heating costs of mini split vs gas?

Jells

I have several 500 ft apartments currently heated by gas fired, non-condensing hydronic. They're 3 room railroads, 2 larger rooms with a smaller kitchen between, and they would seem good candidates for a 2 head mini-split system instead.

I think having the built in AC would make them attractive to tenants, but not if the heat costs are over the top. I'm in NJ where energy costs are high. Searching around finds a lot of promotional hoopla, but little apples to apples comparisons. Any straight poop out there?

TAG

Mini splits are great problem solvers -- but with NG a conventional system is the way to go.

JUGHNE

How is the NG metered now and how would the AC/HP's be metered?

Jells

JUGHNE said:

How is the NG metered now and how would the AC/HP's be metered?

Both are on the tenant's own account if that's what you're asking.

pecmsg

You’ll need Both.
Minis are great but you do need a back up!

Jells

pecmsg said:

You’ll need Both.

Minis are great but you do need a back up!

Can you elaborate? Why would it still need the hydronic baseboard if the pump can provide enough heat? These little places don't take a lot of BTU's. Most of them have less than 30' of fins.

JUGHNE

There are claims of heat delivery at extremely low outdoor temps. They will deliver some heat but perhaps not enough.

IIWM, I would install the units but look into an adjustable low ambient shut down to stop the unit from operating at perhaps 20-25 degrees. Operational shut down but keep the compressor crankcase heater always on.
IMO, the more a heat pump has to defrost shortens it's lifespan.
And running the boilers at lower outdoor temps puts them in the right window of operation.

You have a separate boiler for each 500' apartment? What size are they?
It is easy to over cool 500 feet with 2 inside heads, even though the AC is modulated....I would go with the smallest heads available. Is there anything smaller than 9000?

Jells

Having to keep the baseboard system seriously reduces the charm of the minisplit for me! I didn't want to get into the heating system because it always gets certain people riled, it's circulating water from the WH using a bronze pump, a perfectly legal way to heat here. I do not have any way to install a condensing unit of any kind. So the appeal of the Minisplit is instead of upgrading the current system with a heat exchanger block, I do away with it entirely, and get the benefit of ripping out the ugly and space consuming baseboard heaters.

TAG

The Mitsubishi hyper's work very well. I have a three head and a Carrier 5 speed doing my home and office in NJ. The house has a radiant system -- but when I was doing work on the house and system the HPs had no problem keeping this warm.

It's all a question of getting the correct size and specification .. even with 3+ COP .. NG is still cheaper vs .20 electric where I am.

You have to do the math ....

I'm doing another carrier and minis in my new build in PA. --

GW

Ductless and tenants: maybe not the best move.

Ac—- maybe they run for a few months.

Heating and cooling——that’s a lot of run time and the blower wheels may get gooped. If the tenants are super clean then maybe it’s fine.

I just can’t imagine using ductless for heat when you have nat gas.

Jells

GW said:

I just can’t imagine using ductless for heat when you have nat gas.

Well, neither could I except that everyone keeps raving about how efficient they are! Like I said, hard to get real numbers. Maintenance is a good issue to raise, and another reason I am not willing to bend over backwards to install condensing heating units. I've heard so many maintenance stories, vs the minimal maintenance of a hot water tank.

Zman

Plug in your local rates and see what you come up with. The boiler is probably 80%, be careful with the COP for the mini, they drop off quickly at lower temps.

I ran the numbers for a buddy of mine outside Ashville NC a while back. With the cheap electricity and moderate climate, the was a winner. My guess is that you power is a bit expensive and weather a little cold in NJ.

Jells

Zman said:

Plug in your local rates and see what you come up with. The boiler is probably 80%, be careful with the COP for the mini, they drop off quickly at lower temps.

Cool. Stupid question: When doing these comparisons, I've always added my supply and delivery charges to get my "real cost" even though it is never seen as such on the bill. Correct?

Kjmass1

Jells said:

Zman said:

Plug in your local rates and see what you come up with. The boiler is probably 80%, be careful with the COP for the mini, they drop off quickly at lower temps.

Cool. Stupid question: When doing these comparisons, I've always added my supply and delivery charges to get my "real cost" even though it is never seen as such on the bill. Correct?

Absolutely. You have to pay regardless so include it all.

500sf can definitely be served by 1 minisplit based on load, especially if these units are stacked or side by side, their true heatloss could be very small. Definitely avoid a multihead as the minimum modulation will be way too high for a unit this sized. You'll probably want to stick with Mitsubishi as they can get down to like 2,000btu/hr.

