2017 690.7(A) Maximum Voltage

Status
Not open for further replies.
X

Xamacho

Member
Location
Texas
Occupation
Electrical Engineer, PE
Hello all,

For the new options to calculate the Maximum Photovoltaic Source and Output Circuits 690.7(A), can anyone point me in the right direction of what option 3 is stating:

(3) For PV systems with a generating capacity of 100 kW or greater, a documented and stamped PV system design, using an industry standard method and provided by a licensed professional electrical engineer, shall be permitted.

Specifically what does "industry standard method" mean? Does this mean a modeling software?

Thanks,

Michael Camacho, P.E.
 
Xamacho said:
Hello all,

For the new options to calculate the Maximum Photovoltaic Source and Output Circuits 690.7(A), can anyone point me in the right direction of what option 3 is stating:

(3) For PV systems with a generating capacity of 100 kW or greater, a documented and stamped PV system design, using an industry standard method and provided by a licensed professional electrical engineer, shall be permitted.

Specifically what does "industry standard method" mean? Does this mean a modeling software?

Thanks,

Michael Camacho, P.E.
Click to expand...


I think you will find that this is a more practical issue when it comes to maximum continuous current, rather than voltage. The industry standard method for maximum voltage is already described in the NEC, by adjusting Voc to the ASHRAE low temperature, with the Voc temperature coefficient. This will depend only on the module and the site location's climate. The 6 am temperature will be experienced by any module at a given site, regardless of design decisions like orientation or equipment selection. The cold temperature Voc algorithm is conservative, because as soon as Voc tries to cause any current flow, the modules will heat up and the voltage will drop as current rises.

Imax by contrast, used to have 1.25*Isc as the only way to calculate it. But this depends on a lot more design factors, such as orientation and irradiance available to the modules. By allowing us to deviate from a one-size-fits-all rule, it allows us to use a more ambitious design for DC circuit sizing that more accurately reflects the current that there will be in reality.
 
Thank you for your responses.

The reason I was looking at Max Voltage (vs Current and Conductor sizing) was due to a big string count in a cold environment (-15 degrees Celsius). When we put 19 modules in a string, the total Voc(@-15) was 1015V, putting it above the 1000V designed system, and above the inverter max voltage. I was curious if this new calculation rule would get us below the 1000 volts. However, because this site is still using 2014 NEC, I changed the string count to 18.


I just received the 2017 Mike Holt Solar Photovoltaic Systems, and it states that one standard is from Sandia National Laboratory. Publication SAND 2004-3535 Photovoltaic Array Performance Model. It is free and I just downloaded it and will read it later on today.
 
Xamacho said:
Thank you for your responses.

The reason I was looking at Max Voltage (vs Current and Conductor sizing) was due to a big string count in a cold environment (-15 degrees Celsius). When we put 19 modules in a string, the total Voc(@-15) was 1015V, putting it above the 1000V designed system, and above the inverter max voltage. I was curious if this new calculation rule would get us below the 1000 volts. However, because this site is still using 2014 NEC, I changed the string count to 18.


I just received the 2017 Mike Holt Solar Photovoltaic Systems, and it states that one standard is from Sandia National Laboratory. Publication SAND 2004-3535 Photovoltaic Array Performance Model. It is free and I just downloaded it and will read it later on today.
Click to expand...

But realize that these models are just predictive estimates. Many/most inverters have a register where they store the maximum DC voltage they have seen. If they ever record voltage in excess of the manufacturer's published maximum the warranty on the inverter is in jeopardy.
 
Xamacho said:
Thank you for your responses.

The reason I was looking at Max Voltage (vs Current and Conductor sizing) was due to a big string count in a cold environment (-15 degrees Celsius). When we put 19 modules in a string, the total Voc(@-15) was 1015V, putting it above the 1000V designed system, and above the inverter max voltage. I was curious if this new calculation rule would get us below the 1000 volts. However, because this site is still using 2014 NEC, I changed the string count to 18.


I just received the 2017 Mike Holt Solar Photovoltaic Systems, and it states that one standard is from Sandia National Laboratory. Publication SAND 2004-3535 Photovoltaic Array Performance Model. It is free and I just downloaded it and will read it later on today.
Click to expand...

