navitron
 
Renewable Energy and Sustainability Forum
Welcome, Guest. Please login or register.

Login with username, password and session length
News: Anyone wishing to register as a new member on the forum is strongly recommended to use a "proper" email address - following recent spam/hack attempts on the forum, all security is set to "high", and "disposable" email addresses like Gmail, Yahoo and Hotmail tend to be viewed with suspicion, and the application rejected if there is any doubt whatsoever
 
Recent Articles: Navitron Partners With Solax to Help Create A More Sustainable Future | Navitron Calls for Increased Carbon Footprint Reduction In Light of Earth Overshoot Day | A plea from The David School - Issue 18
   Home   Help Search Login Register  
Pages: [1] 2 3   Go Down
  Print  
Author Topic: Sunny Island frequency 'blip'  (Read 5458 times)
camillitech
Hero Member
*****
Offline Offline

Posts: 5430



WWW
« on: September 20, 2016, 03:13:25 PM »

Thought this may be of interest to fellow 'Islanders',

the grid tied 2.5 Proven I've been running 'AC coupled' for 18 months now surpassed expectations right up to 'Storm Henry' when, after several hours of throwing out almost double it's rated output it 'fried' the core  Cry Twas my own fault I guess for using a WB6000 inverter on what was originally a 2.5kW turnip  whistlie Anyway, lesson learned I fitted a new core, changed the inverter to a recently acquired Aurora 3.6TL and clipped the power curve at 3.2kW.

This was a bit of an experiment as unlike the SMA kit the ABB/Aurora stuff isn't designed for 'off grid' AC coupling. The beauty of the Aurora is that the power curve is really easy to program and it'll connect to the grid as low as 80VDC unlike the SMA's 230VDC. Not only that but my Aurora is an Irish one so it doesn't have to wait 3 minutes to reconnect (as per G83/2) it will do it in as little as 1 second (though the default is 20 seconds)

Well, this worked a treat and didn't even 'drop out' when the generator was started up so I was well chuffed. On Sunday however I arrived home early from work to discover the batteries were only at 75%. This will be on account of mother in law, who's obsession with cleanliness and ironing would challenge the largest 'off grid' system. She fills the kettle to the top constantly then just keeps boiling it throughout the day. Her logic being that it'll make a quicker cup of tea when I come in  banghead

Anyway, with people coming around for dinner and my boy still not having had his 'half hour shower' I flashed up 'Harry' for a couple of hours. I stayed in the shed long enough to watch the transfer switch close and saw that the trusty Lister was feeding the house loads. The Aurora continued feeding the island grid and all was peachy. Several bottles of red later I goes to switch Harry off, by which time it's blowing a good Force 6 and the batteries are at 90 odd %. As soon as the transfer switch opened the Aurora 'dropped out' with a 'frequency out of range 50.99Hz'  Cry



Straight line = Hz spiky one = current, in true SMA style - is +  banghead 

You can see the genny starting at 18:29 and stopping at 20:37, then the frequency rises to almost 51Hz for an hour. As soon as it dropped out I got me laptop and examined the Aurora's settings. Sure enough it said 'frequency range 49.something to 50.49Hz, which is why it wouldn't connect. Strange thing is that I'd recently widened those parameters to +/- 1.5Hz  Huh and try as I might, even with the highest password I couldn't change it. Methinks there may be some time limit on this password.

After 15 minutes of messing about with the laptop I turned the Aurora off, had a large Black Russian and went to bed, woke up at 2:00AM to a gentle breeze, turned on the Aurora and all was 'peachy' once more. Last night, when I finally got around to examining the data I saw that clearly defined 'jump' of 1Hz for an hour. I'm wondering if it is heat related due to the generator being in the same shed as all the inverters. My FSPC (frequency shift power control) is always turned off, is it perhaps a glitch in the SI software?

Anyways I'm not too worried, especially if I can widen those frequency parameters again and keep them there.

Cheers, Paul
« Last Edit: September 20, 2016, 03:15:58 PM by camillitech » Logged

http://lifeattheendoftheroad.wordpress.com/

'Off grid' since 1985,  Proven 2.5kW, Proven 6kW direct heating, SMA SI6.OH, 800ah Rolls, 8kW PV ,4xTS45, Lister HR2 12kW, , Powerspout pelton, Stream Engine turgo, 60 x Navitron toobs and a 1500lt store. Outback VFX3048 and 950ah forklifts for backup,
Scruff
Hero Member
*****
Offline Offline

Posts: 1252



« Reply #1 on: September 20, 2016, 04:39:26 PM »

I don't think heat will affect frequency. Usually it's derived from mechanical RPM, a resonant device, a clock or an oscillator.  Huh
Logged
Scruff
Hero Member
*****
Offline Offline

Posts: 1252



« Reply #2 on: September 20, 2016, 04:45:10 PM »

Methinks there may be some time limit on this password.


