Spiking generation

mickyduck

https://www.sunnyportal.com/Templates/PublicPageOverview.aspx?page=0786c32c-0ae8-4671-a50c-fa0f603f061e&plant=d05829e5-eac5-4c41-8428-f497616c5bbb&splang=en-US

strange how this regularly spikes then drops. Can anyone explain what is happening. I assume its something to do with the inverter as obviously cannot be the sun / cloud

Martyn1981

mickyduck wrote: »

https://www.sunnyportal.com/Templates/PublicPageOverview.aspx?page=0786c32c-0ae8-4671-a50c-fa0f603f061e&plant=d05829e5-eac5-4c41-8428-f497616c5bbb&splang=en-US

strange how this regularly spikes then drops. Can anyone explain what is happening. I assume its something to do with the inverter as obviously cannot be the sun / cloud

Hiya, if it's not the weather, is there any chance that the inverter is shutting down? Two possibilities;

1. Overheating, is it located anywhere that is very hot. Need to check the inverter specs, but possibly reduces it's output from around 40C upwards. Just checked and SMA say it's temp range is up to 60C, but I can't find info on the temp when it starts to reduce output to protect itself.

2. Over-voltage on the grid. The grid should run at 230V (-6% +10%). The inverter might/should shutdown if grid voltage goes over a set amount. My SE2200 did start switching off one day, when grid voltage hit 253V (+10%). If this is the problem, then it might be possible to change the point it shuts off 'IF' it's been set below 253V. Some can be set even higher, but I'm not sure if that's permitted by our DNO's. Your SB4000TL should display/report what the grid voltage is.

Hopefully it's neither of those options, and something much simpler.

Mart.

mickyduck

Hi Martyn
Thanks for the reply
if you scroll back a couple of days (hover mouse over bottom diagram left edge until arrow appears then click) you will see it very dramatically illustrated on the 8 and 9 July where the weather was perfect. Inverter located in Garage so not overly hot......... certainly not as hot as it would have been in my loft.

rogerblack

Note that the generated power - kW - is a nice smooth (ish) curve.
The integrated kWh measures are screwy in several ways.

Firstly - for it to be total yield - it should be a line, rising over the course of the day, not closely following the generation curve, as is clearly ridiculous.
Similar to the green area in http://pvoutput.org/intraday.jsp?id=22852&sid=20740&dt=20130711&gs=0&m=0

Ah - reading the graph more closely including the mouseover indicates it's kWh over 15 minute periods.

Secondly, the spiking.
My first guess would be that this is entirely a data problem, brought about by improper averaging of the submitted data.
It's - IMO - way too regular for something like the inverter overloading - which would never, ever be _that_ consistent,

Mousing over the graphs over several days reveals more oddness.
If you look carefully at the graphs, you see that the 'paired' data is quite obvious, even as generation drops - however it stops at some point.
A few moments inspection shows that it's not changing at a given yield.
However, it is exhibiting this behaviour at between 9AM and 7:15PM - whatever the generated outputs between this time.

Is there any alternate way to get data out of the inverter?

mickyduck

Hi don't know if I can collect the data any other way this comes via my Sunny Webbox... I think its probably a data collection issue as if I only collect it hourly theses spikes are not apparent, I only recently switched it to 15 mins hence the reason I have just noticed this anomaly.

zeupater

Hi

I agree with Roger - it looks like a timing issue to me .... something like aggregating x data readings in one period & x+1 in the next caused by misaligned clock settings with differing collection frequencies between different pieces of kit would certainly cause a similar effect.

I haven't looked at the specs of the Sunny WebBox since before we had our system and the only member who I can think of who has one is Freejuice and I haven't seen a post from him(?) for a while ... the pdf manual (http://www.sma.de/en/products/monitoring-systems/sunny-webbox.html#Downloads-9251) isn't loading correctly on my machine at the moment, but if you have access to a copy check for setting the unit time, also look at the time on your inverter .... don't do this whilst you're generating, wait until after sundown else you'll likely either get a data gap or a really large spike !

Just a guess ...

HTH
Z

Dave_Fowler

I'd agree, I too think it may be a problem with the difference between the reading interval on the inverter and that of the sampling graph. For instance if the inverter sends a reading of the power generated every 2 minutes then in one 15 minute interval there may be seven readings summed together and in the next 15 minutes there would be 8 readings. If you look at the figures, there is a roughly 7:8 ratio of the low to high readings. If you are able, can you ask for readings every 10 or 20 minutes?

Dave F

mickyduck

Dave_Fowler wrote: »

I'd agree, I too think it may be a problem with the difference between the reading interval on the inverter and that of the sampling graph. For instance if the inverter sends a reading of the power generated every 2 minutes then in one 15 minute interval there may be seven readings summed together and in the next 15 minutes there would be 8 readings. If you look at the figures, there is a roughly 7:8 ratio of the low to high readings. If you are able, can you ask for readings every 10 or 20 minutes?

Dave F

The reading interval is only selectable at 15 mins or an hour... so reset to hourly and have a nice smooth curve, thanks for the responses.