Simulating presence of Grid Power

mickatdunge

Hi

I'm in the UK, I've been off grid 15 years using: Air X, 6KVA Lister Generator, Xantrex 3024e and 720Ah wet nicad battery store. However I'm now going to take advantage of the UK Government's Feed in Tariff. My installer has put up 3.5kw of Solar monocrystalline panels with 1 x Sunny Boy 3800 driving a 240 vac usage meter (only 240 VAC meters are authorised under the FiT). Unfortunately he now realises that the SB 3800 needs to detect something like grid power in order to produce power (he did know I was off grid and I think had missed the implications here).

I just want 240 VAC presented to my Xantrex 3024e. I had planned to use its two battery voltage detection relays to switch in sufficient loads to soak up excess power from the panels. I was also going to double the size of my battery store. Unfortunately both remain theory until the supplier comes up with a solution which produces any power at all. From the research I've done, it would seem that one (or perhaps 2) Sunny Island 2224s would be the answer as they allow a bi-directional AC bus which would a) simulate the grid's presence b) allow loads (perhaps including the 3024e) to be switched in and (I think) c) charge the batteries if I don't want to use the 3024e.

Are there any other alternatives that I might be missing? I appreciate that it's not my job to fix it but I'd feel happier if I felt confident in the solution as I'm nervous about the installer at the moment.

Thanks - Mick

Ralph Day

Re: Simulating presence of Grid Power

So Mick, you can take advantage of a feed in tariff (as do I, see signature),without being tied to the grid? Would the purpose be to get paid for the power you produce yet only you consume? An interesting concept. Our fit programs in Ontario required your connection to and sale only to the grid (no use of any power before the grid gets it).

There are several threads about convincing grid tie inverters that a grid is present...when there isn't (or it's a micro-grid). I'm sure there will be responses with advice, not just observations like mine. Good luck.

Ralph

Cariboocoot

Re: Simulating presence of Grid Power

Welcome to the forum Mick.

One question: how are you going to take advantage of a government feed-in tariff if you are not actually going to feed the grid? I may be confused, but that sounds like a misrepresentation of the install. If they are going to allow this even though you are feeding only your own private system, why buy a GT inverter set-up that is basically a pain in the anatomy to synch to an off-grid system?

Xantrex's XW series and the SMA Sunny Island will synch to GT inverters in an off-grid application. Most other OG systems have difficulty with it as they are not designed to synch with anything external and don't have any means of regulating the incoming AC power. That means they will simply reverse-feed it to the batteries until the batteries over-charge and burn up.

Many here have pondered this same issue: how to get regulation of the battery charging by adding AC loads or draining off DC or disconnecting the GTI above a certain battery set point. Still waiting for positive results.

Usually they want to do this because they're looking for power from the GT system when the grid goes down. I just don't see the point in your case; you seem to be expanding your off-grid system with equipment that is entirely wrong for the application in order to take advantage of a government incentive that wasn't meant to be used in this way. Maybe you can clarify the whole feed-in tariff criteria for us?

BB.

Re: Simulating presence of Grid Power

Ralph asked the same question I was going to ask "... the government is willing to pay you a Feed in Tariff where you generate the AC power, pass it through a 230 VAC 50Hz (I assume UK power) to a load bank where you simply burn off the excess power????".

Oh well... Incentives work, even when they’re perverse.

Any who, Yes, besides the Sunny Island (I am sure a very well designed and expensive solution), there is the Xantrex/Schneider XW Hybrid inverter system--Both which will do battery charge management "natively".

And (I think) you can use some Outback and Magnum TSW inverters (I think)--but you have to design/implement your own battery bank storage solutions. Mfg. may or may not support. Many TSW inverters are actually bi-directional and can connect to a (smaller) GT inverter and act like a "local power" grid. GT Inverters usually shut down if frequency variation is worst than +/- 1% (or ~1/2 Hz) from 50/60 Hz regulation. So, some cheaper inverters and most generator sets do not have accurate enough frequency control to support GT Inverters.

OutBack Power
Magnum Energy Inc.

By the way, if they will pay for power from an island Grid Tied Inverter--See if they will allow you to put the revenue meter on your Off Grid Inverter output... You will "log" less energy generated--But it is, at least, a more reasonable number (paid for energy you actually generate/use vs all power including that being burned off in a load bank).

Also--on another subject. How do you like your NiCad battery bank?

