Most efficient way to use a heat pump?

[nola mike]

Feel free to skip the long preamble and go straight to the question.
My vacation place has a heat pump, in addition to my undersized insert. I keep it at 42' when I'm not there. It usually doesn't have too hard a time keeping it at that temp, and I disabled the strip/aux heat. When I go there, if I'm getting there at night the stove won't be able to bring the house up to temp before bedtime. So I remotely turn up the HP to 55-60'. Since I don't care how long it takes to get up to temp (I can turn up the heat in the AM), I don't need the aux heat. Last year however, we had a particularly long cold snap; single/some neg numbers at night, teens during the day for >1 wk. The HP was barely able to keep the place above freezing, and the compressor was running non-stop. I got a huge bill.

SO...
At what point is the efficiency of the compressor offset by extended run times? Or I guess, how many hours of compressor time = 1 hour strip time? Obviously that depends on the HEER of the unit (which I don't know ATM) and the power of the strips (which I have no idea about). And the outside temps I suppose. Hmm, this is becoming complicated.

 

[begreen]

Extreme cold will tax any heating system, especially if the place is only moderately insulated. What heat pump make & model?

 

[Corey]

Your "CoP" or Coefficient of Performance - basically heat pump "efficiency" with a nod to the fact the "efficiency" can be well over unity / 100% under the proper conditions - will change with temperature. In warmer temps, the CoP can easily be 2 - 2.5 or greater...meaning for every KW you spend 'pumping' heat, you get 2 - 2.5 'KW' of heat in the house. With lower and lower outdoor temps, the heat gets harder and harder to 'pump' so overall less and less efficient.

You really take a hit when the outside temp falls below freezing because the outdoor coils will ice up from condensation. If it's above freezing, the unit can shut off for a while and let nature melt the ice away. Below freezing you have to kick on a defroster circuit and/or run the pump in 'AC' mode to get some heat back in the outdoor coils and melt the ice away. Either way, you're 'pumping' some heat back outside.

The manufacturer might have a CoP table, or at least a recommendation when it is no longer efficient to run the heat pump. I would suspect mid 20's or so would be the limit...possibly a bit lower, or higher, depending if your electricity is cheap or expensive, respectively.

 

[woodgeek]

Goodman published some tables for COP versus temp, and they say the COP is still 1.5 or so at -10°F for my SEER 16 unit. That does not include defrost effects, but during most cold spells it can get pretty dry.

Bottom line, if it's a newer vintage (say less than 10 years old and SEER 13 or better) it will save some money to run it down to -10°F or colder.

As for running it cheaply, I would be tempted to program the stat to step up a few degrees around noon, and to step down a couple degrees at midnight. Might force the system to make more BTUs when its warmer outside?

 

[nola mike]

@begreen : It's a Trane, I'll get the model this week when I'm down there.
Interesting about the CoP. Does that mean that as long as the CoP is >1, I'm better off pumping, regardless of how long the compressor is running?

 

[begreen]

The COP is just a measure or efficiency. It doesn't take into account energy costs. My guess is that the Trane is not up to the job without being able to kick into the strip heaters. We have a similar high efficiency American Standard unit that switches over to the strip heaters at 24F. Contrast that to good mini-split systems from Mitsubishi, Fujitsu and Daikin that are able to heat well at zero without switching to resistance heating.

 

[woodgeek]

@begreen : It's a Trane, I'll get the model this week when I'm down there.
Interesting about the CoP. Does that mean that as long as the CoP is >1, I'm better off pumping, regardless of how long the compressor is running?

Yes, if all you are asking is energy per BTU for HP versus strip heat. Then there is 'wear and tear' on the compressor. If the manufacturer says to shut it down at some temp, do so, otherwise I say let it rip.

 

[peakbagger]

sounds to me like you need to look into a few minisplits like Mitsubishi Hyperheats. They are optimized for cold weather operation

 

[maple1]

Is there a chance your unit depends on the aux heat strips for defrosting?

So by disabling them you also disabled defrosting?

 

[nola mike]

Is there a chance your unit depends on the aux heat strips for defrosting?

So by disabling them you also disabled defrosting?

Yes, that's possible (probable?). Don't want to damage the unit, so I'm going to reenable them. Just want to figure out the best strategy for their use. Like I said, I hadn't seen a stretch of cold anywhere near that in my 8 years in VA.

 

[velvetfoot]

I was fooling around with my Ecobee thermostat recently, although I don't have a heat pump. You can strategize with your thermostat on the different stages or aux heat and when to turn them on.

