Solar Mini Split or ... Solar and a Mini Split

[Plum Crazy Rob]

As a test you can run a fossil fuel heater in the garage during the day, then shut it off overnight.
Record how long you ran the heater and what temperature you reached and dropped to overnight.

I was going to suggest the same, good idea to test...

I don't think the EG4 heat pump is of the hyper heat type which can operate efficiently at low ambient temps. If it's not then it's not optimized for winter use at sub freezing conditions which is that you're trying to do.

Correct, not by design, so it doesn't have a pan or compressor sump heater so shouldn't be used in a continuously super cold climate where it could turn into a block of ice or damage the compressor. However, I have used mine down to -3F without issue, and it did pass the Energy Star 5F cold climate test. I feel like some of the el cheapo mini splits have generated a lot of "too many defrost cycle" complaints.

 

[Plum Crazy Rob]

his writeup is great real world experience and the wattages noted are great reference material!

Sure. I think a 3rd 450 Watt panel would have been nice.

The most I have ever seen my 24k draw is 3000W, with Grid assist (on Heat). 2700W Solar, which is my array size. 2200W in Air Conditioning mode, that's probably ball park for what my array puts out on a hot 100-110F summer day here.

 

[PVGeezer]

It's a really weird world where PV modules cost less per square foot than dual pane glass windows. But here we are. PV is about half the price per unit area of windows. It's only about 1/4 as efficient so it's got another factor of 2 or so to go but getting closer. We're also getting close to the point where cheap direct resistance heating with PV beats using a heat pump.

 

[Madcodger]

We're also getting close to the point where cheap direct resistance heating with PV beats using a heat pump.

Umm... What's the math behind that? Because I think you'll likely find an error in it.

 

[PVGeezer]

Umm... What's the math behind that? Because I think you'll likely find an error in it.

Well since you're challenging me you should really present the math showing I'm wrong but I'll bite.

Base assumption here is that were doing sunny day heating only. That of course gives the direct DC system an advantage but it's also the premise of this thread. I'm also assuming I'm doing it here in my climate using my cost data.

I just put in a 32Kw PV system and 54k btu/hr of hyper heat heat pumps this year. All in materials cost for the heat pumps was $8500. So that's $8500/(54000/3.412)= $0.54 per thermal output watt.

PV modules are at around $0.25 per STC watt so around $0.29 per NOCT watt. My hybrid inverters and ac switchgear came in around $0.50 per watt but I'll be generous and cut that in half assuming straight grid tie inverters. My array structure and DC BOS came in at $0.28/watt so combined cost (PV + inverter+BOS) /inverter efficiency is around $0.82/.95= $0.86 per ac electrical watt. COP runs about 2.5 in my climate so that reduces thePV+inverter+BOS cost to $0.86/2.5= $0.34 per thermal output watt. Adding the heat pump the combined cost is about $0.88 per delivered thermal watt. Plus I still have to use the grid for load matching.

If I used direct DC resistance heating instead I could connect to simple baseboard heaters, which dont care if they get DC or AC, so no inverter or inverter efficiency loss. But I will have impedance mismatch loss between the resistive load and the PV array IV curve. I've run those kind of numbers before and they typically work out to around 30% mismatch loss. The baseboard heaters cost about $0.07 per thermal watt and of course have a COP of 1.0. So no COP based cost reduction for the PV or DC BOS. So combined I get $0.07+ ($.29+$.28)/.7 = $0.88 per DC electrical wall or thermal watt.

Such a coincidence. I promise I wasn't trying to make them come out the same.

I can't use the EG4 hybrid heat pumps in my coldish climate but if I did the heat pump cost per thermal watt number would be better and you wouldn't need the inverter so the numbers would swing the other way. Thats why I said we're getting close to the breakeven point, not quite there yet.

I didn't include any PV or heat pump installation costs. Having recently installed both Id call that a wash. Personally Id rather connect MCs than flare linesets. Also I didn't include the tax credits, which are more favorable for PV than for heat pumps.

And what could be simpler than a pv array connected to a baseboard heater with a mechanical thermostat? Nothing to break or maintain.

Anyhow, let me know if you find a math error, I certainly could have made one.

 

[Plum Crazy Rob]

Well since you're challenging me..

Hey, that was quite profound! Lol.

 

[Plum Crazy Rob]

Here's a sample using my bedroom "air battery" and an EG4 12K hybrid.

Hi 48F low 26F - Sunny
Std 2x4 const R13 wall / R30 ceiling, about 450 sq ft.

Heat to 75 during day, drop back to 61 at night:


Uses 0.5 kwh of Grid, see following chart. I have data from another room / another 12k shows add *1 kwh* if you want to maintain 75F overnight with Grid.