Can you change the date on the SI?  fingers crossed!
Logged
book_woorm
Full Member
***
Offline Offline

Posts: 114


« Reply #3 on: September 20, 2016, 04:57:50 PM »

Are you monitoring the temperature in your plant room? Mines got 2 thermal stores in it and a west facing roof. In the summer I knew it got up above 30degC courtesy of an old mercury thermometer but when I instrumented it with electronic data recorder sensors I got a shock. Some of the equipment peeked at 50degC and some individual ICs/power MOSFETS got up to 90degC, because they could not lose heat fast enough. The crystal on an Arduino was not far behind at 82deg and it was definitely not running at the standard 16MHz as all the time controls were out. The Aurora frequency almost certainly controlled by a crystal so heat may be a culprit. Needs careful checking.
Logged

2.4 Kw Kyocera Panels (west facing) Feronius inverter; Sonenkraft Solar Thermal with Twin 180Lt & 280 Lt Thermal Stores; SAP 'A' rated property with UFH & wood burner. Full weather compensation on the UFH buffer temperature & differential controller decides where the heat from the wood burner goes.
camillitech
Hero Member
*****
Offline Offline

Posts: 5430



WWW
« Reply #4 on: September 20, 2016, 05:42:19 PM »

Methinks there may be some time limit on this password.


Can you change the date on the SI?  fingers crossed!

I could always give it a try  whistlie
Logged

http://lifeattheendoftheroad.wordpress.com/

'Off grid' since 1985,  Proven 2.5kW, Proven 6kW direct heating, SMA SI6.OH, 800ah Rolls, 8kW PV ,4xTS45, Lister HR2 12kW, , Powerspout pelton, Stream Engine turgo, 60 x Navitron toobs and a 1500lt store. Outback VFX3048 and 950ah forklifts for backup,
camillitech
Hero Member
*****
Offline Offline

Posts: 5430



WWW
« Reply #5 on: September 20, 2016, 05:50:28 PM »

Are you monitoring the temperature in your plant room? Mines got 2 thermal stores in it and a west facing roof. In the summer I knew it got up above 30degC courtesy of an old mercury thermometer but when I instrumented it with electronic data recorder sensors I got a shock. Some of the equipment peeked at 50degC and some individual ICs/power MOSFETS got up to 90degC, because they could not lose heat fast enough. The crystal on an Arduino was not far behind at 82deg and it was definitely not running at the standard 16MHz as all the time controls were out. The Aurora frequency almost certainly controlled by a crystal so heat may be a culprit. Needs careful checking.

Hi BW the 'power station' not only has the generator in but it's well insulated too, serious 'cock up' on my behalf but hindsight is a wonderful thing. I do have an LCD max/min thermometer in there and both the SI and Aurora record heatsink and other temperatures so I'll have a look tonight and see if there's any correlation between the two.

Cheers, Paul
Logged

http://lifeattheendoftheroad.wordpress.com/

'Off grid' since 1985,  Proven 2.5kW, Proven 6kW direct heating, SMA SI6.OH, 800ah Rolls, 8kW PV ,4xTS45, Lister HR2 12kW, , Powerspout pelton, Stream Engine turgo, 60 x Navitron toobs and a 1500lt store. Outback VFX3048 and 950ah forklifts for backup,
Scruff
Hero Member
*****
Offline Offline

Posts: 1252



« Reply #6 on: September 20, 2016, 06:12:02 PM »

Yurp..my bad...

https://en.wikipedia.org/wiki/Crystal_oscillator

https://en.wikipedia.org/wiki/Crystal_oven
Logged
bautsche
Jr. Member
**
Offline Offline

Posts: 86


« Reply #7 on: September 20, 2016, 07:29:50 PM »

Hi Paul.

Can you give us a more detailed graph of the Frequency only for the same time period? The resolution isn't really allowing me to see what the SI is doing. Or you could let me have the source data (you've got my email address).