 

[woodgeek]

Yes, that's possible (probable?). Don't want to damage the unit, so I'm going to reenable them. Just want to figure out the best strategy for their use. Like I said, I hadn't seen a stretch of cold anywhere near that in my 8 years in VA.

I disagree. In defrost the HP takes heat from the house to melt the ice, and just fires the aux to heat the air out the registers for owner comfort (adding the heat **downstream** of the HP coil). There may be an issue if the indoor temp is 42°F, but probably not I would **guess**.

Good news: with a massive El Nino shaping up, we might not get any real winter weather in the mid-atlantic region at all. It is very unlikely to look like a repeat of the last two years.

 

[maple1]

I disagree. In defrost the HP takes heat form the house to melt the ice, and just fires the aux to heat the air out the registers for owner comfort (adding the heat **downstream** of the HP coil). There may be an issue if the indoor temp is 42°F, but probably not I would **guess**.

Good news: with a massive El Nino shaping up, we might not get any real winter weather in the mid-atlantic region at all. It is very unlikely to look like a repeat of the last two years.

Is that true of ALL heat pumps?

(I have no idea, hence the question...)

 

[iamlucky13]

Goodman published some tables for COP versus temp, and they say the COP is still 1.5 or so at -10°F for my SEER 16 unit. That does not include defrost effects, but during most cold spells it can get pretty dry.

I have a Goodman (16 SEER, 9.5 HSPF), but I don't think it has a humidity sensor or other means of better determining when it can skip or delay a defrost cycle. They seem to run as a matter of course, so I assume it continues to hold efficiency back regardless of how dry the air is.

Using Goodman's chart, extending assumptions about how HSPF relates to COP way out of the reasonable range, and factoring in modest duct loss, I'd guess my net COP drops to 1 at around 0 deg F.

Even beyond the bad assumptions I made, there's more to this, because heat pump output declines as the temperature drops. Mine puts out a little less than half the heat at 0 deg F as it does at the 47 deg design temp. If the heat pump is sized to a balance point around the average winter temperatures, as is the norm, it genuinely can't keep up at the extremely low temperatures. It will run all the time, and still fall behind.

At those really low temperatures, you're accruing a lot of run time on your heat pump, but it's saving you little to no money. I'd set the aux heat lockout at a temperature more like 10 degrees, and wait and watch what your bills look like. Comparing will be hard, however, because outdoor temperatures vary by year.

 

[nola mike]

It's actually not usually an issue when I'm not there. Keeping a house at 42' is not typically that difficult. Normally, the worst we see at night is low 20's and daytime is rarely below freezing. Most of the strategy would come from when I'm trying to warm up the house from 45-->60. At this point, the T-stat says "Oh chit, I'm 18' behind, better turn on aux!". Or, at some point, "Oh chit, I've been running the compressor for 4 hours and I'm still not up to temp, better turn on aux!". I have to see how I can better tune the oh chit setpoints on my ecobee--I had some trouble with it last time. Also looks like my indoor setting is 45', which is the lowest the ecobee allows. Was 42' on the old stat.

EDIT: Actually was able to find some interesting info from my ecobee. I can get a graph of runtime/day. It also gives me the outside temps. I have a few data points in Jan/Feb, but for example, with outdoor temps of X, the compressor ran for Y hours. These are in order of ascending outdoor temps (average for the day temps, I assume).
X/Y
10.2/20.8 hours
14'/10.3
16.6/14.9
18'/12.3
19'/9.6
20.1/9.1
21.5/9.5
24.6/6.2
24.8/14.3 <--this was after the 10' day
26'/6.2
29.4/4.8
30.8/4.9
31.6/4.9
32.2/5.3
36.6/7.1 <--an example of when I turned it up prior to arrival. Indoor temps 53' here.
37.7/2.1

So maybe I remembered it worse than it was. A bit difficult to interpret because these are average temps, so what it's doing at any specific time is unknown. Still interesting.

EDIT #2: Looked at the hourly stats for that 10' day, and the day before. Here's outdoor temps v. runtime/hour. Indoor temps were 43-45' for every data point unless noted. Outdoor temps are based on web data, not actual temps.

24'/28 minutes
21'/28
19'/39
18'/36
17'38
16'/39
15'/59
15'/60
15'60
15/19 (10am)
16/18
17/27...(noon)
...7pm
14/60
8/60
4/60
4/60
8/60 (11p)
12a
11/60
10/60
9/60 (indoor 42')
8/60 (indoor 42')
8/60 (indoor 42')
7/60 (indoor 42')
6/60 (indoor 41')
6/60 (indoor 41')
6/60 (indoor 42')
7/60 (indoor 42')
9/60 (indoor 43')
11/60 (indoor 44')
12/60 (indoor 45')
13/28 (indoor 46') 1pm, don't know why it seems to have caught up here?
15/27
16/22
18/24
18/34
17/37
14/60 (7pm)
8/60
4/60
4/60
8/60
After that there are 10 more hours of 13-19' temps with 60 minute run times.

