Solar or wind powered aircon

[Downunder35m]

I visited a friend of mine who still loves living very rural and self sufficient despite being well past the 80 year mark now.
Grows his own potatoes, veggies and all, including have chickens and sheep of course.
Anyway...
Getting a bit older now means his comfort needs during the summer are increasing, especially on these very humid days.
His idea now was to ask me if I could make him an aircon that he can run on solar and / or wind power alone....

It is only for a small bedroom, so no huge requirements but NOT having mains power available puts a big dint into things.
After lots of talks and a few drinks we settled on the idea of abusing a car aircon salvaged as whole from the wreckers.
The idea shall be to run the compressor mainly using an old wind turbine that was used to pump bore water up back in the day.
As an additional support a little motor from an electric mobility scooter or such might come in handy.

My problem is that I seem too be too old or too confused to actually calculate or at least guesstimate if any of this actually feasible.
I know that a car aircon only works properly above the 120 RMP region but I don#t want to build something up that in the end won't even make the compressor turn.
Of course it would need a pulley with more mass to overcome the initial hesitation to run, especially since stalling means more problems along the way.

How can I figure out what sort of power is required to ensure proper operation?
And would a little scooter motor even be strong enough?
They come with about 500W and that alone seems quite on the limit already...

[liquidhandwash]

If your willing to spend a little money I would suggest buying some secondhand panels and a solar VFD, you can then drive a 3-phase motor and your aircon pump. I've seen a great manufactory of VFDs but it is bookmarked on my other computer that's stuck in a boot loop and now I can't find it. They have them on amazon looks like they also have single-phase VFD. The best part no battery is required.
I've played around with wind turbines, there are 3 rules before you even start. Location location location! And also they are high maintenance and a huge pain in the arse to service.

[liquidhandwash]

That's weird the link I copy-pasted seems to disappear, is it blocked?

www.amazon.com/Inverter-Variable-freque ... 6YMYV?th=1

[Downunder35m]

A friend of mine currently has to rebuild the front end of his car...
His loss, my gain.
We disconnected the belt and then I used my trusted Makita to drive the flywheel at full speed.
Needless to say that nothing happened as the car system demanded the fan to run and some other bogus.
And so the clutch got hard wired with a little switch.
Let's just say the Makita won't mind dealing with big bolts but did not have the power to get the compressor going.
It turned to slow obviously but also drained the battery so quickly that the protection kicked in - oops...

A compressor that is not from a big V8 car might be better but it still seems like an awful lot for a mid sized wind mill.
Back to the drawing board....

A quick check on what sort of BTU values WOULD be required based on room size and expected temperatures I started looking for systems in that range.
And if I shall trust what our energy saving and so on websites suggest I would need at least 2.5kW for the cooling power.
AU standards are a bit weird, so for those from those unknown parts of the globe Aussies have not discovered yet:
Those 2.5kW are for the cooling, not for the power the compressor and fans consume.
The later comes at in about 1000W.
I converted a rather old portable split system I got for cheap (out of gas) to R600 a few years back.
It was rated for 2200W or just over 9A with the original refrigerant.
Running on R600 after extending the thin metering tube by about 25cm resulted in a power consumption of just over 800W under full load on a 32° day.
Mind you though that this oldie also uses a quite oversize rotary compressor inside for a smaller footprint.
I just wish they would have placed it in the outdoor unit LOL

Anyway:
After checking a few small aircon systems at the local hardware store today I found quite a few systems using less than 300g of refrigerant, be it R134a, R32 or for one even R290 - with just 120g of refrigerant inside.
Funny enough this R290 one was not only the cheapest in the range between 2 and 2.6kW but also had the lowest power consumption.
Was a 2.4kW system using LESS power from the plug than a 2000W system on R134a....
And THAT got me thinking.....

Imagine you could get a DC compressor that you adjust in speed through a certain RPM range, let's say from 2500 for low to 3000 or 3500RPM for full speed.
With that you could just go totally over board on the budget and use an Arduino or such to control things.
But what exactly would be the best approach here in terms of getting the best efficiency AND cooling?
And so I wasted a few hours going through this thing we call the world wide web.....