Also: I note that you have FSPC turned off, but have you turned off AFC (Automatic Frequency Control)? On my SI5048 that's "250.11 AfraEna". From the more detailed data I would be able to see if that might have pushed your frequency up. Basically, the SI tries to keep an average frequency of 50Hz. This is to allow clocks that work on the power frequency not to drift overall. So if your genny produces a slightly reduced frequency and the SI is therefore forced to adopt that, after the genny stops, the SI will lift the frequency above 50Hz for a while to allow the average frequency to get back to 50Hz. Does that make sense?
A more detailed graph of the frequency or the source data for it for the period from some time before gen run until well after the end of the "wrong" high frequency would be able to confirm this.

HTH
Eric
Logged
heatherhopper
Sr. Member
****
Offline Offline

Posts: 352


« Reply #8 on: September 21, 2016, 12:43:27 AM »

Paul I agree with Eric regarding the graph. Told you manipulating data becomes a bit of a minefield!
The scale for the frequency does not really show any variation clearly. There is some variation just discernible during the Generator run period which suggests that you do have "AfraEna" enabled (SI synchronises and hands frequency control to the Generator) and Harry is leading the frequency control. Not sure how you would operate without it - would Harry not be an unusually well behaved generator if he were able to match the normal SI Hz control and how would the SI react if AfraEna was disabled and Harry ignored the nominal 50Hz?
You can display two data sets with completely different scales on one graph. Right click your graph, select "Edit", right click again and select "Insert/Delete Axes", select the option for a secondary Y axis and one of your data sets will be aligned with the right hand axis, alter the left and right Y axes scales as appropriate for the two data sets. Apologies if this is not quite the exact method in Excel but the principle is the same.

Looking at this data is a bit difficult bearing in mind your rather unusual set-up. I think you posted a schematic at some time but I can't find it?
By
Quote
watch the the transfer switch close and saw that the trusty Lister was feeding the house loads
I assume you mean Harry was started by the SI and connected through the AC2 connection supplying both Batteries and house grid?
What is the source of the "current" data you display? I assume it is "TotBatCur" from the SI data set but the overall profile looks decidedly odd to me. What are the spikes?

All my gear is housed in a building with temperature variations of just above zero to 30+ degrees C (latter sometimes for days in the summer). I have never seen any adverse effect on inverter behaviour but can't speak for the Aurora. Battery capacity is, of course, a very different thing and a pain in the neck and the way the SI and your DC chargers see and compensate for this may be worth some consideration.
May just be a glitch in the SMA software but with your complex system I would keep looking for a cause. Lots more data to trawl through!
Logged

Off grid AC coupled, 6kW Proven, 2.8kW PV, SMA SI/SB/WB Inverters, 4x576ah Rolls batteries @ 24v, 25kW Biomass Boiler, Wood Stoves, Spring/Well water. Sorry planet - I did try.
bautsche
Jr. Member
**
Offline Offline

Posts: 86


« Reply #9 on: September 21, 2016, 07:01:07 AM »

Just to be clear: When the SI connects to the generator the generator always governs the frequency. AfraEna will have no impact on this either way.

AfraEna will only impact the frequency when the SI is in charge of it but it will take account of what happened while the generator was in charge of it.

Re whether the genny was started through the SI or manually: that makes no difference. The SI will always use the genny when it detects it, even if it didn't start it itself.

HTH
Eric
Logged
heatherhopper
Sr. Member
****
Offline Offline

Posts: 352


« Reply #10 on: September 21, 2016, 12:07:51 PM »

Quote
Just to be clear: When the SI connects to the generator the generator always governs the frequency. AfraEna will have no impact on this either way.

AfraEna will only impact the frequency when the SI is in charge of it but it will take account of what happened while the generator was in charge of it.
Indeed Eric - looks like I misread the function - I stand corrected. My little 2224s don't have this option which makes me wonder if it is integrated in them or not present at all?! I have certainly never seen frequency fluctuation outside of generator operating times other than that I associate with FSPC which is very clearly led by battery voltage. Frequency, without excess AC coupled generation, is always 49.9-50.0 Hz. I am curious now and will look at my data but doubt I'll be able to separate any averaging from all the noise created by the active FSPC.

Quote
Re whether the genny was started through the SI or manually: that makes no difference. The SI will always use the genny when it detects it, even if it didn't start it itself.
I was really just asking if Paul connected his generator to the house loads via a bypass switch (I think he has one) between the "Island Grid" and the house on this occasion.
Quote
I stayed in the shed long enough to watch the transfer switch close and saw that the trusty Lister was feeding the house loads. The Aurora continued feeding the island grid and all was peachy.
seemed a bit ambiguous. Also why I queried the "current " data source.

