What's the update interval for SolarEdge monitoring?

[jasonvr]

Well, I certainly like "easy and cheap"! That Rainforest Eagle solution looks very interesting. Not having to open the service panel to attach CTs is a big plus for me. I haven't had any luck finding an original model Eagle. Amazon no longer has it in stock. It's probably because the new Eagle-200 is due to come out later this summer. Hopefully, SCE is willing to provision it as soon as it's released.
Is it still best to "buffer" the data through Wattvision first, or should I simply add the Eagle directly to PVOutput as a Secondary Device? I've read that by going through Wattvision, the data samples are averaged which results in more accurate consumption measurements overall. I wonder if this is still true; but since I like to "experiment", I'll probably try it both ways.

I tried going directly from Rainforest and I wasn't too happy with it. Went back to Wattvision

Interesting about the Eagle-200. Looks like the big changes are adding Wifi instead of just hardwired Ethernet, as well as what appears to be a Zigbee hub, though I'm not 100% sure on that. The Wifi would be useful to me so that I can move the Eagle onto my guest Wifi network along with most of my other IoT devices that are inherently insecure.

[Woodworkerii]

...I'm getting my consumption information by using the net calculations on pvoutput and combining my SolarEdge data with the data from my Rainforest Eagle Zigbee monitor. Unless you're looking for something revenue grade, it's a pretty easy and cheap way to go. It links wirelessly with your smart meter and reports the net number on the meter. pvoutput is smart enough to combine that with the SE production numbers and figure out the consumption side of things with reasonable accuracy...

Well, I certainly like "easy and cheap"! That Rainforest Eagle solution looks very interesting. Not having to open the service panel to attach CTs is a big plus for me. I haven't had any luck finding an original model Eagle. Amazon no longer has it in stock. It's probably because the new Eagle-200 is due to come out later this summer. Hopefully, SCE is willing to provision it as soon as it's released.
Is it still best to "buffer" the data through Wattvision first, or should I simply add the Eagle directly to PVOutput as a Secondary Device? I've read that by going through Wattvision, the data samples are averaged which results in more accurate consumption measurements overall. I wonder if this is still true; but since I like to "experiment", I'll probably try it both ways.

[JSchnee21]

Thank you everyone, especially @Woodworkerii for the screen shots, I think I finally have PVOutput setup correctly (fingers crossed). For those of you who might be a PVOutput Noob (like me): A few high level steps to get you started:

1) Obviously create a PVOutput account and setup your system
2) Perhaps not as obvious, while making a Donation to PVOutput is OPTIONAL, many/most of the really cool features require a small annual donation -- so do it! You'll need PayPal, so get this setup first
3) For weather data, you'll need a Weather Underground account AND and WU developer API key. Both of which are free for personal low volume use. But, as a die hard fan of WU, and a PSU alumni, I had to give them some money as I use their website / apps / services everyday.
4) Thank you so much for the extended data config info and syntax of the C=>F conversion command -- but it still took me a few minutes to realize you could flip through the different graphs by clicking on the little squares (-:
5) I had to scout around a bit for a local weather station that had a solar radiation meter -- WU makes these somewhat arduous to find.
6) I think I finally got both my production and consumption data setup correctly in PVOutput. I was hoping using "BOTH" would work. But it didn't seem to (Production stopped logging). I had to set my primary SE as "Solar" and setup and secondary SE (same site and API key) as Consumption. This seems to have fixed the issue.

Now all I need to do is figure out how to get my installer to enable the SE Monitoring Portal "Charts" and "Reports" features so I can graph per panel voltages, etc.

-Jonathan

http://pvoutput.org/list.jsp?userid=58372

[jasonvr]

Once I get it all "dialed in", I'll probably look into adding hardware to monitor the consumption side. This solar project started out as a "smart investment" (which it certainly is); but has now turned into a "hobby". Having fun!

I see you're just south of me in Lake Forest which means you are on SCE and should have a Zigbee capable smart meter. I'm getting my consumption information by using the net calculations on pvoutput and combining my SolarEdge data with the data from my Rainforest Eagle Zigbee monitor. Unless you're looking for something revenue grade, it's a pretty easy and cheap way to go. It links wirelessly with your smart meter and reports the net number on the meter. pvoutput is smart enough to combine that with the SE production numbers and figure out the consumption side of things with reasonable accuracy. You do have to contribute to pvoutput for this feature, but it seemed like a good and cheap investment to me (and you needed to contribute to get the higher interval from SE to pvoutput anyways, so you're probably already capable of the net calc).