[Downunder35m]

There is a few ways of going quick and dirty with a budget aircon.
Like using a TX valve meant for commercial sized systems that comes with a not too big sensing bulb.
But doing so would mean that we struggle with our speed control.
It totally upsets the TX valves if you change the flow rate and pressures....
A meter way is to use a TX valves from a car.
Those check the temperature in the return line from the evaporator to regulate the flow.
Problem is that this can be very stubborn when it comes to abusing them with R600 due to the lower temps created.
Last but not least would be the Danfoss way and going with an externally controlled TX valve - electric to be precise.
Sadly those ARE quite costly and so not really and option here.
How can fix something that is not broken and that we did create yet?

On the way home from the hardware store I stopped at scrap yard to check if there is anything to give me a great lightbulb moment.
And I got a few of them
On a very old and very rusty wall unit I noticed that it was a weird R22 system.
For starters there was very thick tube mounted horizontally at the outlet side of the evaporator and a similar one on the inlet of the compressor but mounted vertically - not a dryer, this was inline as well.
Inlet and outlet connections on these were "on the top", leaving most of the tube below the vertical one had both lines going from the top.
The metering was dune with a "bulb" to get the low pressure line to a larger diameter followed by about 1.2m of coiled up capillary tubing - three of them, with each going into a separate coil of the evaporator.
Only logical conclusion for all those gimmicks is that they allowed for quite some over filling of the system.
Or if you prefer a massive operating range for the outside temperatures.
Surprisingly though is was not a reverse cycle system.
Those large diameter additions allow for quite some liquid refrigerant to be stored safely and out of harms way for the compressor.
And in return, if you really needed to, you could crank the thing up to full on a 40° day and still get closing to freezing point air coming out.
Too bad it was totally wrecked.

And then there was this ice and water dispenser someone seemed to have dropped from a mountain top...
TINY R290 compressor with (according to the label) only 48 grams of R290 in it.
From what I could gather from the wreck it looked like it produce two ice cubes at a time LOL
The ice maker was a piece of copper pipe with bumps going into the ice cube moulds.
The "evaporator" was a copper line wrapped around a large cup sized piece of stainless - the cool what comes out of the cold water line.
So it went through the ice maker then around the "cup" to end in long wire frame lines acting as the condenser.
In hindsight I think the owner was not impressed with the performance and had a non verbal conversation with the thing before dumping it off...

So how about we would combine dirt cheap with very old fashion and a refrigerant never intended to be use in such a contraption?
A relatively small evaporator with slightly oversized tubing in terms of diameter and length.
A nicely sized evaporator from an old wall unit as well.
Then some salvaged liquid collectors from the scrap yard.
Ideally though we would opt for some portable unit from the scrap yard in the range of 3.5 to 4.5kW for the cooling capacity.
For the power consumption on the label of the cooling power is not specified it means between about 4.8 and 7.2kW.
We salvage what we can and need from the inside and housing for later use.

General idea (on paper) here:
What such a big portable unit provides for the evaporator and condenser is about ideal for a nice R600 conversion.
In many case we can even be lucky and find one that already has multiple feed lines going into the evaporator, with some extra luck for the condenser as well.
Size wise something from a small car would be ideal but I really wouldn't want to try to join those Aluminium lines with copper ones.
With the outdoor unit being rather low on the ground and the inside unit high on the wall we have the added benefit of being able to use the same oversized piping the original system used for what we need to connect inside and outdoor unit.
The added liquid storage for the condenser means we won't end up with liquid refrigerant getting into the compressor.
The added collector on the outlet side of the evaporator allows us to get away with quite a bit of overfill in the condenser.

Now for the fun part with the micro controller....
With such a system we do have no issues with the filling as a bit more won't hurt.
And even if outside temps go extreme we would have enough volume available to keep the temps down.
What we don't have without a proper TX valve is control over the subcooling and superheat.
That's where the Arduino comes in.
Ideally we would to keep both in the range of 10 to 15 degrees without the evaporator going too low below the dew point.
How then could be accomplish that while also making sure we won't mess up our pressures?