-Bill

stephendv

Re: Simulating presence of Grid Power

Yes, as crazy as it may seem the UK pays FITs based on the generation of renewable energy. It doesn't matter whether that energy is used off-grid or fed into the grid. And they're about to revise the FIT rates because of the huge takeup of what is effectively a reverse robin hood tax.

Anyway, back to Mick's problem. The sunny island's would be the ideal solution as they'd replace your current xantrex and manage the whole system for you very nicely. You'd need at least 3.5kW of battery inverter, so the 2 x 2024 would work or else the older model 4024 (this model can't do generator boost, which the newer 2024 can do).
A nice feature of the system is that the frequency output by the SI increases in proportion to how much surplus power is available. So you can use things like frequency dependent relays to turn loads on whenever there's surplus power available, see DILCs at the bottom of this page: http://www.windandsun.co.uk/Inverters/sunny_island_inverters.htm

If you'd like to use other bidirectional battery inverters (outback, studer, victron) - they won't control the load for you, so you could use the relays as you suggest or put a dump load controller across the battery- but then you're limited to dumping heat. As Bill mentioned the Xantrex XW is another inverter that will increase the grid frequency in order to nudge the sunny boy offline when the batteries are charged. But this is an all-or-nothing system, whereas the sunny island is variable. For the marginal difference in price between the xantrex and sunny island (in Europe the price difference is small, whereas in the US the sunny island is much more expensive) - I'd go with the SI.

Cariboocoot

Re: Simulating presence of Grid Power

I can see giving incentives to people who supply their own power as it relieves some demand from the grid.
That being the case, I don't see why Mick went with GT equipment instead of metering his existing OG output and/or buying additional OG equipment. Probably more bureaucratic misunderstanding (they're good at that). Why anyone should be paid for generating and using power they don't actually need is beyond all logical comprehension. But what has logic (except for the convoluted variety) ever had to do with government?

Like Stephen said, the Sunny Island is probably the best bet here.
Or if the existing OG system meets your power needs, connect the GT system to the grid and sell all the surplus.

mickatdunge

Re: Simulating presence of Grid Power

All

Many thanks for the quick and informed replies, I'm really impressed. There don't seem to be many people here who understand the problem, let alone have the answers. I'll research the answers but the Sunny Island seems to be the favourite approach. I think they might make a 'backup' inverter, so I'll have a look at that too. If I've got to replace my nice 3024e then as the supplier has already installed the SB 3800 I can start with one SI 2224 and afford another in a year's time.

As Stephendv has confirmed, the UK's FiT has two components about 3 UK pence per kilowatt hour for actually feeding electricity into the grid and about 43 pence for actually generating the power. The perverse incentive is that it encourages you to waste power (although 3.5 kw isn't a particularly huge consumption) but it's the case that, if I weren't using PV, I'd have to turn the generator on: I'd argue that the PV is better than burning oil. Plus, I'll be able to have electrically heated hot water in the summer and a freezer (rather than using LPG for both).

Thanks - Mick (off to pub - after I've watched the ISS go past)

stephendv

Re: Simulating presence of Grid Power

mickatdunge wrote: »

All
If I've got to replace my nice 3024e then as the supplier has already installed the SB 3800 I can start with one SI 2224 and afford another in a year's time.

From what I remember about their system, you have to match the size of the battery inverter to the size of the PV installed, so you'd need an inverter with at least 3.5kW capacity. Would recommend you also post this question to a uk forum, there are a handful of UK off-gridders there who are in the same boat: http://www.navitron.org.uk/forum/

Yes 'coot it's a bureaucrats dream. Firstly they installed a ridiculous approval system where only approved panels, inverters and installers would be allowed for any FIT system- the equipment was even listed by model number, so the wrong model of panel would mean that you can't claim the FITs
Then all the approved generation meters were AC- and since almost all off-grid charge controllers are DC based - they can't (or couldn't) be used to claim FITs.
They wouldn't let you measure the output of the battery inverter because you might have the diesel generator on... but if you'd modified the generator to run on renewable oils then you could claim the FITs.

Someone needs to invent a machine that harnesses the energy of paper shuffling across a desk.

