Any tips on my setup?

[sensij]

'On average', gotcha. So if I travel to an off grid site knowing the sky will be clear for the days required of me I should expect better output.

This interpretation of the word "average" is part of what made my post yesterday objectionable. No, your understanding and expectations are not aligned with what the model is showing.

respectfully, if you have questions about pvwatts, it might be to start a new thread. This one has enough confusion in it already.

[-TX-]

'On average', gotcha. So if I travel to an off grid site knowing the sky will be clear for the days required of me I should expect better output.

[SunEagle]

Because while the sun may be up for 10 hours a day your panels are only producing for 7.03 hours a day on average. A solar pv panel needs to have the sun light hitting it more than just at a glancing angle to produce close to it's nameplate wattage. If you look at a graph you will see a "hump" as the sun comes up with the peak around Noon. If the panel produced full output as soon as the sun hit it the graph would look like a rectangle.

[sensij]

The solar radiation column is displayed in units of insolation. Many people are more familiar with the term "sun hours", but the values are the same. The insolation isn't affected by array size. Read the help files for more information.

For an off grid application, losses will be more than 14%, the AC energy it suggests will be available using the default loss factor is way too optimistic.

[J.P.M.]

Yeah, I don't understand the math. How does 1.9kWh (PV) minus 14% inefficiencies equal 7 kWh of radiation in July, for example, when we get sun for 10 hours a day?

Seems very conservative. Which is good in a sense, but still....

kWh.png

FWIW, see my 03/07,1:08 P.M. comment.

[-TX-]

Yeah, I don't understand the math. How does 1.9kWh (PV) minus 14% inefficiencies equal 7 kWh of radiation in July, for example, when we get sun for 10 hours a day?

Seems very conservative. Which is good in a sense, but still....

kWh.png

[sensij]

I think you would be better using all 3 panels, even if you stay at 12 V. Don't let the 520 W bogey scare you... look closely at the Page 10 Sunking linked. 520 W is the max charge power... in other words, that is the 40 A output spec translated to the input charge power, accounting for the efficiency of the CC. Even if your panels would be capable of producing more than that, the mppt system will pull the operating point away from the maximum, to stay within the 40 A output limit (as described in the preceding paragraph on page 9). In that same table on page 10, you can see that the max installed PV array power for the 12 V setup is listed as 1560 W, which is simply 3X the array charging power rating, although Voc and Isc limits might come into play as well as you approach that maximum.

By increasing the your array size, you are increasing the number of hours the batteries can be charged at 40 A. Your batteries will thank you later.

[Wrybread]

> However, if the Voc is really that high, forget what I said about needing a 4th panel. You could support a 24 V battery with the three panels in parallel, more efficient overall and less expensive wiring.

Converting the battery bank to 24 volts is a really good idea and gracefully handles a lot of the issues. It's easy enough to get a 24v inverter, but the rest of my RV runs on 12 volt, so I'd probably just use a step down transformer, which is easy enough. Only one thing pulls a lot of power, which is the furnace blower (80 watts!), but I could easily do without that since I mostly use a portable Buddy Heater since it's so much more efficient.

I think my plan of attack will be to connect the 2 panels in parallel for now. They're flat mounted and unlikely to exceed 520 watts, though its definitely something to watch for. I'd love to keep the 3rd panel up there so I can switch it on during cloudy periods (and to piss off SunKing), so if there's enough room I'll probably do that too.

And if it looks like that system won't work, I can convert the battery bank to 24 volts and go from there, or consider upgrading the charge controller.

[sensij]

Voc varies with temperature. You'll see a temperature coefficient on the panel's datasheet... typically around -0.3%/K, or -0.14 V/K. If the panels are 46 Voc at STC (25 deg C), it needs to be adjusted for the minimum temperature they'll see. Minimum design temp for Stockton is -3 deg C (see link below), so that would be a 3.92 V increase, raising the 46 Voc panel to 49.92 V. Two panels in series just barely clears the 100 Voc limit for the Tracer charge controller.

http://www.solarabcs.org/about/publi...map/index.html

However, if the Voc is really that high, forget what I said about needing a 4th panel. You could support a 24 V battery with the three panels in parallel, more efficient overall and less expensive wiring.

[Wrybread]

Two 46 volt solar panels in series is 92 volts. Are you seriously saying that's not the case?

But anyway, as I said about 10 times, I'm wiring them in parallel.

> As for shade, no problem, you have NO POWER to worry about.

Sorry to be the one to have to tell you this, but you're not making any sense.

[Sunking]

Where are you getting your info from? Here's the manual:

From the manual, page 10. You might try using the USA Model, not New Zealand's.

http://www.aasolar.co.nz/Instruction...A%20Series.pdf

It clearly says the maximum Voc input is 100 volts. But even if it was 92 volts, that's still perfect for two panels in series (each of which is 46 volts open circuit).

Wrong. does not matter if the Controller Voc is 92 or 100 volts, you cannot use two panels in series. The panel spec is at 75 degrees. You have not applied any temperature correction factor of 1.25. That means if the panel spec Voc = 46 volts is adjusted to 46 volts x 1.25 = 57.5 volts.

As for shade, no problem, you have NO POWER to worry about.

[J.P.M.]

OK, more formally, it is approximating the hourly output by using the angle of incidence based on he solar position at the midpoint of each hour (xx:30),combined with irradiance data in the selected TMY file. The point being, the hourly production it models will not reflect real world observed values over shorter periods of time, even if the environmental and equipment assumptions are perfectly matched to real world conditions.

Close enough. Thank you for the correction.

[J.P.M.]

Thanks Sensij, great info. And man, tough crowd on this website. An odd amount of bickering over semantics.

It may seem like semantics to the uninformed, but it isn't. It's the way the model works.

Saying it calculates averages is simply incorrect, as a quick read of the help screens will enlighten. Saying so may lead others who are unfamiliar with what they're doing to incorrect assumptions/conclusions, and result in perhaps costly errors based on the incorrect information.

Take what you want of the above. Scrap the rest.

[Wrybread]

Thanks Sensij, great info. And man, tough crowd on this website. An odd amount of bickering over semantics.

[sensij]

PVWatts does not calculate hourly averages.

OK, more formally, it is approximating the hourly output by using the angle of incidence based on he solar position at the midpoint of each hour (xx:30),combined with irradiance data in the selected TMY file. The point being, the hourly production it models will not reflect real world observed values over shorter periods of time, even if the environmental and equipment assumptions are perfectly matched to real world conditions.