Use Power Transfer Switch Backwards for Grid Backup

[Guest]

I have a Server system that I wan't to power using solar, but have the grid available as a backup.

The way I'm planning on doing this is as follows:

I'm planning on having a solar array that charges batteries, and a power inverter to transform the power to AC power for the Servers. At first, due to cost, I'm not going to be generating enough solar power to run 24x7 and I thought the best solution would be to use a power transfer switch, but in reverse. I would connect the power from the inverter to the spot on the inverter you would normally connect grid power, then power from the power grid to the generator spot. This way once my inverter cut out the power grid would supply power until solar could take back over.

Is this ok to do? The load would only be 200 watts at first, then 500 total.

[boostinsteve]

500 watts 24/7 is not a load I would do this with. A better solution would be just to use the grid since you have it. You are going to find out very quickly that what you take off grid like this is going to cost you so much more money.

To answer the question, you could, but not ideal for your batteries. They are going to get beat to hell cycling them to the low voltage cutoff that most have.

To get technical, I work for ASCO. I know far more about transfer switches than the average Joe. The way I would do it is get a nice charge controller, for instance a midnite classic with the whizbang, and use the SOC function to close one of the aux relays. Once it is closed, most ATS's I have ever seen have a function called remote transfer to emergency. I would use that and wire the grid to normal, and the inverter to emergency. This protects your batteries far better, and will allow you to stretch the time ran on batteries as far as possible. Good luck

Edit:A lot of what you are wanting to do will depend om the equipment you have. In this case since you have a grid, what about grid-tied.

[Mike90250]

In your situation, I would suggest a Grid-Tie installation, and back up your computers with computer rated UPS boxes. Non-computer rated transfer switches switch slower then UPS's and will likely glitch your computers a couple times a week. Just takes one poorly timed glitch during a disk write to mess things up nicely.

[boostinsteve]

In your situation, I would suggest a Grid-Tie installation, and back up your computers with computer rated UPS boxes. Non-computer rated transfer switches switch slower then UPS's and will likely glitch your computers a couple times a week. Just takes one poorly timed glitch during a disk write to mess things up nicely.

Yeah, even with a transfer switch, you still need a power bridge such as an ups. A typical ATS will have a transfer time of 50-100 msecs. Add in any time delays, and you have a much longer time between power and no power.

[Sunking]

Yeah, even with a transfer switch, you still need a power bridge such as an ups. A typical ATS will have a transfer time of 50-100 msecs. Add in any time delays, and you have a much longer time between power and no power.

A real Line-Interactive or Dual Conversion UPS has no interruption or switching time. Critical equipment runs on the inverter 100% of the time.

[Sunking]

spilegi your idea will not work, and even if it did piss poor way to do ir. You don't need solar to do what you want. Do what the pros do, use a Dual Conversion UPS. Rectifier > Battery > Inverter

[thorbjornw]

I have considered something similarly, but on a much larger scale. I am installing a 46 kW solar system on a hotel in Nicaragua with a 144 kWh 48V Ni-Fe battery bank (already ordered the batteries and the solar panels), but I am scratching my head regarding the chargers and inverters. My plan is 9-10 XW+ MPPT80-600 Charge controllers and 6 XW+ 6848, but I am unable to understand the configuration possibilities. As my load is up to 25 kW, i have to install in parallel, and seems I will need to use an External 200 Amp contactor, to avoid damaging the inverters. However, from the documentation I understand that this would make it difficult/impossible to maximize self consumption. I expect only to use the grid in the early morning hours, part of the year, for grid back-up. The rest of the time the PV system will produce enough for our consumption and charging the batteries. So I would like the grid to cut-in in the early morning hours, when the batteries are depleted - automatically. Therefore the idea of a reversed transfer switch seemed to me a possible solution. Anyway, when there is nor battery nor PV, there should be no problems connecting the the grid.

[Mike90250]

I have considered something similarly, but on a much larger scale. I am installing a 46 kW solar system on a hotel in Nicaragua with a 144 kWh 48V Ni-Fe battery bank (already ordered the batteries and the solar panels), but I am scratching my head regarding the chargers and inverters. My plan is 9-10 XW+ MPPT80-600 Charge controllers and 6 XW+ 6848, but I am unable to understand the configuration possibilities. As my load is up to 25 kW, i have to install in parallel, and seems I will need to use an External 200 Amp contactor, to avoid damaging the inverters. However, from the documentation I understand that this would make it difficult/impossible to maximize self consumption. I expect only to use the grid in the early morning hours, part of the year, for grid back-up. The rest of the time the PV system will produce enough for our consumption and charging the batteries. So I would like the grid to cut-in in the early morning hours, when the batteries are depleted - automatically. Therefore the idea of a reversed transfer switch seemed to me a possible solution. Anyway, when there is nor battery nor PV, there should be no problems connecting the the grid.

I'll advise you to stop buying stuff and hire a solar electrical engineer. It will save you a bunch of $ in the long run.
Consider the $ you invested in the batteries, your education expense.