My advice- make a post over at greenbuildingadvisor and they'll get you sorted out on the minisplit front. You'll want to get a Manual J done and it will pay for itself.

Zman

I would do it that way. Just take the total bill and divide by quantity.

Jells

> @Kjmass1 said:
Definitely avoid a multihead as the minimum modulation will be way too high for a unit this sized.

Well that takes it off the table, since one head would not work with this long 3 room apartment. It's a 11x13 room on either side of a 8x9 plus a 7' hallway. The doorways would not distribute the air well.

Jells

> @Jells said:
> > @Kjmass1 said:
> Definitely avoid a multihead as the minimum modulation will be way too high for a unit this sized.
>
> Well that takes it off the table, since one head would not work with this long 3 room apartment. It's a 11x13 room on either side of a 8x9 plus a 7' hallway. The doorways would not distribute the air well. Window units FTW.

Kjmass1

> @Jells said:
> > @Kjmass1 said:
> Definitely avoid a multihead as the minimum modulation will be way too high for a unit this sized.
>
> Well that takes it off the table, since one head would not work with this long 3 room apartment. It's a 11x13 room on either side of a 8x9 plus a 7' hallway. The doorways would not distribute the air well.

> @Jells said:
> > @Kjmass1 said:
> Definitely avoid a multihead as the minimum modulation will be way too high for a unit this sized.
>
> Well that takes it off the table, since one head would not work with this long 3 room apartment. It's a 11x13 room on either side of a 8x9 plus a 7' hallway. The doorways would not distribute the air well.

You could look in to a miniducted unit with the airhandler in the middle to serve the adjoining rooms.

John Mills_5

The low temp models like the mentioned Hyper Heat heat great and efficiently well below zero. Do the math with your local rates and see which way is best. Nice to keep the boiler heat for backup if someone's unit goes down. Got a place for a conventional air handler? Be cheaper for a little 1.5 ton cooling only conventional split and stay with boiler heat.

Jells

John Mills_5 said:

The low temp models like the mentioned Hyper Heat heat great and efficiently well below zero. Do the math with your local rates and see which way is best. Nice to keep the boiler heat for backup if someone's unit goes down. Got a place for a conventional air handler? Be cheaper for a little 1.5 ton cooling only conventional split and stay with boiler heat.

John, this is a tiny 500 sq ft apartment! The little closet for the 40 gal tank is a significant loss of space. If it can't provide heat I'm happy to let the tenants keep using 4k btu window units.

Kjmass1

Are these units stacked or side by side? Have you run a heatloss analysis?

Jells

Kjmass1 said:

Are these units stacked or side by side? Have you run a heatloss analysis?

Both. No analysis, but it's an old un-insulated brick building. Most of the units are adequately heated by 130-40 deg water in less than 30' of baseboard. I once knew the btus for that, but my brain has jettisoned it. One top floor unit with 3 sides exposed needed more emitter so I added a couple of kickspace heaters in surface mount cases.

I'm tempted to get rid of the baseboard in other units and just go with those kickspace units. Much cleaner designwise. Baseboard always ages badly and looks nasty.

Zman

https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&ved=2ahUKEwip6fWnkKHhAhWOwMQHHQ7aA8cQFjAAegQIARAC&url=https://www.slantfin.com/images/stories/Technical-Literature/ratings_fineline30_r.pdf&usg=AOvVaw2Ucgrd_l6_b6srqShphSOP

mikeg2015

John Mills_5 said:

The low temp models like the mentioned Hyper Heat heat great and efficiently well below zero. Do the math with your local rates and see which way is best. Nice to keep the boiler heat for backup if someone's unit goes down. Got a place for a conventional air handler? Be cheaper for a little 1.5 ton cooling only conventional split and stay with boiler heat.

The hyper heat models rely on heat strips in the pan to allow effective defrost at very low temps. NET COP if outdoor unit is fully loaded will probably be close to 1.0.

You won’t find good data on defrost power consumption or the pan heaters.

Mini splits biggest advantage is zoning, part load efficiency and no leaky ductwork.

kenjohnson

I just did the math on a Mitsubushi HyperHeat unit compared to gas (my old Monitor gas space heater failed and is being replaced with a mini-split). The COP of the HH mini split is something like 3.4 at 20-30 degrees (output 22,000 BTUs, input 1.8 kW, or ~6,000 BTUs) and 2.0 at 5 degrees (output also rated 22,000 BTUs, input not stated).