That's the beauty here - those temps and the pressure directly relate to each other.
Some bogus tutorials tell you that your system is properly charged if the equalised pressure of the system is identical with what the PT charts reads for the current ambient temp.....
Think about what damage this could do....
Why?
Because the gas turns to liquid based on the temp AND pressure.
As a liquid it needs far less volume.
The only thing you can confirm this way is what type of refrigerant is in there but NOT how much....
So if we know the ambient temps for inside and outside we can determine what pressures we SHOULD have in the evaporator and condenser.
By using our subcooling and superheat we can calculate the required pressure for those ambient temps.
And guess what?
That PT chart also tells us what temps we should have for a certain pressure.
Of course there is some losses her so we have to add between 2 and 4% to get really accurate....
So far so good but that means we need quite a few temp sensors - how many ???

One each for the ambient temps of course.
The ideally two on either side of the evaporator and condenser, making 6 already.
Of course the last thing we need on a hot day is our compressor cramping out due to overheating.
Unlikely with our excessive use of refrigerant but if not taken care VERY costly.
And if we are honest for the compressor we would prefer two sensors as well.
One to check hot hot the housing gets the other to ensure our inlet line is not starting to freeze over - indicating that shortly after we face liquid refrigerant destroying or compressor.
Damn, that's a lot a PID programming and a quite complex matrix to combine all values.
All this just turn the compressor on and off and if need be increase the RPM'S ??? Hmmm....
Let's recap here....

If we start flooding our evaporator it starts to ice up.
And if we starve it we end up with a lot of ice right on the entry points - because our pressure is way too low.
The first means the outlet side will quickly go below the 4°C range, while the later means we won't reach our goal for the superheat.
Means we can get away with just one temp sensor place shortly after the capillary tube - or one of them anyway.
On the other side of the system we want to be about 10 - 15° above ambient temp.
Remember that it is heat pump so we can only ever remove heat.
How does that help us here?

If we fail to deliver not enough refrigerant our temp sensor for the evaporator AND the condenser will will indicate BELOW optimum values.
If we deliver too much refrigerant the temp sensor for the evaporator will first go UP - due to the liquid refrigerant not evaporating into the line as a mix of liquid and vapour.
Then however it will quickly start to drop with the coils to start icing over.
On the high side we will see how the ideal temp keeps dropping fast while the refrigerant boils off in the line going to the condenser.
We optimise our capillary tubing for the metering either by just length if you want to go all the way or as a small diameter coil(s) right on the inlet side of the condenser.
The later requires far less capillary tubing while the first means standard tubing for the connections.
Not sure which is preferred yet ...
And the ideal metering of course shall be done for the lowest speed of the compressor....

Where are we with this then at this point?
We should have a system that provides optimum performance on a day just hot enough to turn the damn thing on.
On a really hot day though not even our embedded liquid storage will help us to get anywhere near the dew point for the evaporator.
But didn't I just say we optimised our system ???
Yes, but on and for a "cold" day
As we use oversized cores it means our fans now pump too much hot air through the evaporator to provide proper cooling.
The cores simply transfer more heat energy from our room or away from our condenser simply because the amount of refrigerant he have in use is not enough.
And so we crank up the speed for the compressor.
We know when we have to do this based on the ambient temps and our subcooling and superheat going bonkus.
Means we only now crank up the speed to actually give the system the flow rate required to deliver enough liquid / gas mix to the evaporator for it cool down the entire way.
We can just keep it coming until our inside temp has dropped enough or our temp sensor indicates we are at risk of freezing.
Throttle back for a while until the temps are fine again then ramp it up a bit less and repeat until the ideal speed is reached.

What about the fans ?
Glad you asked
Since we already control the motor speed through the micro controller it of course is only logical to include the fans here.
And in reality we have no other option than to use a good PID based PWM control for the fan speed.
If you made it till here, you might as well read on for the complexity of our PID control and why exactly we go this route - in theory so far.