 

[Plum Crazy Rob]

And what could be simpler than a pv array connected to a baseboard heater with a mechanical thermostat? Nothing to break or maintain.

Guessing you're well aware of this, but for any others who might come along - baseboard heater switchgear and thermostat would need mods to prevent DC arcing etc.

 

[Solar1]

...and resistance heating, I would assume, would likely last a lot longer than a heat pump?

It's crazy that it's even close...

 

[PVGeezer]

Yes for sure you'd need a DC rater contactor. You'd can drive that with one of those little $5 digital thermostats. And an IMO disconnect. No need for overcurrent protection if you match up a heater to each PV string.

Or you could use a direct DC connected immersion heating element with a water tank to store the heat for nighttime use. A 500 gallon insulated HDPE water tank costs about $2k. HDPE is good to 180F so if you allow a 40 degree temp swing the tank would store about 49kwh. So that's $42/kwh. Compares favorably with DIY LiFePO4 batteries, even if you figure in the heat pump COP. Connect to a pex slab heat system or some wall radiators and youre good to go.

Edit: If you already have an outdoor wood boiler you could connect the solar heated tank(s) into that and have a PV-Wood hybrid system.

 

[PVGeezer]

...and resistance heating, I would assume, would likely last a lot longer than a heat pump?

It's crazy that it's even close...

Oh yeah. If I were doing pv because I was a prepper (I'm not) I wouldnt care if it was more expensive. Our fancy electronics will be dead in 10 years. A DC pv array will be just fine 30 years from now, and you can stock a couple of spare resistor heating elements for a few dollars. Good luck getting replacement parts from China after the zombie apocalypse.

 

[Madcodger]

Well since you're challenging me you should really present the math showing I'm wrong but I'll bite.

Base assumption here is that were doing sunny day heating only.

That's a big part of the error I propose you have - it's an impractical heating system except under ideal conditions. Heat is required when it's cloudy, overcast, and of course at nighttime (which encompasses a majority of the 24-hour "day" during heating season). So, sure, you might be able to purchase hardware (heating elements) that for some limited period of time might be a less expensive way to generate a given number of BTUs compared to a heat pump. But the moment conditions become less than ideal, which is a majority of the time, that system loses out to a system that moves and concentrates heat (i.e., a heat pump) instead of having to generate it. Heat is present in the air even under terrible conditions for solar, so accessing it with any method that requires less energy than it takes to make use of it is superior to a method that can at best offer a return of 1:1. Systems that generate heat directly using electricity almost always operate at a coefficiemt of performance (COP) of at best, 1:1. While "100% efficient" sounds great, it pales in comparison to a system that operates at a COP greater than 1:1, which is true for almost all modern heat pumps. It's the same reason that heating hot water with flat plate collectors or evacuated tubes is no longer a major contributor to solar water heating. Once heat pump water heaters entered the mix, concentrating, moving, and storing heat quickly become a more efficient solution, that can work 24/7 if one has a mens of storing the heat / energy.

 

[Solar1]

Well, that settles it! LOL

Actually, I just ordered a non solar 2 ton heat pump. So...now to order the inverters (probably EG4 3000EGH-48s) and put up the solar panels I have on hand. Another big project

Pretty amazing that with the cost of panels we can even consider resistance heating...that would have been unthinkable 10 years ago.

Of course, I'm hoping that I'll be able to shave a few degrees off the temps in the summer as well. And should be able to add another 2 ton unit once everything is in place.

Thanks for all the input and the detailed analysis.

 

[PVGeezer]

That's a big part of the error I propose you have - it's an impractical heating system except under ideal conditions. Heat is required when it's cloudy, overcast, and of course at nighttime (which encompasses a majority of the 24-hour "day" during heating season). So, sure, you might be able to purchase hardware (heating elements) that for some limited period of time might be a less expensive way to generate a given number of BTUs compared to a heat pump. But the moment conditions become less than ideal, which is a majority of the time, that system loses out to a system that moves and concentrates heat (i.e., a heat pump) instead of having to generate it. Heat is present in the air even under terrible conditions for solar, so accessing it with any method that requires less energy than it takes to make use of it is superior to a method that can at best offer a return of 1:1. Systems that generate heat directly using electricity almost always operate at a coefficiemt of performance (COP) of at best, 1:1. While "100% efficient" sounds great, it pales in comparison to a system that operates at a COP greater than 1:1, which is true for almost all modern heat pumps. It's the same reason that heating hot water with flat plate collectors or evacuated tubes is no longer a major contributor to solar water heating. Once heat pump water heaters entered the mix, concentrating, moving, and storing heat quickly become a more efficient solution, that can work 24/7 if one has a mens of storing the heat / energy.

So, having asked me to present my analysis, did you even bother to read it? It doesn't appear that you did.