 

[nola mike]

So looking at that data, it seems reasonable to use the compressor exclusively at anything >20-25' or so. Starts to struggle once you get into the teens. It would also seem that with a <50% duty cycle, I still have some reserve, so could heat up the house to 60' if given enough time, with the compressor likely running 100% * X hours. Of course, I'll have much greater heat losses at 60 v 45'.
II don't know if any of this brings me closer to an answer, but having fun geeking out with numbers and graphs

 

[woodgeek]

I have a Goodman (16 SEER, 9.5 HSPF), but I don't think it has a humidity sensor or other means of better determining when it can skip or delay a defrost cycle. They seem to run as a matter of course, so I assume it continues to hold efficiency back regardless of how dry the air is.

Bad news: I know way too much about Goodman defrost controllers, having hacked mine a number of different ways over the years, and gotten a decent boost to COP for my efforts.

You are correct @iamlucky13 the stock controller is 'dumb' and just cycles every fixed amount of run time, settable at 30, 60 or 90 min intervals with a jumper. BUT, it is still adaptive because it runs the HP in reverse only until the coil reaches 50°F, at which point the ice is melted. If the coil is dry and there is no wind, this can take as little as 90 seconds. If the coil is loaded due to high outdoor RH, it can take 3-4 minutes of reverse running per cycle to melt the ice. In heavy snow or freezing rain (not a lot of hours) it can take 6-7 minutes per cycle.

Obviously, the amount of energy spent in defrost is proportional to the frequency and duration of defrost events. I set my timer to 60 minutes (versus the 30 min my installer selects). It has never been a problem, and cuts my defrost costs by nearly 50% for moving a jumper. And then, in dry weather, the unit does these real short 1.5 minute defrost runs every 60 run minutes....not a huge hit compared to wetter weather with it bogging down in long defrost cycles.

The simplest 'hack', BTW is to put a 'delay on' timer on the 24VAC line between the defrost controller and the aux contactor. I noticed that there is enough hot air in the ducts that there is no cold air out the registers for 2 mins or so into the reverse cycle. So I set the delay timer for 2 minutes. This has the effect that the aux only kicks on 2 minutes into the defrost cycle. I get no cold blasts, and during dry weather, no aux is used during defrost.

 

[iron]

i may be wrong, but older heat pumps struggle at temps <32*F. newer heat pumps, i believe, are better up to 0*F. at least, this is what i vaguely recall reading a year or so ago when looking into replacement systems for our house.

 

[begreen]

That's too large a generality iron. You can still buy units that will only work down to about 32F. These are more commonly sold in southern states, but available nationwide. The mini-splits are the champs when heating down in the zero degree range. Fujitsu, Mitsubishi and Daikin have some good units for this application.

 

[iamlucky13]

Bad news: I know way too much about Goodman defrost controllers, having hacked mine a number of different ways over the years, and gotten a decent boost to COP for my efforts.

Very interesting post. In our humid climate, I don't think I'd want to go 60 minutes without a defrost, though. I'd have to keep a very close eye on it on a foggy day if I were to try it. That jumper is under the cover in the corner of the condenser that protects the wiring?

I also noticed that the timer does not seem to reset once the heat pump shuts off. So if it accrues 29 minutes and then shuts off, the next time it comes on, it goes almost immediately into defrost.

On a tangent, I can't say I've been thrilled with my Goodman. It does seem to be heating roughly in line with expectations, but it was only 3 months old when the condenser fan died (parts replaced under warranty, labor donated by the installer, who griped that Goodman doesn't reimburse him, even for 1st year failures like that). It's now got 4 winters on it. For the last 2, every now and then, the reversing valve seems to stick when it comes out of defrost. It makes a hiss like the valve normally does when it switches, then a growling/rattling noise and fails to produce any heat. I have to manually shut it off and let it cool down for 10 minutes or so, after which it starts up normally.

I'm keeping an eye on it, at the suggestion of the installer, who even though he's admitted he questions Goodman's quality these days, is pretty confident that at a minimum the compressor is bulletproof.