[Downunder35m]

There is several ways of influencing our subcooling and superheat.
To get a little grasp I will try to give a few examples.

If what goes into our evaporator is too hot (assuming a properly filled system) then it means our condenser struggles to provide enough of a temp difference.
By increasing the fan speed if what comes out of our condenser is too hot we fix our superheat until it matches.
Since all this also affects the other side of the system we can just use the outside fan speed to control our system until it reaches the limit on low speed.
This might be the preferred way here as the others mean we either need to adjust the metering or the inside fan speed.
As we check and control all temps anyway this leaves us with fan speeds based on ambient temps in relation to the temp difference the user sets for the inside.
Basically what an inverter based system would do these days if you go full size.

The problem with our PID and PWM control for fans and compressor is that in a HVAC system nothing changes instantly.
Means we first to check and determine how long it takes for the cores heat up / cool down.
Then we need a table with the relations of inside and outside fan speeds based on the ambient temps and user settings.
Last but not least the speed control for the motor to crank things up if needed....

So what DC compressor can we get then ????

[Downunder35m]

The compressor is still the major problem here...
For a 12 or 24V DC system we get them for R134a, R290 AND R260.
While we can use a R134a and also a R290 compressor they come with a downside.
Both are meant for rather high pressures and have a quite small displacement.
Considering these 12V models all run at the same speeds and that we want to use R600 anyway....

You see what just happened ? Sense
We use an oversized R134A or if lucky R290 evaporator and condenser.
And now use a compressor pumping significantly more gas per stroke through the system then what it was designed for.
At least that was until I realised how damn hard it is to find a decent sized R600 DC compressor....
Even a 2kW unit already comes with an awful lot in terms of BTU.
The biggest DC compressor I could find for R600 though only provides around 400 BTU.....
Keep in mind though that this based on running in a optimised R600 system.
But even for or modded R134 system it simply means our compressor won't be able to handle the volumes required for those cores.
Quite a shame actually if you consider the power requirements for the R290 and R134a compressors.
They wouldn't make totally different controllers, just different firmwares maybe.
So why not provide a R600 compressor maxing out what those motors are able to provide with a few high displacement options?
The big one I found is 60 model but uses far less power than a R290 or R134a 35 model.
I fail to see why they could not make one with around 120cc instead of the 60 I found to max things out.
With that we get in the ballpark required for a small room.....
Until then though.....

Ok, but what sort of power requirements are we talking here for the 60cc R600 compressor?
104W.....
That is at full speed and full system pressure.
On a good day and low speed only around 48W.
So going in the 200W range here is still very feasible if you ask me....
400W if we can go for a 24V solar system instead.

The displacement and control we need seems only be possible with a variable displacement compressor from a car and adding an electric motor to it.
And ideally we would like something in the 150cm³ range.
Things that from a few Toyota models no one seems to use them and sticks with fixed displacement models that require a clutch.
Starting to wonder why so far no one came up with a 12 or 24V DC aircon of proper capacity....
Same for why we use inverters to control compressor speeds and not variable displacement compressors outside certain applications.....

[liquidhandwash]

I had a thought, a big hot water service the bigger the better, filled with coolant, circulated by a pump, to a 12-24 volt fridge/freezer also filled with the coolant powered by solar. The hot water service becomes your cold battery. in the weeks before the heat wave you run the freezer and pump dropping the temperature of the coolant as much as possible, then on hot days run the coolant through a radiator and fan. I've got a 12 /24-volt fridge freezer that gets down to -20. might take extra insulation and a few months to drop 400 litres of coolant down to that temp.

[liquidhandwash]

And then in the winter, you could use the solar panels and the electric element on the hot water service to heat the coolant, and you now have a heat battery..

[Downunder35m]

I like the water battery idea.....
But last time I checked for something smaller (my still) it turned out that the tank is a relative thing one you keep circulating things.
200 liters chilled down to just 4 degrees sounded great but after a few runs I reverted back to letting the water run through 3 aircon condensers again.
In order to get a half decent cooling effect you need to have the core sitting at under 10 degrees.
Even with a 500 liter tanks chilled down to 0: How long until the return flow heat the tank to a point where the cooling becomes a thing of the past on a 40 degree day ?