Or perhaps you didn't understand my calcs and are just covering up your ignorance by raising the obvious point that most people want heat at night and not just when the sun shines?

Sunny day heating is specifically the use case in this particular thread. I stated as much at the beginning of my analysis. And everyone knows that heat pumps move heat and have a higher COP than resistance heat does. You did notice that I assumed a COP of 2.5 for the heat pump in my calcs didn't you?

So those are not a math errors. I'm still waiting for you to identify where the math error is that you seemed to be so certain was there.

There are numerous other use cases besides sunny day heating where it is not a good use of resources to invest in heat pumps vs additional PV. For example, you mention heat pump water heaters. These can be a very poor investment.

In my case, a two person household with low hot water use in a cold climate, I was much better off simply adding PV to run my conventional resisistance water heater than I would have been spending my hard earned dollars on a HPWH. Our water heat use amounts to only about 2 MWh annually using resistance heat. I can readily supply that with under $1000 of PV. I would still need about $300 of PV to supply a HPWH, not to mention the PV required to power the additional load on the home heating system because the HPWH is actively cooling the house the whole time it's running. Also Id have to buy a complex $1500 device to replace a perfectly good, simple and reliable water heater. And Id have to count on replacing that fancy water heater at least once and probably twice during my PV array's lifespan.

So a HPWH would be a really poor investment for me. If OTOH I lived in Florida and had 6 teenage daughters who all took hour long showers every day, my economics would be totally different....

So the takeaway here is, don't listen to the marketing BS from folks trying to see you the newest widget. Instead of rushing to a simple minded one size fits all conclusion, take the time to understand and do the math associated with your specific use case. You'll be glad you did.

 

[Plum Crazy Rob]

I like the way you think @PVGeezer

Would you mind if I posted your PV powered Resistance Heating vs PV powered Heat Pump analysis to this other thread here for some more thoughts on it??

 

[PVGeezer]

Sure go ahead. Let's see if we can get some actual feedback on the math. I've done a lot of this kind of analysis over the years but there's always more to learn. Thats why most of us are here, to share and learn..

 

[Madcodger]

So, having asked me to present my analysis, did you even bother to read it? It doesn't appear that you did.

Or perhaps you didn't understand my calcs and are just covering up your ignorance by raising the obvious point that most people want heat at night and not just when the sun shines?

OK, dude. You do you. I’ve tried to be polite, but you respond like this. Good luck with your approach. Others can do as they wish after considering all options.

 

[Plum Crazy Rob]

I appreciate your input also @Madcodger

I've literally got your BARNPARTMENT saved to my Solar folder lol:


I know all of our heads are exploding a bit after what @PVGeezer wrote and feeling a little silly about our "Heat pump gadgets"

 

[PVGeezer]

OK, dude. You do you. I’ve tried to be polite, but you respond like this. Good luck with your approach. Others can do as they wish after considering all options.

Peace, man.

You asked for the math, stating that I had made an error. I took the time to provide it. Common courtesy then suggests you should take the time to review it and provide feedback. You didn't. You just said you didn't agree with the base assumption of the thread (daytime PV heating).

BTW, if it's because you didnt follow all of it, and I'm not saying it is, note the word "perhaps", that's perfectly ok. Just say so, I'll be happy to explain further.

We all should be here to learn.
This kind of energy cost analysis is something I happen to be pretty good at. There's plenty of other things I'm not so good at and I'm sure there are things I could learn from you. I'm still open to having a constructive discussion.

I do agree with you 100% that everyone can do what they wish after considering all options. It's how to choose among those options that's at the heart of much of the discussion on this forum. My point is just that these decisions are not necessarily obvious and can almost always be improved by an analytical approach.

 

[PVGeezer]

I know all of our heads are exploding a bit after what @PVGeezer wrote and feeling a little silly about our "Heat pump gadgets"

I like my heat pump gadgets too. Like I said, I just two weeks ago installed 54kbtu/hr of "hyper heat" cold climate heat pumps. I got two Blueridge (a Midea brand) triple head units. 6 heads total. I still have to install the lineset covers.

I got them rather than doing something with DC PV resistance heat for exactly the reason Madcodger suggests: they are versatile and will provide heat whenever I ask them to. I might even use them for cooling once or twice in the summer.

I love em. So quiet and they use less than half the energy of the old central unit they're replacing, which was so bad it went on resistance strips below about 35F. That's gone but I still have the strips and air handler installed. I might try to do something else with that stuff, like maybe put in a water to air heat exchanger and an outdoor wood boiler, if I can find one at a reasonable price. Lots of wood around here. Not sure yet.

