capacity is always a good thing for mechanical things.

It is explained like this in the thread.

> if you have a large compressor, you pay more for it with every revolution the trick is to figure the maximum BTUs or watts you will need, and design a little bit smaller. Plan to use a supplimental energy source to fill in during extreme conditions.

Worth noting the AC_Hacker paid $25 for the AC he used for this.

I don't think it's the temperature delta in this case, at least not between the room and outside. I suspect internal equilibration of the unit when power cycled and also energy transfer design capacity on the indoor side of the unit are to blame.

> Most of the consumer advice is very low level - this Heat Geeks site seems better that most, but it still lacks enough detail for me to get into the numbers

You'll probably want an undergraduate physical chemistry textbook for this.

I think it takes a while for the juices to equalise and reach equilibrium before it’s running at max efficiency. You don’t want a duty cycle of say 15mins. You want say 1.5hrs. Something like that. (Numbers very hand wavy this is not my field directly) With an over designed system it stops and starts too quickly and this invalidates the warranty on commercial systems.

Modern heat pumps drive the VFD and circulation pumps as needed (i.e. the pumps run slower when the heat demand is low).

The main factor for good efficiency is low output temperature- underfloor heating works at ~35 deg C, radiators work at ~55 degC (because of smaller surface area).

I sized my heat pump based on historical data of heating oil /diesel consumption- since I had the data of past two winters, I could reliably determine the maximum continuous heat requirements for my house.

It also makes the environment itself uncomfortable in many cases - excessively drying the air, or causing weird moisture/condensation issues (as it won’t fully cycle the air in the house before stopping).

Essentially, you want your HVAC system to run for long enough it’s a gradual transition in temperatures inside and the equipment can run for awhile each time it needs to run to avoid the start/stop problems and let it ‘settle in’.

Not sure what the right terminology is, however, it's due to it turning off, on, off, on again and again...

When it is slightly undersized, it'll run continously and also prolonging compressor life...

Compare this to being able to drive continously at the same speed vs. having to stop/start every 500m or so...

duty cycle, i think.

GSHP don't have as much issue with low temps as the below ground temp is fairly stable year round.

Brine temperature varies over the heating season by up to 10 deg C.

I really tried to verify the horizontal coil length and installation depth recommended by installers from the first principles, but I had to give up, as there were just too many unknown inputs: the type of soil (thus, specific heat and heat conduction), water content (which varies over the year), received solar energy (shady/sunny).

Reply on news.ycombinator.com