But I have a nice co-worker that still visits the family in the Philippines twice a year....
She showed me something interesting on her phone that a local store uses.
They run their little walk in freezer and cool house on water - literally.....
She promised to take some more pics next time
Either way:
They have a 200 meter long pipe from upstream going to a water wheel and evaporative, fan forced cooler box that houses the condenser from an old refrigeration truck.
The fan runs on solar and batteries.
The evaporator sits in the walk in freezer and again all is power by 24 from solar and batteries.
So far no big deal really....
But they use the water wheel with a pulley to drive a CUSTOM build compressor.
Claimed to have been salvaged mostly from a commercial unit that had a mechanical oil pump added as the drive shaft was extended to hold the pulley.
Imagine a cut in half drum with this contraption mounted inside.
At the bottom of the drum the oil pump that just provides a steady flow to drown everything.
I know what you gonna say now but they fixed this by adding a pressure gauge and 9kg propane tank for camping use.....
Yes, every few days they open the valve to replace the refrigerant that escaped through the non existing compressor housing LOL
I suggested to at least use some high pressure oil ring seals or such but even after showing her pics she did not understand what I mean.
She is a great personal carer but has really no clue about anything mechanical or technical...

Sadly I neither have the 3m water wheel they have, nor a creek or such to power it.
Otherwise I would try a compressor from a car or truck.
There is some quite nice projects that people tried or succeeded with.
My landlord finally agreed that it is time to at least fix some of the issues outside.
I am hoping to get the OK to put two 200W panels for 12V / 24V output on the garage.
With a little tracker to keep them in the sun of course.
Assuming that at least in the morning and on less hot days I can use some of the juice to charge batteries I guess I could afford about 800-1000W for a few hours, depending on the battery charge.
But let's go with a more cautious approach and say I end up with about 600W I can use for 8 hours.
4800Wh isn't much with a conventional heat pump system that uses a compressor.
Having said that....

There is quite a few ways how we can cheat with a compressor based system.
I experimented a bit with a salvaged window mount aircon that someone discarded on the nature strip, sadly without the remote....
Propane provides quite cold temps when boiling off.
But in a NORMAL system the goal is to keep the condenser about 10 to 12 degrees above ambient while the evaporator is kept around the dew point for the lowest and around 10 to 15 degrees below the temp of the ingoing air.
Depends on the design and capacity.
Ok, how to cheat here ?

The only things we really need to worry about are the pressures and how much liquid is in the heat exchangers.
Means we opt for a condenser that is big enough for our refrigerant amount to stay in the butter zone of just 10 to 15 degrees above ambient.
(We can improve on the temp with water sprayed onto the core that we get from a tank)
The trick is to fully abuse the evaporator part by using a heat exchanger.
Like for your water cooled PC
A nice block with coolant and refrigerant lines inside.
The control is done by a motor driven TX valve that gets it's position based on the OUTLET temp of the refrigerant line.
An added expansion vessel, like by using a "pot" in the line to the compressor takes care of any too large fluctuations in case something get out of parameters.

In such a system we can really go overboard in terms letting the refrigerant go way below and temps and above any flow rates we would allow in a normal system.
All a matter of design to ensure there is enough volume to expand into while all the liquid boils off before leaving the core.
Minus 40 is no big problem and a good coolant can take this with ease.
Some old chest freezer can be sealed off as a coolant tank.
Only problem is that a big tank and lines running to some old radiator in the house is not really doable when you are renting....
And as a direct system without a tank the controls are a nightmare....
The coolant flow and temp needs to match what the compressor and heat exchanger provide.
A high pressure cut off switch needs to be added and also a bunch of temp sensors on top what is already there.
I guess there is a reason after all why the future of cooling is passive and not active....
Wondering when someone will restart the ammonia based way in "reverse" with passive cooling....