I have a small cabin next to my house in north central WA (climate zone 6b). Winter nighttime temps range down -15F, though +20s are more common. This winter daytime temps have varied from -5 to mid 30s.

The cabin is poorly insulated and not very often used in winter. I keep it at around 45F so the pipes won't freeze. Until this year, it only had baseboard heat. After some remodeling of my main house, I moved a Mitsubishi MSZ-GE12NA-9 and MUZ-GE12NA2 over to the cabin. Thanks to this forum, I found the secret Mitsubishi site with efficiency tables. It shows a COP of 2.87 @ 17F for this model. The instruction book just gives a "guaranteed operating range" of -4F. Neither place shows what the efficiency may be at, say, 0F.

The lowest thermostat temp the minisplit allows is 60F. The first question is whether keeping the cabin at 60F with the minisplit will be more costly than keeping it at 45F with the baseboards.

I assume when the cabin is in use and heated to 70F, I should usually use the minisplit. But that leads to the second question. At what low outside temp should I use the baseboards because the minisplit gets even less efficient?

I would be happy for any advice, including to references I could check myself.

I have had some emails with Mitsubishi. The rep first told me that there was no difference in heating efficiency at lower temps, just in the amount of output, down to the -4F where the unit would go on standby. Lower outdoor temps = less BTUs.

I pointed out what the table mentioned above says about the COP dropping at lower temps. He said he had only been looking at the HSPF when he said the efficiency didn't change. Then agreed that it did drop as temps dropped.

I wanted to know how much it dropped when the temp got below 17'F. Also, the bottom line question above about when the minisplit becomes less efficient than the baseboards.

At this point in the exchange, the rep said I obviously knew more than the typical residential customer and had the info to figure it out by myself. Actually, I don"t. But maybe it is too complex or there are too many variables to figure it out.

Randy, try asking your questions with their Tech Support people...

That is what I thought I was doing. In any case, I'm not sure how I'd get through to more techie people. But maybe it doesn't matter that much. The difference is probably marginal

The answer to your question is too complex requiring a simulation model. To get it simply run both stand alone look at your utility bill.

Posted By PARAHOMES on 25 Jan 2017 05:26 AM
The answer to your question is too complex requiring a simulation model. To get it simply run both stand alone look at your utility bill.

I think you're right about the simulation model being too complex. But the utility bill won't help, either, at least at the margin I'm talking about. Outdoor temps are too variable. They would have a greater effect on utility bills. I'd have to do some kind of calculation based on hourly temp averages over the month of the bills, or at least by weekly reading of the meter.

I am still wondering about whether there is any efficiency data for my model below 17F, down to its rated -4F shutoff. I couldn't find one.

Going back to the simulation model point. Maybe there is a simplified approach. (I am a total newbie at this stuff, I should add.)

500 sq ft house, R-11 walls, R-18 ceiling (if I'm lucky), 15%-20% of outside surface is windows -  crappy old aluminum double-glazed windows, floor sits on dirt and rocks, except one room with a bit better than R-11 insulation under floor. Climate zone 6-B. Is there some formula or software that says what BTU I need for a target temp of 45F vs 50F, either keyed to average or a specific outside temp, like 15F?

I think is probably safe to say that the heat pump is never going to be less efficient than electric baseboard.

I'm using a Spokane airport weather tower... seeing dry bulb temps down to -15, wet -25F max load. Using Seer 18/9.6 HSPF ductless MS? Electric baseboard 100% eff.

With 50 F inside set points.

Running baseboard only all year ~ 8kBTU/r difference better than HP. Yes marginal. Annual utility bills ~ $10/yr more for HP.

When it becomes more efficient hard to tell w/o major zoom to the hourly graphs are close to the same, blue HP, orange BB overlay....

This seems odd.  9.4HSPF Heat pump should be much more efficient than baseboard at most temps you see in Spokane... what am I missing?