Cariboocoot

Re: Simulating presence of Grid Power

Another fine example of having people who don't understand a situation create the regulations that govern it. :roll:

solarvic

Re: Simulating presence of Grid Power

An addition to #4 post I am researching AC coupling my Magnum inverter that I have. The latest model Magnum MS4024PAE model has software that will AC couple with a gt inverter. When your battery gets charged it raises the cycles to 60.6 cycles and the gt inverter will shut off. I don,t have the pae model so I will have to provide a dump load to prevent overcharging the battery. Magnum recomends a GT inverter with not more than 90% of the magnum inverter. :Dsolarvic:D

stephendv

Re: Simulating presence of Grid Power

stephendv wrote: »

From what I remember about their system, you have to match the size of the battery inverter to the size of the PV installed, so you'd need an inverter with at least 3.5kW capacity.

Turns out I was wrong about this. You can install PV up to 2 x the rated size of the battery inverter, so a single 2224 would be fine with 3.5kW of PV. Details in one of the sunny boy docs about tying to SI on sma.de site.

mickatdunge

Re: Simulating presence of Grid Power

stephendv wrote: »

Turns out I was wrong about this. You can install PV up to 2 x the rated size of the battery inverter, so a single 2224 would be fine with 3.5kW of PV. Details in one of the sunny boy docs about tying to SI on sma.de site.

Stephen
Thanks for confirming this. I've just read through the technical specification for the SI2224 and I began to suspect that this must be the case. I think the 2200 watts refers to its ability to invert that much power from batteries (which is fine by me). The only worry I have left is whether my 6KVA Lister Genset will be too much for it (given that, at 240vac, that equates to the 25 amp max for input).
Apparently I (or my installer) also need a SMA Remote Controller to configure the SI and also a RS885 comms port to plug into the SB3800. They make it quite hard (and expensive) don't they?

Thanks again - Mick

stephendv

Re: Simulating presence of Grid Power

An overpowered genset won't be a problem as long as you don't try and draw more than 25A through the SI while the generator is on. I guess you've seen the charger specs: 80A, so typically you can only charge with around 2.3kW which will be a waste both for the 6kVA lister and the 3.8kW of PV- it'll also make a dent in your FIT payments.
The remote controller will always be useful, so that you can see what's going on in the system. You only need the RS485 module in the SB to set it up initially, once it's configured, you can take it out. I'm sure your installer would have one spare that he could use to program the thing with.
Yep, it is quite expensive - firstly because you've gone the AC coupling route with a grid tied inverter that costs almost 3 times as much as a DC charge controller, and then you're working with SMA which is not the cheapest. But if you really have to go the AC coupling route then in my opinion the SI + SB combination is your best option.

mickatdunge

Re: Simulating presence of Grid Power

stephendv wrote: »

An overpowered genset won't be a problem as long as you don't try and draw more than 25A through the SI while the generator is on. I guess you've seen the charger specs: 80A, so typically you can only charge with around 2.3kW which will be a waste both for the 6kVA lister and the 3.8kW of PV- it'll also make a dent in your FIT payments.
The remote controller will always be useful, so that you can see what's going on in the system. You only need the RS485 module in the SB to set it up initially, once it's configured, you can take it out. I'm sure your installer would have one spare that he could use to program the thing with.
Yep, it is quite expensive - firstly because you've gone the AC coupling route with a grid tied inverter that costs almost 3 times as much as a DC charge controller, and then you're working with SMA which is not the cheapest. But if you really have to go the AC coupling route then in my opinion the SI + SB combination is your best option.

Hi Steve

Resurrecting an old thread here. The system's been working fine for the last 18months or so and now has almost 1000 amp hour at 24 volts of NiCad (thanks eBay). I picked up a second SI2224 and was wondering if I could increase the AC capacity of the system by using them in tandem? I fancy (or fantasise, I'm not quite sure) about an electric car but think I'd need more power to charge it at c3000watts than the single SI2224 would be happy with. A Nissan Leaf has 24 KwH of battery which is pretty much what I could produce from my system on a Sunny Day. Free motoring is an appealing prospect for local pottering around at least. What d'you think (about operating SI2224s in tandem not the car!)

Thanks - Mick

stephendv

Re: Simulating presence of Grid Power

Yep you can absolutely use them in tandem but they MUST be configured to do this and they must both have the communication boards installed. All explained in the manual.
Having another SI, means you can draw more from the battery- but for electric car charging, I would have thought that you'd be more interested in charging it from solar, not from the battery. Charging an electric car through the SI's would effectively mean twice the charge discharge losses as you discharge the nicads and charge the electric car.
If car-charging was your only requirement, then it would have been better to first get more solar PV and then get another sunny boy so that you have more power available on the AC side. With your existing 3.8kW sunny boy, you could go up to 4.5kW or so of solar so that you have a steady 3.8kW output in realistic conditions. Which should enough to charge the car.