[jasonvr]

Another small change to consider in your settings... The "summary" column in the extended data is set to "last". If you set that to "change", when you look at your data in the daily view instead of the intraday view, I think it will display the range for the day instead of just the last value.

So I just tried this out on my system and unfortunately it doesn't work like that. According to the help info, "Change" is "The last value minus the first value of the day". Sound like it basically shows you the temperature change from beginning of day to end of day (i.e. 00:00 to 23:59) which doesn't seem very useful other than showing if you have a general warming or cooler trend day over day. Seems like Highest or Average would be best for temperature.

[Woodworkerii]

I get Fahrenheit by adding a rule in the "Extended Data Rule" section:
v7 = 1.8 * v7 + 32;

It would be v9 instead based on your setup

jasonvr, the temp conversion formula did the trick. Thanks for clearing that up. I think I tried a variation of it, but I didn't get the syntax right. I probably wouldn't have gotten it to that format because (to my non-engineering brain) it looks like it would cause the calculation to run an endless loop. In fact, I wasn't even sure if that was the right field to enter the formula. It all seems to be working well now.

Another small change to consider in your settings... The "summary" column in the extended data is set to "last". If you set that to "change", when you look at your data in the daily view instead of the intraday view, I think it will display the range for the day instead of just the last value.

sensij, I set the summary column to "change" as you suggested. Because I only have 3-days of data collected, I'm still watching to see the effect of various changes to the settings. I guess I'll still be "experimenting" for a bit. Once I get it all "dialed in", I'll probably look into adding hardware to monitor the consumption side. This solar project started out as a "smart investment" (which it certainly is); but has now turned into a "hobby". Having fun!

[J.P.M.]

How do you know if your system is able to report panel temperatures? I have Jinko Solar panels with a SolarEdge inverter.

Most do not. Additionally, I'd be skeptical of those that do. Most sensor attachment methods are such that they are usually not a very good indicator of cell temp. Also, I've measured cell /panel temp. variations over my array of about something like 2-5 C. depending on P.O.A. and wind vector, making more than a few sensors necessary. FWIW, I can usually and easily find even adjacent cell temp. variation of a deg. C. or so.

Also, keep in mind that panel temps., average or otherwise, are not ambient temps.. and the GHI (Global Horizontal Irradiance (or insolation)), needs to be converted to P.O.A. (Plane of Array) irradiance or insolation if information on efficiencies or efficiency comparisons or other parametric information is sought.

[sensij]

How do you know if your system is able to report panel temperatures? I have Jinko Solar panels with a SolarEdge inverter.

There are no commercial systems I am aware of that attempt to measure or report panel temperature. Only home-brew instrumentation for those with the time and money to burn. Solaredge will tell you inverter temperature.

[macaddict]

If cell temperature (or at least the backsheet surface temperature of a PV panel) is measured by a device that has the ability to communicate it at some intraday interval, it can be sent directly to PVoutput. You are aware of the complications involved in getting that measurement accurately and reliably, so I won't go into that here.

How do you know if your system is able to report panel temperatures? I have Jinko Solar panels with a SolarEdge inverter.

[J.P.M.]

PVoutput can be linked to a Wunderground site and pull the data available. Usually, that is ambient temperature at that station and irradiance measured by a horizonal sensor, if the station is equipped for that.

If cell temperature (or at least the backsheet surface temperature of a PV panel) is measured by a device that has the ability to communicate it at some intraday interval, it can be sent directly to PVoutput. You are aware of the complications involved in getting that measurement accurately and reliably, so I won't go into that here.

P.O.A. irradiance is not produced or provided by PVOutput, although they have a rough "insolation" function that attempts to show what clear sky output on that day might look like for the location and array orientation entered by the user. I was never successful in figuring out what particular model they use, but it clearly didn't handle temperature correctly when I last looked into it.

PVoutput allows some basic function writing using the uploaded data, as in the C to F conversion described a few posts ago. It doesn't really support the more complicated equations necessary to combine time, location, array orientation, ambient temperature, and wind speed to implement models like those used in PVWatts or SAM, or the extensions to those models like what you have worked on that are even more tuned to your system.

When I get my weather station active again, I would like to use the CEC model that I translated in python over the winter to see how it compares to my actual conditions and live output. I would still rely on the NOCT model of cell temperature, at least at first, until I can start experimenting with small footprint sensors that could mount directly to the panels. I would be able to upload the model output to the extended values in PVOutput, along with any intermediate calculations (like P.O.A. irradiance) that are useful to see.