Logged

Off grid AC coupled, 6kW Proven, 2.8kW PV, SMA SI/SB/WB Inverters, 4x576ah Rolls batteries @ 24v, 25kW Biomass Boiler, Wood Stoves, Spring/Well water. Sorry planet - I did try.
bautsche
Jr. Member
**
Offline Offline

Posts: 86


« Reply #11 on: September 21, 2016, 12:11:19 PM »


Quote
Re whether the genny was started through the SI or manually: that makes no difference. The SI will always use the genny when it detects it, even if it didn't start it itself.
I was really just asking if Paul connected his generator to the house loads via a bypass switch (I think he has one) between the "Island Grid" and the house on this occasion.
Quote
I stayed in the shed long enough to watch the transfer switch close and saw that the trusty Lister was feeding the house loads. The Aurora continued feeding the island grid and all was peachy.
seemed a bit ambiguous. Also why I queried the "current " data source.


[/quote]

The SI has an internal transfer switch which locks the SI grid and the genny together.
Based on Paul's graph, he didn't disconnect the SI from the house. (that's my reading anyway, and there'd be no reason to do that, either, that I can see).

Eric
Logged
camillitech
Hero Member
*****
Offline Offline

Posts: 5430



WWW
« Reply #12 on: September 21, 2016, 01:22:50 PM »

Good afternoon fellow 'islanders' and thank you.

lost my internet with no prospect of a return until next week so I'm just getting online occasionally, whilst out and about.

Pretty sure I disabled 'AfraEna' but a good point and I'll check. Those queer spikes on the 'Battery Max Charge Current' are the immersion coming on H and the generator was started manually via the SI's 'start function'. Yes I think my comment re the transfer switch may have been a little misleading, I was meaning the SI's own internal one. One thing I did notice whilst having a look again last night was a big increase in Heat sink and chip temperatures whilst the generator was running. However this fell rapidly once Harry was stopped.

I'll send you both the data if I can get near a satellite dish this evening or tomorrow morning, meanwhile I'll have another play with it myself. I'd put it on here if I knew how to, I'm sure there must be a way.

Anyway, gotta go, battery dying and things to do, many thanks for the pointers, Paul
Logged

http://lifeattheendoftheroad.wordpress.com/

'Off grid' since 1985,  Proven 2.5kW, Proven 6kW direct heating, SMA SI6.OH, 800ah Rolls, 8kW PV ,4xTS45, Lister HR2 12kW, , Powerspout pelton, Stream Engine turgo, 60 x Navitron toobs and a 1500lt store. Outback VFX3048 and 950ah forklifts for backup,
Ted
Global Moderator
Hero Member
*****
Offline Offline

Posts: 3919



« Reply #13 on: September 21, 2016, 04:38:58 PM »

You could put the raw data on here if you can save as txt or xls and then just attach.
Logged

Volunteer moderator
bautsche
Jr. Member
**
Offline Offline

Posts: 86


« Reply #14 on: September 22, 2016, 07:12:19 AM »

Hi Paul.

Here's what your frequency graph looks like on its own:


I find the two spikes earlier in the day interesting, too.
I have actually graphed InvFrq (Avg) and the first spike is at 05:06:20. Nothing is out of the ordinary for the min and max frequency, but the average seems low. Looking at the rest of the data, there doesn't seem to be anything else spectacularly happening at the time, the most I can say is that your battery discharge went up to -12A from -5.7A, but given the size of your bank, that really shouldn't have that kind of impact.

Ignoring the two spikes that peeked my interest and going back to your problem with the frequency:
I appreciate that your config says you have AfraEna enable turned off, but from the graph it sure doesn't look like it.
I wonder whether we are all misreading what happens if you turn it off in the manual. I've had that before with mine where from the manual I expected a particular behaviour just to find out when I queried it with SMA that I had completely misunderstood and the end-user description of it was not sufficiently clear.

Anyway, if you average your SI frequency from the start of the day until 18:15:14 you get 49.98Hz and from 18:30:29 (just before you started the genny) until 21:40:36 (just after you frequency drops back down to 50Hz) you get an average of 49.97Hz.
It sure looks like your SI tries to achieve an average frequency of 50Hz to me, i.e. that (from my understanding of how it works) AfraEna is in fact set to "true".
You may want to double check the setting and if it is indeed set to "false", I'd first ask SMA what the expected behaviour is....
That behaviour doesn't match with my expectations as far as I understand the documentation.

HTH
Eric

Logged
Pages: [1] 2 3   Go Up
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.20 | SMF © 2013, Simple Machines
SMFAds for Free Forums
Simple Audio Video Embedder
Valid XHTML 1.0! Valid CSS!