I have a Fujitsu RLS12 mini-split (12k BTU rating) heating a relatively uninsulated 1600 square foot 2 story 1926 bungalow that I will start a complete renovation on soon. For 3 years, it has been set to 60 degrees F and just puts out heat. When it gets cold and can't keep up, the oil-fired furnace will kick on and deliver some heat. For the most part, the mini-split has kept the house above 50 degrees F all the time. I have used only about 30-75 gallons of oil in a year (more three winters ago, less last winter and this winter, which were warmer). In a climate that averages (historically) about 7500 degree days in the winter.

So, I would say you could run your Mitsubishi the same way and save a load of money. Set the baseboard electric to 45 degrees F, and let the minisplit go. When the temperature drops below 60 deg F, it will just keep going. If it stays too cold for too long, the baseboard will come on. If it warms up during the day, the temperature will start to recover to 60 degrees F as the minisplit can start putting out more BTUs. Either way, the minisplit will be more efficient to run than baseboard.

Thank you all for your comments. Looking at the last one, I guess I will have a very rough answer when I get the next electric bill for my cabin. I say rough because I don't have any locally valid degree day stats. I will just see if there is any obvious drop in electric usage. It will also be rough because of irregular use of the cabin.

By the way, kjohnson, I keep the cabin at 50F (when unoccupied) with baseboards at 45F.

The GE12 drops off in efficiency & capacity a lot quicker than the FE12 or FH12 at sub-zero (or even sub- +20F) temperatures, and isn't really a great choice for a location that regularly hits -15F. (What's the elevation there, anyway? That too affects both efficiency & capacity.)

The lowest setpoint you can deliver with the remote is 60, but it's possible to customize a remote that simply turns it off when the temps are above 50F, if you're really into it. The temperature sensor circuit in the head may also be tweakable to "think" that it's 60F when it's only 50F, but I've never taken at close enough look at it.

More recent-vintage Fujitsus (IIRC all RLS2 series & newer) have a "Minimum Heating" mode that keeps it at 50F, even though the lowest programmable temp with the remote is also 60F.

Posted By patonbike on 25 Jan 2017 04:58 PM
This seems odd.  9.4HSPF Heat pump should be much more efficient than baseboard at most temps you see in Spokane... what am I missing?

And it's even more efficient when the indoor temp is 50F instead of the 70F indoor temperature at which HSPF is tested. Throw out the HSPF number, look at the COP at a comparable delta-T and modulation level.

eg. 70F-17F a delta of 53F is included in the HSPF test, but the modulated level at which efficiency is tested at 17F isn't always at maximum-modulation,(it's usually a some lower output) and you may have to dig a little deeper. But running at +50F indoors the same delta-T would occur at -3F, and the efficiency & capacity at -3F would be comparable to (but not identical with) the published COP at +17F in an HSPF test submittal.

Most extended capacity charts in the engineering manuals don't go lower than 60F for an indoor temp, so you're a bit out of luck there too.

Bottom line, the HSPF is a bit meaningless when operated like this, and barely meaningful even in normal set ups, since the sizing relative to the average load has a huge influence on the as-used efficiency of modulating heat pumps. If oversized for the load such that it rarely modulates it can be hard to beat a seasonal average COP of 2 in a zone 6 climate, and the same is true if it's so undersized that it runs at max-speed most of the time. When it's above +10F outside the actual efficiency can be quite a bit higher at max speed than at min, but at sub-zero temps it'll be a COP of 1.5-2 independent of modulation level. At -15F and colder the GE12 is likely to be deliver better than a COP of 1, and with defrost cycles it might be less than 1.

In a location that sees extensive hours a sub-zero temperatures you'll need a pan heater to keep the outdoor unit from destroying itself on built-up defrost ice. I don't know if they even make one for the GE12, but they do for the FE12 & FH12. Mitsubishi will void the warranty for pan-ice/fan damage in a US zone 5 or colder climate unless the pan heater option was installed. It only comes on when necessary, but it also cuts into the as-used COP/HSPF

Couple of points, Dana1. And please keep in mind I am just trying as a homeowner to understand this. I am somewhat of a computer techie but have no experience this kind of engineering.