 

[Madcodger]

OK, I'll jump back in to say that I shouldn't have said "math" error, when I really meant an error in practicality. A person can "prove" many things with mathematical precision, but whether the thing proven is of great use is another story. If you have a great way to store your sun-derived heat that might make it far more practical (large thermal mass of water or stone, perhaps) but you need to add that to the cost. Otherwise, talking about the efficiency of generating heat only under ideal conditions is perhaps a fun and interesting exercise, but to what end? Tell you what - I'll try to remain open minded about the concept, but I do think one needs to design around practicality and not just theory. Open to ides about storing that energy...

 

[PVGeezer]

I think the main point I was trying to make is that a high tech high efficiency high cost system isn't always the best solution, even if you have to store the energy. Take for example the heat pump water heater. In my case it was much cheaper to just add a little more PV to my system while I was designing it than it would have been to install a HPWH. I'm on grid so my energy storage is "free".

Re thermal energy storage I agree that can indeed be problematical. I think it can be competitive in cost with LiFePO4 batteries, which lets face it, are still pretty dang spendy. Here's my math on that:

Cost of 500 gallon insulated HDPE water tank: $2k. HDPE is good to 180F so assume say a 170F max temp and 130 min operating temp for a 40F temp differential for thermal storage.

Thermal storage is 500x8.3x40= 166kbtu=48.6 kwh. Let's assume a heat pump COP of 2.5 so the equivalent electrical storage is 48.6/2.5 = 19.4kwh. Or $103/kwh.

You can buy an EG4 14.3kwh wall mount battery for $3300 or
$231/kwh.

So the water tank wins. That is, it does if you have a place to put it and a way to distribute the stored heat to your living space.

If you have an central HVAC system then you can install a water to air heat exchanger in it as is commonly done in the outdoor wood furnace industry. Those are pretty cheap. And that kind of system would allow you to add the wood furnace so you could have a PV wood hybrid system. That could be a good solution for some folks. I kept my central air handler as I've been thinking about maybe doing this myself.

You'd still have to find a home for the water tanks though. That is a bigger problem than cost. Most folks would need several. My daily winter heat load is typically around 500k so I would probably need 3 or 4. I could put them in the basement except I couldn't get them down there unless I rented an excavator and took out a wall.

I wonder if a flexible bladder tank would work? Id need to get it in place and then cover it with insulation somehow.....

 

[Solar1]

No insulation needed the water bladders would just warm the basement and then the heat could rise to the house

I like the idea of using old school simple tech. The simpler cars of the 90s and 2000s are so much cheaper to run than new ones if you have any sort of repairs involved. Same with this, a cheap resistive heat strip vs a complex heat pump. The first time ...or two someone has to come and fix it and the heat strip starts looking mighty enticing. Not to mention keeping the inverter going, it's a complex system we are putting in place.

I think the take away is that solar production has become SO much cheaper than in the past. Now if we can get the batteries there we are done! Same as with electric cars, they will be great once we get the battery tech figured. Heck, they would have eclipsed gas cars back in the day if they had good batteries.

 

[PVGeezer]

Good point on the tank insulation. Just insulate the basement well, put a couple floor registers in and youre done, let natural convection do it's thing. It might be a problem to use conventional immersion heaters in a bladder tank though.

Next challenge would be to find a simple reliable solution to improve the impedance match between the resistance heaters and the PV array IV curve. This can be done without that but it would help improve the economics. An mppt controller would work if it had a wide enough output voltage window.

There are specialized mppt water pump controllers like that but I think most of them rely on the motor inductance to store energy to stabilize the output voltage. Might be wrong about that, it's probably been 20 years since I took at look at that technology. We should try to find one of the DIY electronic guys here to design one.

Re electric vehicles, there are very few problems with batteries in them. Most of the problems are with all the fancy systems they put on cars these days, gassers and electrics.

How could anyone ever have survived without all those backup cameras, proximity alarms, Bluetooth interfaces, automatic headlight dimmers, etc?

I own two EVs, a Bolt and a Lightning. The batteries are just fine, and they have essentially zero maintenance. They've been great fun to drive. They do have all those electronic gizmos on them but so do all the ICE vehicles. They cost me zero out of pocket to charge, and I don't think I paid any more for them than I would have for equivalent gassers. Id never go back to to ICE myself.

 

[Solar1]

Well, I think you are correct the batteries now are fine...it's just the cost to produce them...but that will change

I have a lot of gassers but can't wait to get an electric!

And yes, it's all the gizmos that cause the problems. Although I have been disappointed to hear about tesla replacing multiple motors on high mileage cars. I mean it's an electric motor we have been making those for a long time and they should last almost forever...at least WAY longer than a gas engine but it appears they aren't there yet. Maybe "they" is just Tesla ... maybe the car makers with more experience will do better...or it will just take a while for the manufacturers to get it together. But really the motor should never need to be replaced.