I've seen a lot of jiggering around with what low temperature to use over the years. We are dealing with weather, local conditions, climate change over 30 years, etc. Lot's of factors that should indicate to a designer to be more conservative even if it means shorter strings. The NEC does not provide a prescriptive requirement. So basically it is up to your best judgment as a designer, keeping in mind that if you spike above the upper voltage limit it might cost a lot of money both in repairs and restringing the whole array for a lower string voltage. So choose your temperature wisely. There is almost always pressure from someone on a project to choose a higher temperature to get longer strings, and usually that someone will not suffer if the choice turns out to be wrong.
 
pv_n00b said:
I've seen a lot of jiggering around with what low temperature to use over the years. We are dealing with weather, local conditions, climate change over 30 years, etc. Lot's of factors that should indicate to a designer to be more conservative even if it means shorter strings. The NEC does not provide a prescriptive requirement. So basically it is up to your best judgment as a designer, keeping in mind that if you spike above the upper voltage limit it might cost a lot of money both in repairs and restringing the whole array for a lower string voltage. So choose your temperature wisely. There is almost always pressure from someone on a project to choose a higher temperature to get longer strings, and usually that someone will not suffer if the choice turns out to be wrong.
Click to expand...

Exactly. When I look at the ASHRAE minimum temperatures, as are recommended in the informational note, I laugh, because I can remember the temperature being outside what is indicated in these tables. But if you did design to the lowest temperature on record, like the spit-clinking -40's, you likely will end up with a string size that is out of range for inverter's MPPT operation, which means no string size would be compatible.
 
Carultch said:
Exactly. When I look at the ASHRAE minimum temperatures, as are recommended in the informational note, I laugh, because I can remember the temperature being outside what is indicated in these tables. But if you did design to the lowest temperature on record, like the spit-clinking -40's, you likely will end up with a string size that is out of range for inverter's MPPT operation, which means no string size would be compatible.
Click to expand...

Not so much any more; most modern inverters have a much wider operating range than earlier ones. I remember an SMA TL system I designed several years ago for which there was one and only one string length that worked, and even then we had to make sure that it got inspected in the morning. If we tried to start it up in the afternoon when the modules were hot it wouldn't generate enough voltage to boot the inverter. The system would keep running all day if it was started in the morning, but if there was a power glitch in the middle of the day it would not come on again until late in the day or the next morning.
 
pv_n00b said:
I've seen a lot of jiggering around with what low temperature to use over the years. We are dealing with weather, local conditions, climate change over 30 years, etc. Lot's of factors that should indicate to a designer to be more conservative even if it means shorter strings. The NEC does not provide a prescriptive requirement. So basically it is up to your best judgment as a designer, keeping in mind that if you spike above the upper voltage limit it might cost a lot of money both in repairs and restringing the whole array for a lower string voltage. So choose your temperature wisely. There is almost always pressure from someone on a project to choose a higher temperature to get longer strings, and usually that someone will not suffer if the choice turns out to be wrong.
Click to expand...

And for the other point of view, I find Designers to be overly conservative with VOC. Irradiance will be quite low when these low temps occur. I dont know why everyone will count every volt from a 100 year cold event, but not count the volts gained from low irradiance. Go measure voltage some really cold morning and tell me what it is compared to the standard computed value - a few percent is all you need to get that extra 20 volts you were fretting over to get that extra module in. furthermore it is highly unlikely an inverter will smoke at +20,+40 volts even if it happens at all. Even if an inverter does have a warranty issue, and even if the inverter records the voltage, and even if the voltage has been slightly exceeded, the manufacturer still may not automatically disqualify the warranty. Even if they do, I imagine the labor and hassle are more of a concern than the actual repair/replacement cost....Maybe I am biased as I have been installing these systems in the 35 degree rain in mud 1 foot deep :rant: :D
 
electrofelon said:
And for the other point of view, I find Designers to be overly conservative with VOC. Irradiance will be quite low when these low temps occur. I dont know why everyone will count every volt from a 100 year cold event, but not count the volts gained from low irradiance. Go measure voltage some really cold morning and tell me what it is compared to the standard computed value - a few percent is all you need to get that extra 20 volts you were fretting over to get that extra module in. furthermore it is highly unlikely an inverter will smoke at +20,+40 volts even if it happens at all. Even if an inverter does have a warranty issue, and even if the inverter records the voltage, and even if the voltage has been slightly exceeded, the manufacturer still may not automatically disqualify the warranty. Even if they do, I imagine the labor and hassle are more of a concern than the actual repair/replacement cost....Maybe I am biased as I have been installing these systems in the 35 degree rain in mud 1 foot deep :rant: :D
Click to expand...

SMA does indeed record the maximum DC voltage and void the warranty if you exceed their maximum; I was in an SMA training session a couple of weeks ago and got it from the horse's mouth. Voltage does not change much with low irradiance; I don't consider it. As a designer and a PE my license (or at least my E&O insurance) is on the line when I stamp a design, so yes, I am conservative when it comes to Voc. Of course, as an installer you can change the string length in the field (sorry about your working conditions; I have been there and done that), but if you do it's you who is on the hook if things go sideways.

That said, it's rare for me to have a maximum calculated string length so close to an integer that a few degrees difference either way would change it. YMMV, of course.
 