The GE12 has a min temp setting on the remote of 50°F. It is accessed by using Smart Set, which is otherwise is only meant to create memory settings. Don't ask me why Mitsubishi hides the 50° mode there, but they do it for many models.

I am at 1,560' altitude. I don't know exactly what average night temps are, but I would guess that in recent years they are above 15°. We have a few nights or a week here and there below 0°. Daytimes below 0° are rare. This winter we've probably had 5 nights at -10° or a bit lower.

So far I have not seen any defrost ice under the minisplits. I also have a dual model in my main house, but I haven't raised that issue here because my main heat source when temps are really low is wood.

If my COP is at least 1.5, and usually much more. I am happy.

Thanks for the tip on the "Smart Set" function (it's news to me!)

At 1560' you don't have significant capacity degradation. When you mentioned -15F I thought you might be up some canyon off one of the higher passes or something. The clearer air of the eastern slopes of the Cascades makes for bigger diurnal temperature swings, but also less absolute humidity, and less defrost icing concern. As long as the daily highs go north of freezing semi-regularly you won't pick up much ice in the pan, and the self-destruct risk is low.

If you aren't there regularly to monitor in person, hopefully it's mounted above the historical peak snow-drift level. I'm always amazed when I see them ground-mounted on condenser pad in snowy areas, but it's a really common error.

You might dip below a COP of 1.5 for a few 10s of hours per year, but your average should be north of 2.5, but probably not 3. For some in-situ third party testing in different climates under more normal indoor conditions, see:

http://www.neep.org/sites/default/files/resources/NEEP%20DHP%20Report%20Final%205-28-14%20and%20Appendices.pdf

and

http://neea.org/docs/reports/ductless-heat-pump-impact-process-evaluation-field-metering-report.pdf?sfvrsn=18

Thanks, again, Dana1 for the info and links. As for the snow, there is no problem. My house is 50' from the cabin. I can actually see the compressor from my dining room window. More important, it is on a covered deck. In a blowing blizzard I might see a couple of inches blown up to the compressor, but even that is pretty rare.

My main house compressor is a bit more exposed, but still has quite a big roof overhang. I have seen snow go higher than the raised bottom after several roofalanches. They start with a big pile of snow a few feet from the compressor but can sometimes build up and slide back. I'm well aware of that potential and keep my eye out for it. It happens once or twice each winter. Five minutes of shoveling takes care of it.

Been busy writing a sell sheet mini/erv my proposed design report for a Dr client that never heard of them, typical, said he wants to research it and I bet he ends up with the same questions……this is an interesting one to model see how close I can get.

I see GBA blogger D1 chimed in now we are on the slopes of the eastern cascades skiing?? Hot coco on the house @ RandySea Cascade Cabin Oasis? I’d been more impressed if you predicted the unit was not freezing before he told you but, I do admire the absolute humidity theory to explain. Oh, how simple complex physics can become when needed.The rest of your post made me feel like I just got out of a Mitsubishi factory test lab, been there done that many times & wrote those final test reports you are reading that cannot be used alone in situ like this, can be way off I'll prove to you all soon. Not sure what that all had to do with one at a job site or the answer to the question.

<!--[if gte mso 9]> Normal 0 false false false EN-US X-NONE X-NONE

 “I moved a Mitsubishi MSZ-GE12NA-9 and MUZ-GE12NA2 over to the cabin. "

NA9 wall mount 1 ton indoor/outdoor system, 12NA2 1 ton outdoor unit. 22 SEER, 11.4 HSPF, single zoned.

Have inverter driven compressor valves that continuously modulates compressor or heating and cooling vs stop/start.

Now that I took the time to look deeper, your running 2-1 ton outdoor units 1- 12K indoor, 500 SF, both not shown in NEEPs cold climate database. It explains the issues with HSPF in low cold temp performances (tested @ steady state, sub-17F), provides better info than mfgs. These properties are related and one should not be changed w/o changing the other.

http://www.neep.org/initiatives/hig...-heat-pump

The E-Plus sim model I ran used CC-MSHP VRF flow that allows for more accurate performance at part load and low temps with capacity retention and impact of pan heaters on performance. Retention of heat capacity at colder climates is highly related to system size.

