Questions around a 192 VDC solar system

[Malvineous]

Hi all,

I hope this isn't too basic for such a well educated community, so my apologies if it's off topic.

I'm planning to relocate to a rural area in the next year or two and the idea of generating my own power really appeals to me.  I don't necessarily want to go off-grid, but I would like to have power available during grid blackouts.

I recently picked up a second-hand 5kVA UPS for $50, and I was thinking I could repurpose it as a small residential inverter (single circuit probably.)  It's a "dual conversion" unit, meaning in normal operation the inverter runs all the time even if mains power is present, so having it run 24/7 on DC only should be no problem.  It also runs off 192 VDC (nominal) which is supplied by 16x 12V 5.5 Ah batteries in series, which I like as there should be less loss than a similarly sized system running on 24 or 48 VDC.

Firstly, is going down the track of a ~200 VDC solar system more trouble than it's worth?  I imagine there aren't too many issues doing DIY work with a 24 V system, but are there any hurdles around a 200 V system?  I don't mind getting a sparky out to handle any grid-tied parts of the system and to check over the thing initially, but I'd like to be able to do things like replace batteries and non-grid-tied inverters on my own.

I am also wondering how I would set up solar cells to get power into it.  I thought maybe trying to run at 192 VDC would be difficult if most equipment is aimed at 12/24 V users, but I have seen a few realistically priced MPPT solar charge controllers that will support a 192 V battery bank and 400+ V worth of panels, so I guess this is what I'm after.  When connecting solar panels in series, is there any maximum rated voltage you have to be aware of?  I am not terribly familiar with how the solar charge controllers work.  It looks like they adjust their voltage to get optimal current into the battery, which makes me wonder how high of a voltage will they put in, and could this damage the inverter?  I am guessing that since it's a UPS designed to charge batteries while the inverter is running then it should be able to handle some higher voltage than the nominal 192 V, but I'm not sure what the upper limit is.

I presume these MPPT controllers are designed to have inverters running off their outputs as well as batteries, so having one or more inverters "stealing" current before it gets to the batteries should be no problem?  What happens when the inverters pull more current than can be delivered (e.g. cloudy day or dawn/dusk) - do the charge controllers just deliver whatever they can and then switch off once they can no longer supply any meaningful current?  What happens if you have a 5 kW load and a 1 kW charge controller?  Will it overload, shut off completely, or will it just deliver 1 kW and the rest of the current comes from the batteries?  Can you connect multiple solar charge controllers to the same DC bus (e.g. from panels on two different buildings) or will they compete?

Lastly, is it possible to have both an 'island' system (that stays powered when the grid is off) as well as a grid tie-in system?  I'm thinking along the lines of multiple inverters running off the same DC system, with one (or more) powering residential circuits and one feeding any excess back into the grid.  If so, how exactly does this work?  Logically to me, the solar controller would raise the voltage as more power becomes available, and the inverters would drag it down again as they use current.  An inverter feeding the grid from the same system would switch on once the DC voltage went above a certain point, and it would essentially cap it there, preventing it from going any higher by feeding more and more power back to the grid (until it reached maximum load of course.)  Is this simplistic understanding roughly accurate?

Sorry for the long post and all the questions but any info would be much appreciated!

[Seekonk]

It is probably foolish to design a system around a $50 piece of equipment you just happen to have. Buying 16 batteries is the biggest mistake you can make. Batteries cost about $0.14 per KWH used and only last a few years.  A small battery pack and small sine wave inverter might be a practical way to go.  Connecting them with some grid tie inverters would provide extra power in daylight hours. Heating water is the best thing you can do with a small system in low use periods when no emergency exists.

I have a camp I run on 2KW of panels and just a small 12V battery.  All my hot water is from excess solar power.  I live pretty good for being off grid for months, even have a dishwasher and clothes washer.  Just can't use everything at the same time.  To make solar work you have to pretty talented building your own stuff or spend a whole lot of money that is pretty much wasted.  Go to solarpaneltalk and they will really burn you on your idea. Black boxing it is pretty inefficient.

[fourtytwo42]

Lots of questions there Regarding string voltage I believe the specification limit is 1000V but most commercial grid ties don't work above 6-700V and down to maybe 150V.
Your proposed batteries will store at most 1Kwh and in my opinion that will go nowhere in an outage, for example my oil boiler and water pump for the heating consume ~550W together!
A guy from the Phills on this site had a big 48V system that could keep him going for days but that was some serious battery storage
As someone else said investing lots in this UPS may be a bad idea as if it ever breaks can you repair or replace it ? OTH the leisure systems based on 12/24V are really expensive.
Had you thought of using a standard grid tie (non-Chinese) to begin with just to get you started, batteries add complexity and have short lives unless rated for deep cycle operation. After all a few gas/oil lamps and a woodburner are a lot cheaper and simpler IMOP.
Regarding externally charging your UPS batteries this may be extremely dangerous if it is a transformer-less design as the batteries will be at grid potential!

