Solar expected max values too high?

I am currently running CumulusMX in parallel with WD and had noticed that it was giving higher sunshine hours that WD continuously. Initially I wondered if CMX was the issue however having investigated further I believe CMX to be more accurate than WD now.

Looking at the WD expected max values and comparing them with my maxsolar script it seems like WD is creating values an hour ahead of the current time, now WD used to track my max solar script quite well usually only +/- a few w/m² different. Now I do not know when this started as I had not checked up on this in the past few years :astonished: has anyone else noticed that their sunshine hours are too low and had to alter their thresholds? I should add that my WD clock and daylight savings settings are correct and both WD and CMX are using the same weather station sensors.

When I have gathered more data I will probably email Brian for his opinion.

Stuart

I have been doing some data gathering and this latest build (148) is generating expected max solar values higher than I would have expected. i have attached the table which shows the expected max solar for the time from WD in the first column following date and time and the second column is from my max solar script run for my location. You can see that the WD values are significantly higher than from my script. In previous versions of WD these values tracked quite closely. I intend to gather a full 24 hours of data. I will also install version 145 (if I have it!) after this data gather and before I email Brian to see if it makes a difference for me.

2023-05-04,09:00,585 	507
2023-05-04,09:01,588 	510
2023-05-04,09:02,591 	512
2023-05-04,09:03,593 	515
2023-05-04,09:04,596 	517
2023-05-04,09:05,598 	520
2023-05-04,09:06,600 	523
2023-05-04,09:07,603 	525
2023-05-04,09:08,605 	528
2023-05-04,09:09,608 	530
2023-05-04,09:10,610 	533
2023-05-04,09:11,613 	535
2023-05-04,09:12,615 	538
2023-05-04,09:13,618 	540
2023-05-04,09:14,620 	543
2023-05-04,09:15,622 	546
2023-05-04,09:16,625 	548
2023-05-04,09:17,627 	551
2023-05-04,09:18,630 	553
2023-05-04,09:19,632 	556
2023-05-04,09:20,634 	558
2023-05-04,09:21,637 	560
2023-05-04,09:22,639 	563
2023-05-04,09:23,641 	565
2023-05-04,09:24,644 	568
2023-05-04,09:25,646 	570
2023-05-04,09:26,649 	573
2023-05-04,09:27,651 	575
2023-05-04,09:28,653 	578
2023-05-04,09:29,655 	580
2023-05-04,09:30,658 	582
2023-05-04,09:31,660 	585
2023-05-04,09:32,662 	587
2023-05-04,09:33,665 	590
2023-05-04,09:34,667 	592
2023-05-04,09:35,669 	594
2023-05-04,09:36,671 	597
2023-05-04,09:37,674 	599
2023-05-04,09:38,676 	602
2023-05-04,09:39,678 	604
2023-05-04,09:40,680 	606
2023-05-04,09:41,683 	609
2023-05-04,09:42,685 	611
2023-05-04,09:43,687 	613
2023-05-04,09:44,689 	615
2023-05-04,09:45,691 	618
2023-05-04,09:46,694 	620
2023-05-04,09:47,696 	622

What effect is that having on your sunshine hours

Way too low, I normally use 75% as a threshold for sunny but now reduced to 70% which has improved it a bit. I should add that CumulusMX tracks my maxsolar script just like WD used to do and has more realistic sunshine hours using a 75% threshold.

Stuart

1 Like

Mine was 55% to 60% to get something like normal sunshine hours instead of 75% when I tried 147/148 in April as I have max solar line on my graph it is a lot higher than b145 and when I tried the latest b148 on Tuesday evening then max solar line was still too high so put b145 back on

Well having gathered sufficient data I just re-installed V145 and immediately the first value recorded for expected max solar was within 4w/m² of that showing in my script. So there is something wrong with this in later versions. Once I have some more data probably tomorrow I’ll email Brian to get his view.

Stuart

I have this morning emailed Brian with an explanation of what I have observed about the expected max w/m² values in build 148 together with data collected from both build 148 and 145. I will update again as soon as I have any news.

Stuart

Thanks for your time on this Stuart, will be interesting to see what Brian says

Harold

Thanks…

Looking some more at the version history for WD Brian used a new compiler from build 146 on the 13th January 2023 and I’m wondering if this issue actually started then but I have not got anything between 145 and 148 to test with.

Stuart

Willowsford Farm - WD Archive (novawx.dscloud.me)

1 Like

I just downloaded 146 to test. Thanks for the reminder.

Stuart

Well my initial view is that 146 does not create an issue, the maxsolar issue was close to my script value for the first minute of running. I will watch it for a while to check though.

I think now I’ll wait to see what Brain has to say.

Stuart

PS. Just checked and 146 is tracking my script values as expected, so not this build.

It was worth trying, though, to pin it down a bit more :slightly_smiling_face:

Well having had no reply from Brian I decided to try build 148 from 20th May. The current expected max values calculated by WD is still too high and so I have tried various settings on the Solar CP to adjust it to bring it inline to where I believe it should be, it was running about 11 minutes ahead of where it should have been.

It needed a minute offset of around 11 minutes on the’ Max Solar Time of day’ tab. this now runs inline with my maxsolar script and my CumulusMX install. I’ll watch it closely for a while to see how it goes. I suspect the minute offset needed will not be constant as before it was nearly 30 minutes ahead early in May when I checked initially.

Anyway currently my 75% threshold would be OK for sunny. Now I have no idea if Brian has actually changed anything in this latest update as there is no mention in the history, so I need to wait and see how it goes.

Stuart

I’ve been looking at my captured solar data this morning and it looks at present that my 11 minute offset is making WD track quite well to what I expect the max solar for time values should be, it’s about +/-5 w/m² which is negligible. So far so good. But does not explain why this is needed now in WD as I have never used it previously.

Stuart

I have the 1st of May build of 148 and that appears to have got nearer to the normal as the month has gone on beginning of June will be interesting because it appeared to be worse at the beginning of the month, but I will update to latest build during the week anyway

I was just checking my sunshine hours and solar max again and now find my -11 minute offset is causing the expected max w/m² value to be much too low now. There is something awry with the expected max w/m² calculations . It should not be necessary to alter the settings like this. My values are not tracking anywhere near other sources I can find for these values.

It is only fairly recently that this has changed. I tried quite a while ago emailing Brian but he has not replied.

Stuart

I take it still no word from Brian on this

No to be honest I’ve given up on him. I think he ignores me. Please try yourself if you agree these values are incorrect and see if he replies. I am willing to talk with him if he wants to communicate with me!

Stuart

I agree that something is off, but my situation is backwards. We had clear skies yesterday and WD is showing max values much lower than what my sensor reads. Value is always at 100%+, where typically I run around 92% on a clear day. I’m in the US, PA at 40.3, -76.0.

I found that v10.37Sb147 from Feb 3rd works correctly, but lacks some fixes for WLL history.
I was using build 149 from July and tried the latest EXE and they do not work.

Example for 11:10 AM EDT today with some high clouds:
b147: actual - 650, WD max: 773 (85%)
b149: actual - 650, WD max: 635 (102%)

I’m currently running a hybrid install with b147 EXE and with the other components that were updated after Feb 3, but the WLL history extraction is linked to the EXE so that is still broke, but not an issue.