I applied a Heating Capacity Retention fractions to the model, not sure what yours are so I’ll use values for MSZ-FH12NA2 26 SEER, 12.5 HSPF to see how it would perform in 6. Max heat retention 65% @ 5F /47 F max, 100 % @ 5F/17F maintenance capacities. -5F is the outdoor dry bulb temp at which the fraction is defined. Limited COP points can be misleading so, the sim model uses a liner relationship between this point and the rating point to determine max heating capacity as a function of outdoor ambient temp and sensible/latent heat curves from a TMY3 weather tower.

I’ll keep the party in Spokane since it’s the closest tower I can find at elevation ~ 1850'. Vapor drive based on temperature of water, absorbent state, not air or absolute humidity alone, are the determining factors of frost/thaw cycles but, to simplify the model uses RH most Z6 area winters >60%.   

Pan heaters typically operate when the compressor is running & outdoor dry bulb temp thermistor is < 32F & typically reduces HSPF .1-1.0 points. NA2 uses 106W that won’t have much impact on COP about 15% reduction.

Other values I used:

Heat Capacity Retention factor is X temp (5 & 17F)/47 F.  

Heating capacity offset (difference between heating/cooling capacities) 1,600 BTU/HR

.40(RC) min cooling, 1.20 max. 

Min rated cooling capacity:  2500 min/12K rated = .2btu/hr

4.      Max rated cooling capacity: 13,600 max/12K rated = 1.1

5.      Min/Max Heat flows/ton: 200-400. Cooling: 200-425 cfm/ton

6.      SHR .73

7.      Fan: .07 W/cfm

Ra Ran two conditions,

1.      1. Auto size for max heating/cooling load.

2.      2. Size up to 1.3X for VSR in colder climates when heating exceeds cooling (per ACCA Manual S)

 

-5F retention temp results shown. Auto-sizing shows more heating at max or using Manual S. SML-F (size for max load false, uses MJ/S) SML-T size for max load true used simulated  design conditions). Resulted the same more heating loads in Z6.

 Required MJ8/S capacities are for more efficient unit than your models so your capacity requirements s/b higher. I don’t think a 1 ton is going to cut it according to it try 2.

E-PLUS supposedly more accurate, less conservative, less empirical than MJ8 shows heating capacities under 1 ton. MJ8 50%+ oversizes too much this is much lower...

Both BB(blue) and CC-MSHP(orange) are superimposed in graph charts. When I zoom in close you can CC-MSHP is delivering slightly lower btu/hr stand alone to maintain 50F in a small 500 SF cabin, Spokane, in the month of Jan where your peak loads are. 

As far as stand alone site metered peak power consumption CC-MSHP is far less I’d be running it anytime I could, use BB for back up only.

<!--[if gte mso 9]> Normal 0 false false false EN-US X-NONE X-NONE -->

Here this ones interesting...check out the power modulation based on environmental loads depending thermistor type.

Just curious. If the cabin is rarely used in the winter, why aren't you draining the plumbing between visits? All you need is a valve at the lowest point -- a frostproof hose bib in a primitive cabin, for example, or a wye off the commode supply with a hose over to the shower drain in a fancier one. Open the taps. Flush the commode. Turn down your water heater as far as it will go, or switch to a tankless. Things will be chillier when you do visit in the winter, but huddling around the wood stove is part of the fun.

why aren't you draining the plumbing between visits? All you need is a valve at the lowest point

[Sigh - if only!] The cabin was built in stages over 60-70 years. Part is on dirt + stones, part on concrete, part on a poorly insulated crawl space. It has two full bathrooms at either end, and a kitchen in the middle.
I have no idea how many "lowest points" there may be. That would be impossible to add under most of the cabin.

We could have a plumber come and blow out the pipes. But they would not guaranty that blowing out the pipes wouldn't break a joint, possible where we'd need a jackhammer to dig down to fix/replace it.
And we'd have to blow the pipes after each time we used the cabin.