Probably not the answer you were looking for, but I hope that providing some extra context for those unfamiliar with the terms you were asking about is ok.

Thank you. I got what I asked for and that's to say in a positive way. Maybe more comment later, or maybe a thread discussing what may be some ways to use or some limitations to the PVOutput "output", might be useful.

I appreciate the response.

[sensij]

Sensij: This is a bit off topic, but can you provide some info to help offset my ignorance with respect to where/how PVOutput gets its cell temp. and P.O.A. irradiance values ?

Thanx.

PVoutput can be linked to a Wunderground site and pull the data available. Usually, that is ambient temperature at that station and irradiance measured by a horizonal sensor, if the station is equipped for that.

If cell temperature (or at least the backsheet surface temperature of a PV panel) is measured by a device that has the ability to communicate it at some intraday interval, it can be sent directly to PVoutput. You are aware of the complications involved in getting that measurement accurately and reliably, so I won't go into that here.

P.O.A. irradiance is not produced or provided by PVOutput, although they have a rough "insolation" function that attempts to show what clear sky output on that day might look like for the location and array orientation entered by the user. I was never successful in figuring out what particular model they use, but it clearly didn't handle temperature correctly when I last looked into it.

PVoutput allows some basic function writing using the uploaded data, as in the C to F conversion described a few posts ago. It doesn't really support the more complicated equations necessary to combine time, location, array orientation, ambient temperature, and wind speed to implement models like those used in PVWatts or SAM, or the extensions to those models like what you have worked on that are even more tuned to your system.

When I get my weather station active again, I would like to use the CEC model that I translated in python over the winter to see how it compares to my actual conditions and live output. I would still rely on the NOCT model of cell temperature, at least at first, until I can start experimenting with small footprint sensors that could mount directly to the panels. I would be able to upload the model output to the extended values in PVOutput, along with any intermediate calculations (like P.O.A. irradiance) that are useful to see.

Probably not the answer you were looking for, but I hope that providing some extra context for those unfamiliar with the terms you were asking about is ok.

[J.P.M.]

Another small change to consider in your settings... The "summary" column in the extended data is set to "last". If you set that to "change", when you look at your data in the daily view instead of the intraday view, I think it will display the range for the day instead of just the last value.

Sensij: This is a bit off topic, but can you provide some info to help offset my ignorance with respect to where/how PVOutput gets its cell temp. and P.O.A. irradiance values ?

Thanx.

[sensij]

Another small change to consider in your settings... The "summary" column in the extended data is set to "last". If you set that to "change", when you look at your data in the daily view instead of the intraday view, I think it will display the range for the day instead of just the last value.

[jasonvr]

I get Fahrenheit by adding a rule in the "Extended Data Rule" section:
v7 = 1.8 * v7 + 32;

It would be v9 instead based on your setup

[Woodworkerii]

I have both temperatures enabled. There is an option to push the Wunderground temperature to an extended value instead of the standard location (as well as solar radiation). You can then plot ambient temperature on the extended graph. I also added a formula because it comes in Celsius and I want it in Fahrenheit.

Our new PV system has been online since mid-May. I'm still learning, but with the help of this forum, I was able to switch the inverter's communication method from cellular to Ethernet. A big improvement, but I found that the SolarEdge web browser & iOS app monitoring solutions (privileges as set by my installer) are quite limited. So after reading through some of the past discussions here, I finally got up and running with PVOutput. Donated, and now get 5-min data resolution. Polling interval is set at 15-min to stay within the SE's limit on queries (not to mention the MANY times I hit them throughout the day with my phone app). Using PVOutput, I now have system output, inverter temperature in degrees Fahrenheit, and voltage on the "main" graph/table. On the "extended" data screen v8 and v9, I have Solar Radiation and outside air temperature (API linked from a local/neighbor's station via WUnderground). I hope I'm correct in assuming that the unit for Solar Radiation is W/m^2.
Here's the problem: I set the outside air temperature to report in deg-F from WU, but PVOutput still displays it as deg-Celsius. If I send the WU temperature data to the MAIN graph/table, it correctly displays the temperature in deg-F. There doesn't seem to to be an option on PVOutput to send the inverter's temperature data to the extended graph. jasonvr mentioned that he applied a formula (deg-C * 9/5 + 32?) somewhere to get the extended graph to plot in deg-F. Where/how do I inject this formula? Do my settings look OK?