So let's assume for a moment that the mini-split input is half the output at 5 degrees F, which is what COP of 2.0 would mean. That would mean ~3 kW input. To make the COP be 1.0, the pan heater would also have to consume 3 kW. I find it hard to believe that the pan heater would consume 3 kW constantly to make the COP go from 2 to 1. I could believe that the pan heater was 3 kW rated but more than likely it runs on a 10 or 20% duty cycle, so I can't see it making the COP equal 1.0 - perhaps 1.8 or so.

Since the unit requires 240V service, 20A would be 5kW total available supply. I could see that being necessary for very short periods of time (defrosting).

If you have some data from a technical manual, or measurement system, I'd love to see it.

It has also been my experience with a Fujitsu RLS2H (similar to HyperHeat) that the COPs are still well above 1 at 0 degrees F.

Curious to know what data others may have.

Larry Weingarten

Hello, Just to muddy the waters, have you considered PTAC units? Throw in some jump ducts or other ways of allowing air flow, (like Tamarack makes) and you might have more choices.

Yours, Larry

mikeg2015

kenjohnson said:

I just did the math on a Mitsubushi HyperHeat unit compared to gas (my old Monitor gas space heater failed and is being replaced with a mini-split). The COP of the HH mini split is something like 3.4 at 20-30 degrees (output 22,000 BTUs, input 1.8 kW, or ~6,000 BTUs) and 2.0 at 5 degrees (output also rated 22,000 BTUs, input not stated).

So let's assume for a moment that the mini-split input is half the output at 5 degrees F, which is what COP of 2.0 would mean. That would mean ~3 kW input. To make the COP be 1.0, the pan heater would also have to consume 3 kW. I find it hard to believe that the pan heater would consume 3 kW constantly to make the COP go from 2 to 1. I could believe that the pan heater was 3 kW rated but more than likely it runs on a 10 or 20% duty cycle, so I can't see it making the COP equal 1.0 - perhaps 1.8 or so.

Since the unit requires 240V service, 20A would be 5kW total available supply. I could see that being necessary for very short periods of time (defrosting).

If you have some data from a technical manual, or measurement system, I'd love to see it.

It has also been my experience with a Fujitsu RLS2H (similar to HyperHeat) that the COPs are still well above 1 at 0 degrees F.

Curious to know what data others may have.

You’d have ot install a power meter to get reliable data, and they you’d need ot measure return and supply temp and airflow to know for sure.

You can’t use the breaker size. Breakers just protect wire. You need to look at the FLA (full load amps) and MCA (Minimum circuit ampacity) to see what the most it could draw. Inverter units are special, in that I believe they simply overspeed the compressor in low load conditions (cold weather) to increase total lift but not exceed FLA of the compressor motor. This is not nessesarily effecting, but allows more output.

You have to understand that refrigeration capacity depends on total compressor lift (difference is pressure on suction and discharge).

So consider and unloading 2 stage scroll. (Hot gas bypass). It reduces lift and net volumetric output by bypassing vapor at the discharge back to a midpoint in the compressor scrolls. A variable speed unit maintains lift but reduces volumetric output but slowing down. But here the thing, a 2 ton 2 stage scroll on 1st stage might out 24000 BTU nominal @ 95F, but at 75F, 1st stage might output 23,000BTUs. But draws a lot less current unloaded and at those temperatures.

SO back to minisplit performance at low temps. They overspeed the compressor and they have pan heaters to ensure the ice melts off the coils at very cold conditions. The pan heaters waste energy, and the compressor at that speed probably has a COP of 2.0. Add in the energy for defrost, and depending on load you could be at 1.5 or less. It simply gets hard to remove ice from and uninsulated unit when it’s below 10F outside. Ideally, the units might have insulated cabinets and powered shutters and 2 refrigerant circuits and condenser fans.

I think one company did something similar to this with two compressors in tandem for a high performance cold climate heat pump. I think it used a similar concept to an inverter overspending the compressor once discharge pressures dropped low enough a second compressor boosted it up further. Hallowell Acadia Heat pumps. They went out of business.


As for hte pan heater, it’s probably like self regulating heat trace. it consumes as much power as needed to say at maybe 50F year round. IN very cold weather it might in fact run almost continuously.

I think there’s a lot of things that could be done to increase low temp net COP and net output but would require a lot of complexity and cost.

Ultimately I think hydronics may be a better solution for overall efficiency with heat pumps with an energy storage system. Homes simply need more mass, then you need less peak capacity and can ride through colder weather with less defrost demand. IF you could back off capacity a little, you might be able to keep the coil temp above repoint and reduce frost. Same way a indoor coil in cooling can remain dry if you increase airflow and increase coil size.