pv output

[poheZ]

HI

What do people use to send data (automatically) to pvoutput please?

 

[Scobo]

I'm using a Current Cost Envir connected to a netbook PC via USB and pvbeancounter software to upload the data every 5 minutes to pvoutput.

 

[FB.]

I'm very sceptical of the various devices. A lot of people are reporting numbers which are too good to be true - although I'm sure they like to feel all warm and fuzzy - patting themselves and their installer on the back for skill and ingenuity if their system appears to be well ahead of the rest.

Some have since realised that the sum of all their uploaded readings far exceeds what the generation meter says. The university of Sheffield solar team ask for data from the generation meter - probably for a good reason.

I take readings from my generation meter and type it in by hand. It looks much less scientific without the various other data uplodas, but the generation meter is the gold standard.

Once I have completed a year's worth of data I will stop adding to PVoutput because I no longer believe what many people are uploading and I no longer believe it to be a reliable data source. I will be using University of Sheffield generation-meter-derived data in future and will continue to contribute to their database.

 

[Gary K]

I am also doubtful of some of the outputs claimed, and also input my output manually . However I will continue to input to pvop as it is usefull to have all the data stored for when the inverter dies and all info is lost

 

[Scobo]

Agree that it's not 100% accurate by any means but it a very useful guide to how well your system is doing.
I find the total output tends be out by 1 or 2 kwh per week and adjust it weekly to match our meter.

 

[FB.]

Agree that it's not 100% accurate by any means but it a very useful guide to how well your system is doing.
I find the total output tends be out by 1 or 2 kwh per week and adjust it weekly to match our meter.

The thing is; if we both have 2.5kW systems, with SAP targets of (say) 2000kWh per year, if you don't correct your extra "1 or 2 kWh per week", your system appears to be perhaps 100kWh superior per year (5%). Voila - you've "shown" that your xxxxx brand of panels and xxxx inverter, installed by --- solar co.ltd are 5% superior, which is a significant and very worthwhile extra performance - if it was true.
You adjust your figures periodically to bring them back into line, but I expect the majority of people don't - and they wouldn't want to, if their system was no longer top of the performance league in their area.
Some systems in mine or neighbouring counties are 25% above the second-best-performer in their area, and 35% above the median - even with the same brand of panel, inverter, azimuth, pitch etc. There is no way that a system can be that much different from similar well-installed arrays nearby.

So I'll be leaving PVOutput to those with automated uploads to compete in a land of "put-in-what-you-like" solar readings, so that those of us who input factual data can be compared more scientifically. After all: the more accurate the data, the better the conclusions which can be drawn from it.

 

[Worcester]

If you want an inexpensive system and have an SMA inverter, add in the blutooth piggyback, if bluetooth not already in there, then get a good bluetooth dongle and hook it all up for free via your PC to PVOutput.
OR:Build one yourself: openenergymonitor.org - you need an emonTX, an emonBase, and if you want a display emonGLCD. then display it for free on emoncms.org or pvoutput.org.

BOTH the above solutions are way more accurate than anything else we've come across on the market.

Note the emonTX can be configured for true power/energy by adding in a voltage and phase detection open which the simple CT loop detectors can never compensate for.

 

[wstevew]

I use my Android App. "Sunny Solar Log" this collects data direct from the inverter.

While I accept that a generation meter is more accurate, I can say that over almost 11 months 3122kWh the difference between the inverter total and the meter is around 1.5 kWh which can probably be put down to losses between the inverter and meter. Especially as I had the AC cable replaced with 6mm in August because they installed with 2.5mm TWE originally.

 

[Worcester]

@wstevew - I think your app picks it up direct from the inverter doesn't it?

So it will certainly be more accurate than any of the 'simple' CT based energy monitors (CC, Owl, Wattson etc)

The openergymonitor.org emontx can also use an interface to read the flashing led on the meter if you want 100% accuracy (if the meters are 100% accurate )

 

[FB.]

Yes, inverter stats seem to be more reliable. My inverter (PowerOne Aurora PVI-3600-UK indoor model) reports lifetime generation of 0.5% more than the generation meter.
The difference is probably a 0.5% loss in the few metres of cable from the inverter to the generation meter.

 

[Scobo]

Forgot to add that I use an Optismart which reads the meter's LED flashes paired with the Envir display and yes it's a lot more accurate than a CT clamp.

 

[wstevew]

@wstevew - I think your app picks it up direct from the inverter doesn't it?

Yes via Bluetooth same method as Sunny Explorer uses.

 

[Gnarly]

I use smatool on Linux, look for sma-bluetooth to find the repository

It runs from a 'cron' job every hour, seems reliable, and my sunny boy 3600TL built-in meter tracks the generation meter very closely.
Not too difficult to set up although you do have to build it and issue a few runes to mysql to create the database.

 

[finewine]

I use the brilliant Sunny Solar Log which just does it automatically for me. Great app for my HTC. If interested I am called 3 highfields pv output