Voltage calcs, would you consider the +/- 10%

[acrwc10]

The cut sheet on the modules says the VOC can be plus or minus 10%, when calculating string voltage should I consider the possible 10% increase so as not to exceed the inverter max voltage input of 600 volts?

 

[GoldDigger]

My guess is that the voltage range is meant to cover the voltage change with panel temperature rather than to cover product variations at STC.
As long as you do your own temperature correction calculation for max voltage in winter and starting voltage in summer, I would start with the nominal value.
(That assumes a silicon panel. I cannot say authoritatively about other types.)

 

[ggunn]

The cut sheet on the modules says the VOC can be plus or minus 10%, when calculating string voltage should I consider the possible 10% increase so as not to exceed the inverter max voltage input of 600 volts?

I have never seen a module spec'ed that way. All that I have seen have a Voc number and a beta number for calculating corrected voltage for temperature.

 

[Engineer#4827502]

Is the +/- 10% the tolerance? I've seen flash test data where the VOC was somewhere more in the range of +7%/-3% (power was +5%/-0% but voltage and current individually swung further). +/- 10% would be a pretty big tolerance.

 

[jaggedben]

If the module doesn't have a published temperature co-efficient then you should use table 690.7 to calculate maximum voltage, assuming these are silicon panels.

 

[acrwc10]

These are "REC" modules, the +/- 10% is a range given on the data sheet. Not the temp coefficient, that is also listed and not an issue.

 

[jaggedben]

I'd say it's not required by code.

If you're worried about damaging the inverter, I'd say it's a judgement call, but personally I probably wouldn't worry about it. I'm fairly confident that most brand-name inverters can withstand a voltage 10% higher than allowed without damage, and the max voltage calculation is supposed to be for those occasions which may never actually happen. I.e. record low temp that happens when the sun is up instead of in the middle of the night. Also, it would have to be the case that all your modules would be 10% over for the calculation to be relevant, and what is the likelihood of that? :lol:

I've also seen people commit mistakes that exposed inverters to 800-900V for a short time until someone said "why isn't it turning on?" and fixed it. And I've yet to see an inverter get damaged that way.

 

[Zee]

JaggedBEn, I agree with your excellent points.

One little hair I'd like to split:
I, also, used to say "well it isn't ever gonna be sunny (=high V) and be record cold simultaneously".

In fact, record coldest temps are reached just before dawn.
(It gets colder and colder all night.)
And you know the rest: PV voltage is not proportional to sunlite intensity. Even a little light will trigger max VOC on PV modules.
The third factor is that at dawn the Voc will exist as opposed to the lower Vmp as the inverter has not turned on yet as it is below its power threshold.
So, actually record cold and MAX voc can happen daily...at dawn in winter.

 

[GoldDigger]

Not to mention that on a clear night a panel can reach a temperature lower than the air temperature. The same lack of shade that lets them work well all day also allows them to cool well by radiating into the night sky.
Good air circulation under the panels will reduce this effect.
Mounting flat close to an insulated roof will maximize it.

 

[ggunn]

I've also seen people commit mistakes that exposed inverters to 800-900V for a short time until someone said "why isn't it turning on?" and fixed it. And I've yet to see an inverter get damaged that way.

Even so, there are some inverters that have a "black box" recorder that keep a record of the highest voltage the inverter has ever seen, and if it is presented to the manufacturer for warranty work and has ever been exposed to voltage higher than the published maximum the inverter can tolerate, they may not honor the warranty.

 

[BillK-AZ]

PV module Voc does vary with irradiance, but not a lot:

The above is from a 2009 REC data sheet and is typical.

A PV module data sheet on the REC Solar website (www.recgroup.com) does show a tolerance that is missing on many data sheets:
"Analysed data demonstrates that 99.7% of panels produced have current and voltage tolerance of ?3% from nominal values."

 

[SolarPro]

The cut sheet on the modules says the VOC can be plus or minus 10%, when calculating string voltage should I consider the possible 10% increase so as not to exceed the inverter max voltage input of 600 volts?

No. Use the published Voc value, which is an average value for that bin of modules. Within the bin, there is some variation in Voc values. (In this case, the Voc can vary by ?10%.) But once you start putting multiple modules in series, the Voc values in aggregate will trend toward the average.

 

[acrwc10]

These have all been interesting opinions but, no one has addressed the question from a code requirement approach. Do you really think the comments "it should never be an issue because I have never seen it happen." will hold up to an inspector?:blink:

 

[acrwc10]

PV module Voc does vary with irradiance, but not a lot:
View attachment 11398
The above is from a 2009 REC data sheet and is typical.

A PV module data sheet on the REC Solar website (www.recgroup.com) does show a tolerance that is missing on many data sheets:
"Analysed data demonstrates that 99.7% of panels produced have current and voltage tolerance of ?3% from nominal values."

Then why does there cut sheet say the Voc can be plus or minus 10%, seems like they have an internal conflict.

 

[jaggedben]

These have all been interesting opinions but, no one has addressed the question from a code requirement approach. Do you really think the comments "it should never be an issue because I have never seen it happen." will hold up to an inspector?:blink:

I think we have addressed the code requirements. The code proscribes two methods for calculating max voltage, one if there is a known temp coefficient, and one if there isn't. It does not require accounting for a tolerance figure.

Then why does there cut sheet say the Voc can be plus or minus 10%, seems like they have an internal conflict.

Maybe for the 0.3% that don't fall within 3%?

 

[SolarPro]

Then why does there cut sheet say the Voc can be plus or minus 10%, seems like they have an internal conflict.

Not really. All module manufacturers publish a power tolerance range for specific products. For example, REC modules have a positive only power tolerance of 0 to +5 watts (not %). So these are very tight power bins.

Some manufacturers, like REC, provide additional detail beyond power tolerance. In this case, they have stated that the extreme limit of the Voc range is ?10% for a specific module model. They have also stated that the vast majority of modules within a specific bin have a Voc of ?3% of the value on the data sheet. The latter makes sense, given the tight binning.

I suspect that manufacturers publish additional details like Voc tolerance simply due to the increased use of I-V curve tracers in the field. If someone measures 10,000 I-V curve traces in the field, the manufacturer likely doesn't want to get a phone call w/ a warranty claim just because 30 of these modules show a Voc of -10% of the published value.

What manufacturers warranty, after all, is power, which is a product of both voltage and current.

 

[ggunn]

I try to make sure that my temperature corrected Voc has a cushion between it and the maximum inverter voltage, anyway. It makes me a tad nervous if the ASHRAE numbers give me 599.9V on a 600V system.