1) I can't even begin to imagine how much distilled water a battery bank that size is going to consume,

2) 144Kwh bank @ 48V = 3000ah. I have a 800ah bank, that supports a single XW6048 With 4x my battery, you will barely be able to power 4 larger inverters.
trying to power 6 is a pipe dream, The NiFe batteries have high internal resistance and cannot deliver huge amounts of current

3) Switchgear and cables needed for all this is going to be crazy expensive This install should be using >140VDC industrial gear, not homeowner gear

[thorbjornw]

Thanks for your optimistic advice.
I don't worry about the distilled water consumption. I can get inexpensive gear to solve that (http://www.sbsbattery.com/products-s...deionizer.html). Regarding the battery bank, according to the documentation available, I should charge and discharge at C/4 or C/5. That will give me more than enough for my needs. I am measuring my consumption in real time, and it normally lies within 4-10 kW. Occasionally, during the early evening hours, when all the A/Cs are on, it goes up to 10-12 kW - the 25 kW is the contract I have with the Electricity Company. So I am not worried about that. This is a slow charging and discharging environment.
I am no fan of Lead-Acid batteries. My Rolls back-up 40 kWh Battery Bank, never cycled below 70% DOD and which should have lasted 10 years, is already failing after 4 years with at least 5 bad cells, despite being cared for as a spoiled baby. As for Li-Ion battery banks, they are still expensive and short-lived, and I have no way of recycling them here in Nicaragua. That is the reason going for old-fashioned and well-proven low-tech non-toxic Ni-Fe. It needs maintenance, but that we can provide.
What is not clear to me, however, and what the solar engineers have not been able to answer, is if there are problems with 6 inverters in parallel (the solar engineer is proposing 3 dual inverter blocks). According to the Schneider documentation, this may cause problems during power cuts as they do not all switch simultaneously, so the last switching may blow the fuse. So therefore I may need to put in a contactor/tramsfer switch. But that, according to the Schneider manuals, may make it difficult to do load shaving.
This is the reason for considering putting in a 300 amp contactor to simply go off grid most of the day, and on-grid in the early morning hours, when the battery bank goes down to 40V. The aux port can do that, and control a micro-contactor that can control the heavy duty contactor. That is, if it really is a problem with paralleling 6 XW+ inverters.
By the way, we (my electrician and I) have 4 years experience after installing and running a 4.2 kW solar system, and it has worked well. Except for the back-up battery bank (never used for cycling).
But I can see, you have Ni-Fe batteries. And you don't seem to be happy with them?

[Mike90250]

OK, you think you have those bases covered. I'm only running a single inverter and from what I recall, the XW's cascade, as one approaches 80% load, the next one starts up and so on, But there is a finite number that can be stacked.
An alternative, is to hard wire each inverter to it's own dedicated load, They won't load share, but neither will they cascade fault either. They are limited by their internal transfer relay (i think it's 60A AC).
And as long as you have studied the NiFe batteries, that's OK, but because of their high internal resistance, you will not be able to service much load with them for very long. I'm using 800ah for a house with a daily usage 7.5Kwh
But the XW gear does have a decent low voltage cut out that makes it possible to use NiFe

[thorbjornw]

Yes, I have considered splitting the load up, but the disadvantage is that I will have to have redundancy on each circuit. The problem is, as I understand it, the 60 Amp limit on the internal transfer, when there is a power cut. I will look at it with my electrician. One reason for choosing XW is exactly that I can go down to 40V cut-off voltage. It sounds worrying that you are only able to get 7.5 kWh out of a 48V 800 Ah Ni-Fe bank. That bank should store around 38 kWh, so that is only around 20% of the stored energy. Why is that?

[Mike90250]

I have 3 days of juice in the bank, but the problem is the battery internal resistance. Under load, the voltage drops. A lot. So you have to have extra capacity to allow for that, With lead acid, I had a 400ah bank. 100% oversize to compensate for the high resistance, They used to mention that on the NiFe sites, take a look again for it.

[thorbjornw]

So that would around 60%. I am clear, the Ni-Fe has limited capacity for heavy load. The good thing about our load is that it does not have extreme peaks - it is relatively evenly distributed during the evening and night, with a peak for 10-15 kW from 8 to 10 pm, and after that it keeps running at around 4-9 kW during the night hours. But whether my 3,000 Ah battery bank will be enough, I am not sure. Anyway, we just shift to grid power, when the batteries go too low. I had calculated being able to take 80% out, but it might be too optimistic.
I guess you have put cut-off voltage in the inverter at the minimum (40V)?
Did you follow the factory advise charging and discharging 3-5 times at C/4 to run in the batteries?
And did you try the recombination caps to limit water consumption?

[Mike90250]

.......And did you try the recombination caps to limit water consumption?

I've not seen valved recombination caps that fit the NiFe cells - do you have a source ?

[thorbjornw]

I saw it somewhere in discussions on Ni-Fe batteries, but now I can't find it again. Someone who had experimented with them on some of his cells, and apparently with success. If I find it, I shall send the link to you.