 

[woodgeek]

Very interesting post. In our humid climate, I don't think I'd want to go 60 minutes without a defrost, though. I'd have to keep a very close eye on it on a foggy day if I were to try it. That jumper is under the cover in the corner of the condenser that protects the wiring?

Your 100% humidity is the same as my 100% humidity. 60 minutes has been fine here, it doesn't get choked up. Installers use 30m b/c they are not paying the bills. My installer also told me I was 'crazy' to have a HP in PA, and I should turn it off below 40°F, after he installed it of course.

I also noticed that the timer does not seem to reset once the heat pump shuts off. So if it accrues 29 minutes and then shuts off, the next time it comes on, it goes almost immediately into defrost.

Indeed. The frost does not (usually) melt when it is off, so it is an elapsed runtime timer and that is what makes sense.

On a tangent, I can't say I've been thrilled with my Goodman. It does seem to be heating roughly in line with expectations, but it was only 3 months old when the condenser fan died (parts replaced under warranty, labor donated by the installer, who griped that Goodman doesn't reimburse him, even for 1st year failures like that). It's now got 4 winters on it. For the last 2, every now and then, the reversing valve seems to stick when it comes out of defrost. It makes a hiss like the valve normally does when it switches, then a growling/rattling noise and fails to produce any heat. I have to manually shut it off and let it cool down for 10 minutes or so, after which it starts up normally.

I'm keeping an eye on it, at the suggestion of the installer, who even though he's admitted he questions Goodman's quality these days, is pretty confident that at a minimum the compressor is bulletproof.

Bummer. I had a SEER 14 unit for 6.5 winters that finally burned out a compressor after a very heavy ice storm and freeze up (that was likely caused by a failure mode of a very complex hack to convert my defrost to demand only ).
No problems with my SEER 16 so far, and keeping my hacks to a minimum (just as described above, 60m + timer).

 

[woodgeek]

Is that true of ALL heat pumps?

(I have no idea, hence the question...)

The AUX is downstream of the A coil in all HPs I have ever heard of. If it was the other way, the HP would have to pump heat into a higher temp stream, the higher lift would kill COP and hot refrigerant would get sent to the compressor. Ugh.
In dual fuel systems the A-coil is typically downstream of a furnace, like with a central air system, and the stat has to be wired to run either the furnace OR the HP, and never both at the same time or damage will occur.

 

[iamlucky13]

Your 100% humidity is the same as my 100% humidity.

We get a fair amount of fog, and I'm sure those little droplets being pulled through the heat exchanger love to freeze on the fins. Not to mention, frost, by reducing heat transfer, slowly reduces the efficiency of the unit as it builds up. Even short of damaging the unit, at some point, the savings from reduced defrost cycles would presumably be offset by the lower performance.

There's a bigger gap than I'm entirely comfortable with between 30 minute and 60 minute defrost settings.

Indeed. The frost does not (usually) melt when it is off, so it is an elapsed runtime timer and that is what makes sense.

When it's 50 degrees and the heat pump has been off for an hour or more because of the mild weather, it does, which are the conditions I noticed this happen in a couple weeks ago. A reverse timer, running at a rate proportional to the temperature (dropping to zero as the temperature approaches freezing), seems like it would be effective.

 

[woodgeek]

We get a fair amount of fog, and I'm sure those little droplets being pulled through the heat exchanger love to freeze on the fins. Not to mention, frost, by reducing heat transfer, slowly reduces the efficiency of the unit as it builds up. Even short of damaging the unit, at some point, the savings from reduced defrost cycles would presumably be offset by the lower performance.

We get plenty of freezing rain, snow getting into the fins, fog, you name it. I have watched output using my ecobee charts, and I have never seen any indication of the above issues. Even when there is frost, there is still airflow through the frost, just at lower cfm. This causes the coil temp and BTU output to drop slightly, but the effect on COP is pretty small, because the elec consumption falls with the coil temp.

The difference in SCOP between a 30m timer and a 60m timer btw is about 0.5, i.e. from 1.7 to 2.2 at 32°F, under dry conditions.

When it's 50 degrees and the heat pump has been off for an hour or more because of the mild weather, it does, which are the conditions I noticed this happen in a couple weeks ago. A reverse timer, running at a rate proportional to the temperature (dropping to zero as the temperature approaches freezing), seems like it would be effective.

My HP doesn't defrost at all above ~40°F, and at those temps, the frost doesn't passively melt significantly during the off cycle.

 

[nola mike]

Here's my heat pump info:
York Affinity 8T
Model YZE03811A

EDIT: Huh. Looking at the install manual, seems that this is a 2 stage unit? I don't think the T-stat has it set up like that...