Then using one of them frequency dependent relays you could hook it all up, so that the electric car only charges when there's surplus PV power available.

mickatdunge

Re: Simulating presence of Grid Power

stephendv wrote: »

Yep you can absolutely use them in tandem but they MUST be configured to do this and they must both have the communication boards installed. All explained in the manual.
Having another SI, means you can draw more from the battery- but for electric car charging, I would have thought that you'd be more interested in charging it from solar, not from the battery. Charging an electric car through the SI's would effectively mean twice the charge discharge losses as you discharge the nicads and charge the electric car.
If car-charging was your only requirement, then it would have been better to first get more solar PV and then get another sunny boy so that you have more power available on the AC side. With your existing 3.8kW sunny boy, you could go up to 4.5kW or so of solar so that you have a steady 3.8kW output in realistic conditions. Which should enough to charge the car.

Then using one of them frequency dependent relays you could hook it all up, so that the electric car only charges when there's surplus PV power available.

Thanks again Stephen. Yes I'd intend to charge the car when there was surplus during the day. I guess my lack of understanding is whether the SI would supply more than the 2.2kw it was rated at when there were excess watts coming off the panel or is the 2.2kw only the limit when you're taking power off the batteries? I've actually got another SI (I bought one 'in case' when they stopped making them). I'll have to get it a communication board, I imagine this is the same board that it uses to communicate with the other devices on the network (Web Box and something else I can't remember, perhaps the Remote Control?). Now you mention the manual I remember that there was a concept of master and slave for the SIs. I'll have to research the car now, I think I'll be unable to resist the temptation to bankrupt myself by attempting to save money! I've specced an Arduino device which a friend has built for me, it senses the rises in Hz and switches devices on to use up the extra power. I thought I was going to heat water, I'm not now!

stephendv

Re: Simulating presence of Grid Power

mickatdunge wrote: »

I guess my lack of understanding is whether the SI would supply more than the 2.2kw it was rated at when there were excess watts coming off the panel or is the 2.2kw only the limit when you're taking power off the batteries?

Imagine have a single AC bus, which is your own private mini-grid. The SI pushes battery power into this grid, with max of 2.2kW and the sunny boy puts 3.5kW into this same grid. So on a theoretical ideal day, you could draw 5.7kW from your grid.

I'm not 100% sure which communication boards are suitable (they call them "piggy backs"), the RS-485 one definitely is, but don't know whether bluetooth and serial would also work.

mickatdunge wrote: »

I've specced an Arduino device which a friend has built for me, it senses the rises in Hz and switches devices on to use up the extra power. I thought I was going to heat water, I'm not now!

Very nice! I'm sure there are many sunny island owners who could use a device like that, please keep us updated on the progress. Once you have the arduino up and running you could then set priorities for the loads, e.g. if car is below 60% and water is above 30 degrees, then first charge car (or whatever rules make sense for your usage).

I'm using a raspberry pi and some software to log all the data to a database and turn some loads based on them, details here: http://www.casanogaldelasbrujas.com/blog/island-manager/
Not going for electric car just yet, first will be hot water then maybe electric bicycle trailer

mickatdunge

Re: Simulating presence of Grid Power

Stephen

Thanks for the reply. I'm sorry that I'm such an intermittent presence but I'm working away (retiring soon) at the moment and don't often get the chance. I suspected that the inverters combined on the AC bus but feel relieved now that you've confirmed it - I'd been limiting my usage to about 2.5kw so as no to strain the Sunny Island if it wasn't the case. I'll abuse it a bit more now. I've already put an RS485 board into the SB and SI and I can look at them on my Sunny Webbox. I also upload to the SMA site and track on my iPhone with the app.

The Mango, Raspberry Pi, SQL thing looks very interesting. I'll ask my guru about linking a Pi to the Arduino. As I'm shortly to have more time I might use it for playing with this type of stuff. The Arduino is plugged into the mains, measures frequency and its relays light up predictably but my loads are still attached to some frequency sensing relays I've been using. Again, with more time I'll experiment and let you know how it goes.

As to the car, it's probably a next year thing, it won't be cheap and I could really do with civilising the track to my house rather than wrecking a newish EV. I suppose I really should start with an electric bike but I can still (just) muster enough enthusiasm to leap before walking.

Mick