[David Hess]

Firstly, is going down the track of a ~200 VDC solar system more trouble than it's worth?  I imagine there aren't too many issues doing DIY work with a 24 V system, but are there any hurdles around a 200 V system?  I don't mind getting a sparky out to handle any grid-tied parts of the system and to check over the thing initially, but I'd like to be able to do things like replace batteries and non-grid-tied inverters on my own.

High voltage DC has additional hazards not present with high voltage AC so just for safety reasons, I would preferentially use 48 volts.

Solar charge controllers have minimum and maximum input and output voltages and currents.  Presumably they include protection against fault conditions.

A MPPT solar charge controller works the same as any other solar charge controller except for providing greater power under conditions where maximum power is being drawn from the solar panels.  When supplying a battery system and load, excess load will be drawn from the battery.  The batteries should be protected against excessive discharge.

Lastly, is it possible to have both an 'island' system (that stays powered when the grid is off) as well as a grid tie-in system?

I am sure I have seen some grid tied inverters which support a battery bank for offline operation but a separate transfer switch and inverter could be used.

[IanMacdonald]

I imagine this is one of the UPS's that use a tray of NP6-12 gel cells? If so the energy density versus cost ratio is not very favourable for this kind of use. The main problem here is that the relatively high voltage demands numerous small batteries in series. You would be better to investigate a lower voltage inverter so that a few large deep-discharge batteries could be used instead. 

Modern MOSFET inverters do not suffer the losses at low voltage that the older BJT or thyristor types did. (Which is probably why that old unit used 192v) 48v would be a good choice. The main issue with anything lower would be the cost of the cabling for the high currents involved. 

[Malvineous]

Many thanks for the detailed responses!

@Seekonk: You're right it would be foolish to design a "proper" system around an old piece of equipment, but I'm thinking more about doing it as a bit of a fun project to learn about the topic and just to see whether it can be done, and of course that means doing it on the cheap.  So if it doesn't go anywhere I won't be out of pocket by too much but I'll have learned a lot in the process!  The 2 kW system you have sounds fine and would certainly be a great place to start as well.  As you say, having to buy 16 batteries of whatever type I want in order to get 192 VDC is going to be the biggest cost.  I like the huge 100+ Ah 2V single-cell batteries but of course I'd be looking at around $40,000 to buy the 96 of these that I'd need to get 192 V!

@fourtytwo42: Yes the batteries that come with the UPS only provide for some 10 minutes of runtime at full load, so I certainly wouldn't be using these.  The unit is designed to accept external battery packs, and while I wouldn't be using the UPS' own battery charger, it is rated to handle up to 10 packs, so 192 VDC * 5.5 Ah * 10 = ~10 kWh total.  But if I'm charging via panels then I don't have to worry about overloading the UPS battery charger and I can put whatever capacity I can afford onto it.  I have thought about using a grid-tie inverter, but I would like to start it off-grid and possibly grid-tie later.

Re the transformerless design, it does appear to be one of these (lots of big heatsinks but no transformers, and the unit is a lot lighter than the other 48 VDC UPS units I have from the same manufacturer which have two hefty transformers in them.)  However the negative terminal on the 192 V bank is bonded to the inverter neutral output, and I only show a reading of approx 1 VAC between either battery terminal and the mains neutral input.  There is a 240 VAC potential between mains active input and either battery connector.  Measuring on DC there's as good as zero volts between either battery terminal and a mains connection (active or neutral, mains input or inverter output.)  I take this to mean that the batteries aren't completely isolated but they also aren't at mains potential either, however I'm not 100% certain how to check this so please let me know if this isn't the right way to test it.

@David Hess: What are some of the additional hazards of DC over AC that would be important in this situation?  I've read mixed opinions about AC vs DC however the takeaway for me is that above a certain voltage it doesn't really matter whether it's DC or AC.  As long as the equipment is rated for the DC voltage in use then it seems to be fine.  Since I'm working with 240 VAC on the inverter output, it seems that adding 192 VDC on the input isn't any more dangerous than that.  One thing I like about the higher DC voltage is that at 5 kW I only need to handle around 30 A of current, whereas 48 VDC means I will have to size everything to handle over 100 A continuously, and I find those kinds of currents a little scary too!

@IanMacdonald: I believe this is a MOSFET inverter (no large transformers, manufactured 2006) so I think the 192 VDC is more about being able to use smaller wiring.  They seem to switch from 48 VDC to 192 VDC around 3 kW, then to 384 VDC at 7.5 to 10 kW.  The design seems to be to keep the DC current below 30 A.