Last edited:
ggunn said:
SMA does indeed record the maximum DC voltage and void the warranty if you exceed their maximum; I was in an SMA training session a couple of weeks ago and got it from the horse's mouth. Voltage does not change much with low irradiance; I don't consider it. As a designer and a PE my license (or at least my E&O insurance) is on the line when I stamp a design, so yes, I am conservative when it comes to Voc. Of course, as an installer you can change the string length in the field (sorry about your working conditions; I have been there and done that), but if you do it's you who is on the hook if things go sideways.

That said, it's rare for me to have a maximum calculated string length so close to an integer that a few degrees difference either way would change it. YMMV, of course.
Click to expand...

I get that the person responsible for a decision will tend to err on the conservative side. I am sure most professionals do it. A heating guy doesnt want the furnace he spec'ed to be undersized so will tend to oversize the system, etc.....Of course there are industry accepted methods for many things that one can use to CYA. Just IMO there is excessive fear about the VOC and it as calculated more conservative than necessary. I think the NEC now allowing it to be modeled reflects that. Here is a somewhat dated article by Bill Brooks I have posted before, and he agrees with me. Being somewhat dated, he is discussing using ASHRAE temps instead of record lows, but I believe going from ASHRAE to modeled is analagous and an extension of that.
http://solarprofessional.com/articles/design-installation/array-voltage-considerations#.W4yEes5KiM8

Dont you run into situations where you get like 1015 volts using temp corrected STC VOC and want to get the extra "half" module in there?

P.S. I have always wanted to see some accurate data on VOC vs irradiance - pretty much all the graphs one pulls up on google are pretty "coarse" and hard to get values for the VOC change. I found this which has some good data, table 4 page 284:

https://ac.els-cdn.com/S18766102173...t=1536021657_effb00b99835de2fb771caf3b8369f7e

The 1000w/sq vs 800w/sq values in the paper agree with what I got by taking module data at STC vs NOCT values and temp corrected for the 20 degree difference: about .5 volts. Thats only about 9 volts difference for 1kv strings, but It starts to drop off more rapidly from there. Around 500w/sq looks like about 1.5 volts per module which would be about "half a module"
 
electrofelon said:
And for the other point of view, I find Designers to be overly conservative with VOC. Irradiance will be quite low when these low temps occur. I dont know why everyone will count every volt from a 100 year cold event, but not count the volts gained from low irradiance. Go measure voltage some really cold morning and tell me what it is compared to the standard computed value - a few percent is all you need to get that extra 20 volts you were fretting over to get that extra module in. furthermore it is highly unlikely an inverter will smoke at +20,+40 volts even if it happens at all. Even if an inverter does have a warranty issue, and even if the inverter records the voltage, and even if the voltage has been slightly exceeded, the manufacturer still may not automatically disqualify the warranty. Even if they do, I imagine the labor and hassle are more of a concern than the actual repair/replacement cost....Maybe I am biased as I have been installing these systems in the 35 degree rain in mud 1 foot deep :rant: :D
Click to expand...

It's pretty safe, and conservative, to assume you will have STC Voc at 20% insolation. Current is very dependent on insolation, open circuit voltage much less so. I've heard the argument that the daily low occurs before sunrise so why worry about it, or the insolation is so low in the morning why worry about it. There is also cold soak to consider, and supercooling due to dark sky radiation that many people don't know about much less consider. Someone can go get a Ph.D. in physics and know down to the degree what is happening and still get it wrong because temperatures fluctuate and are not always what we think they are going to be. Yesterday's record low is broken today.

Many people use the ASHRAE extreme annual mean minimum temperature. But notice the "mean" in there. That is the average of the low temperatures which means half the time the temperature will go below the number being used in the design. I usually use the mean minus one standard deviation, also given in the ASHRAE tables, that means about 80% of the time the temperature will be above my design low temperature. Belt and suspenders.

It's one thing to make a boo boo when designing a 5kW residential system and popping an inverter, it's a little different when one is designing a 1MW+ systems and popping inverters will cost millions of dollars and the EPC will be dragging the designer into court to cover that loss.
 
Last edited:
electrofelon said:
It will in fact be about 11% lower. A big part of engineering is weighing risks with the likelihood of those risks. We just have a difference of opinion as to the likelihood and consequences.
Click to expand...

Likelihood, OK, but consequences? Some if not all inverters record the highest DC voltage they see, and some if not all the manufacturers of those inverters will void your warranty if you go over their published maximum irrespective of whether damage was done or whether a warranty issue is connected with the overvoltage incident that was recorded. That is a pretty severe penalty for a chance taken for a minor benefit, so I don't risk it. And yes, I agree, risk/benefit analysis is a big part of engineering. I don't take risks with minimal upsides and major downsides when there are uncontrollable variables like weather involved. As an engineer I believe it to be irresponsible to do otherwise, but YMMV.
 
Status
Not open for further replies.
Top