Electronics > Power/Renewable Energy/EV's

Batteryless and off-grid system for air conditioning

(1/6) > >>

brumbarchris:
Hi,
As I want to stay away from the weight and chemical content and maintenance requirements of batteries, I am considering a battery-less system that will only operate the air conditioning unit. The logic is that it will only have to operate when there is a lot of sun, so no energy storage would be needed. Just the Solar panels, going to a DC_DC converter which would convert their output to a reasonable stable voltage for an inverter. And then the AC unit would be connected to the inverter.

The AC unit I have now draws some 1.5kW. Considering some margin and considering I might want to replace it with a larger unit in the future, let's talk about a 2kW limit.

So the questions are:
1. Has anyone else tried such a setup? Is it viable in your experience?
2. How many solar panels do I need for that? Is it 3kW instant power? Is it 10kW? (each conversion on the PV--to-220V AC string would have its losses, and also the sun will not always be DIRECTLY on top of the panels)
3. Can you suggest some scenarios considering 250W panels? How should they be connected and in combination with what type of DC-DC converter?
4. Ar there any known and affordable inverter manufacturers which provide pure sine-wave? (this is a must, in order not to damage the AC unit)

Of course, I realize detailed calculations will be needed, but I just wanted to lay out the framework for the initial discussion.

Regards,
Cristian

Codebird:
You are going to need some energy storage, because the AC isn't going to be happy about shutting down every time a cloud passes the sun. You shouldn't need much though, as there is no reason to run AC at night - conveniently, the available energy each day should correlate very well with how much you need. It would also reduce the cost, because the AC will not need to run continuously - you will be able to get away with a smaller panel area, as on a moderately warm day you can charge for ten minutes in order to run for five. You might even be able to do it with supercaps. It's an interesting project, and you are going to need some electronics to manage it (ie, a charge controller of some sort), but engineering-wise I see nothing insurmountable.

I'm assuming that grid power is unavailable, as it would probably work out cheaper.

brumbarchris:

--- Quote ---With an Inverter A/C unit could you modify it easily (relatively) to input DC before the inverter?
--- End quote ---

Well, the AC unit already exists, and it is not an inverter based one. And even if it were, my goal would have been not to modify it, I want to maintain "backwards compatibility" with the grid.



--- Quote ---You are going to need some energy storage, because the AC isn't going to be happy about shutting down every time a cloud passes the sun
--- End quote ---

Turning OFF the AC each time a cloud passes is perfectly acceptable. When it's hot here then it is really HOT and we would love some clouds but we do not get them. This does, however translate into a requirement for the DC-to-AC inverter, which would have to have among its features a relatively smart undervoltage protection (i.e not just decrease the amplitude of the output sinewave, or scramble the frequency or something or anything like that; just pure shut-down.

Anyway, that's one of the reasons for which I consider over-sizing of the PV panel area, to get sufficient energy maybe during the occasional cloud during a bright day. And if I really oversize it, meaning the whole roof covered with panels, then I might even consider using the unit for heating during winter. But of course, that is just a secondary objective.

The grid is available, but I do not want to mingle with it a all. Minly because of the potential regulations regarding pushingcharge on the grid.

Regards,
Cristian

fourtytwo42:
I think there are two things you need to investigate more fully here
1/ what will be the shape of the power delivery curve each day considering the location of your panels, this will also show you what peak installed power you will need to maintain 1.5Kw for the number of hours you want and the months you want. There are many websites for this.
2/ Determine the characteristics of your AC, for example is it an old capacitor start motor, then you will have trouble due to the very heavy inrush and no protection against frequent starts. Or is it a more modern inverter type, these have a lower inrush and can sustain frequent starts as they have inbuilt timers etc.
Also not all inverters you might use for supply are the same when it comes to motor start so depending on your AC characteristics be careful what you buy. I am with you on the battery front BUT for this application it depends if the PV costs for the extra peak power required without a battery is cheaper than the alternatives (battery and charge controller).

I just saw your latest reply, it's a capacitor start motor.......well I think the only practical way to avoid batteries is to use the grid to start it (as you have a grid), in other words invest in a GTI. Yes I understand that means investigating the regulatory requirements but that may be time well spent. There are a lot of "plug in" GTI's that are normally acceptable up to 1kW and there's nothing stopping you using multiples :)
The only problem is sometimes there input range is a bit restricted meaning you have to use expensive 12V leisure panels but again a bit of research can pay dividends.

Another thing I forgot to mention is check your electric meter specification on the web and make sure its not going to do anything nasty with export power, some of them treat it as fraud or charge you for it but a quick word with your local company normally gets this sorted out (meter change etc).

Codebird:

--- Quote ---which would have to have among its features a relatively smart undervoltage protection
--- End quote ---

Not that smart. Any hobbyist could put together a circuit for that - it's an LM393, some resistors and transistors, and a bloody big MOSFET or relay. But there is an easier way.

Give up on 'no batteries.' It really complicates things, because it means you can't use a standard charge controller, can't do MPPT, and you can't operate the AC at all under low-sun conditions (as opposed to operating it at reduced capacity), and the start-up surge is going to be a serious problem. Instead you should design a 'minimal battery' system. Use a battery, but make it a really tiny one - just a couple of cheap SLAs is all you need. Yes, they'll eventually wear out after a few years of use, but they are cheap enough that replacing them won't be a problem. You only need enough battery capacity to meet that start-up surge load and give you about five minutes runtime. Go that route and all you need are panels, battery, off-the-shelf charge controller and off-the-shelf inverter. No custom electronics required.

Navigation

[0] Message Index

[#] Next page