Comparing Production Estimates to Actual Production

[control4userguy]

The point I was trying to make was I was not satisfied with the pre-set calibration coefficient they use for the 20A CT which is what is needed for some solar installs. It is quite a deal to certify under ANSI 12.1 or 12.20. The device is now recognized or qualified for use beyond, say, non-reference measurement. You, as a consumer, would automatically get the 1% accy. unit. With that, you would select anything other than the "revenue grade" CTs.

Someone will have to inform me as to how P3 claims their device is 0.2% accuracy. I can understand not wanting to go for the ANSI rating for a $20 retail device...

[bcroe]

The point I was trying to make was I was not satisfied with the pre-set calibration coefficient they use for the 20A CT which is what is needed for some solar installs. It is quite a deal to certify under ANSI 12.1 or 12.20. The device is now recognized or qualified for use beyond, say, non-reference measurement. You, as a consumer, would automatically get the 1% accy. unit. With that, you would select anything other than the "revenue grade" CTs.

Someone will have to inform me as to how P3 claims their device is 0.2% accuracy. I can understand not wanting to go for the ANSI rating for a $20 retail device...

I find it hard to believe a clamp on current transformer can be certified accurate to
0.2%. Each time it is closed, the halves might not align to that accuracy, and any
microscopic dirt at the mating surfaces will have an impact. A single piece transformer
you have to thread the wire through, MAYBE..........

I have adjusted shunts to that level, but if they heat up, the constantine resistance will
drift. Bruce Roe

[control4userguy]

PS does not use an external clamp as such. I have no idea what's inside theirs... You are mostly correct though, clamps generally run between 1% down to 0.3% of reading. My Yokogawa is a 0.3%'er and requires special attention to use properly.

[control4userguy]

Clamps are clamps. If you are serious about I measurement then you want something like a flux gate transducer (0.01% accy).

[bcroe]

Clamps are clamps. If you are serious about I measurement then you want something like a flux gate transducer (0.01% accy).

If that were actually practical, you would still need to thread the wire through it, or
bolt onto a prepositioned wire (like a shunt). Bruce Roe

[control4userguy]

^Not sure where you are going with this. Accuracy, in the consumer sense, is not absolute. Even then, you can calibrate to some other standard and still be "inaccurate". Welcome to the world of statistics.

[control4userguy]

If that were actually practical, you would still need to thread the wire through it, or
bolt onto a prepositioned wire (like a shunt). Bruce Roe

But, just to carry on with this line, shunts are hugely affected by temperature. If the panel is located outdoors, you would never be a assured of a authentic reading from day to day.

[bcroe]

But, just to carry on with this line, shunts are hugely affected by temperature. If the panel is located outdoors, you would never be a assured of a authentic reading from day to day.

Yea, that's what I said. Bruce

[control4userguy]

OK, I'm lost. Yes, you still need to make physical connection to the transducer. Let us back-up here, the CTs included with the eGauge kit are a snap-on affair. Little to no hysteresis change due to a relatively fixed airgap. We don't need gnat's ass repeatability to tell that the solar array is working as it should.

[bcroe]

OK, I'm lost. Yes, you still need to make physical connection to the transducer. Let us back-up here, the CTs included with the eGauge kit are a snap-on affair. Little to no hysteresis change due to a relatively fixed airgap. We don't need gnat's ass repeatability to tell that the solar array is working as it should.

Right again. It wasn't me who brought up flux gate transducers. Bruce Roe

[control4userguy]

^Whatever.

Pleppik- don't forget to update the 3000 firmware.

[pleppik]

February Production vs Estimtes

Here's the data for February:

Actual production vs. Estimate
Solar-Analysis-1.gif
We produced 441 kWh in February, more than double January, and January was more than double December. Yay for longer days and higher sun angle! This was a little more than the installer's original estimate of 411 kWh for February.

Snow Cover
Solar-Analysis-2.gif
I estimate that we lost about 35 kWh in February due to snow cover. We had only a little snow, but there were several days during the month when one of the arrays was covered. Those also tended to be bright sunny days (of course).

Weather
Solar-Analysis-3.gif
According to insolation data from a nearby weather station, February was 7% sunnier than the average over the past few years. So adjusting the installer's estimate upwards by 7%, I find that the estimated production was 439 kWh, almost exactly what we actually produced.

Deviation
Putting all the adjustments in, in February our production was about 8% higher than I would have expected from the installer's estimates. This isn't as extreme a difference as in prior months, but it's still the case that in every singe month our production has been above expectations.
Solar-Analysis-4.gif

[bcroe]

Here's the data for February:

Snow Cover
[ATTACH=CONFIG]5951[/ATTACH]
I estimate that we lost about 35 kWh in February due to snow cover. We had only a little snow, but there were several days during the month when one of the arrays was covered. Those also tended to be bright sunny days (of course).

Yes our snow storms are frequently followed by a great sun day. Which causes me to be out at first light
next day, removing the snow. Feb was about like last year here, more than double Dec or Jan (per day).

Bruce Roe

[pleppik]

March Production vs. Estimates

In March, my solar panels produced 37% more power than what I would have expected from the installer's original estimates after taking into account snow cover and weather. For those keeping track, that's now six months in a row with more production than expected, and four of the six months with at least 30% more than expected.

Actual Production vs. Estimate
In March, we produced 680 kWh, vs the installer's original estimate of 494 kWh.
Solar-Analysis-2.gif

Snow Cover
There were several days in March with at least partial snow cover on one of the arrays. By comparing the production on the array I could brush clean to the production on the array I can't reach to clear, I estimate we lost 33 kWh in production to snow cover.
Solar-Analysis-1.gif

Weather
Using data from a nearby weather station, March was about 6% sunnier this year than in the past several years. To correct for this, I added about 28kWh to the installer's original estimate for March production.
Solar-Analysis-4.gif

Deviation
So putting all that together, we produced 37% more power than what I would have expected based on the weather, snow conditions, and the installer's original estimate.
Solar-Analysis-3.gif

[bcroe]

In March, my solar panels produced 37% more power than what I would have expected from the installer's original estimates after taking into account snow cover and weather. For those keeping track, that's now six months in a row with more production than expected, and four of the six months with at least 30% more than expected.

The winter before had way more snowfalls than this one. Maybe the installer's estimate
was based on that sort of thing. Was he just as far off in summer? Bruce Roe