[RE-wrenches] large array string sizing

[William Miller]

Jeff:

If I read this correctly, it sounds like you have an inverter with an 
unusually narrow voltage window -- 12 modules is too few and 13 is too 
many.  This complication is above and beyond the physical limitations 
imposed by the structure.  There is too little information in your e-mail 
to allow one to double check the numbers.

I assume that this project is in the bid phase because these details would 
be worked out at that point.  You need to change inverters or modules and 
CAD a design that satisfies the electrical and physical requirements.  We 
recommend that these projects be designed to the nut and bolt level prior 
to bidding.  It hurts when you put that much effort into a job you don't 
get, but it hurts worse wen you win a bid and find out there are unpleasant 
surprises.

Good luck.,

William Miller

At 09:28 PM 12/7/2009, you wrote:
>We are completing an institutional solar project about 120 kW going on a 
>large parking deck. The large spans require some really large beams to 
>support the multiple rows of modules since support columns are as much as 
>60 feet apart. In other words, adding more modules to each multiple row is 
>a really big structural issue and they are at the limit now.
>
>However, no matter how we design the strings, the number of modules in 
>each row can only be divided by 12 and multiples of 12 which is the only 
>combination of modules per string that also works with the number of 
>modules per row, the number of rows per system, and the number of modules 
>stacked N-S related to shading issues between rows. However,the inverter 
>manufacturer wants 13 modules per string to avoid low voltage cutoff on 
>hot summer days. if I go to 13 modules per string, I am really hitting 
>voltage max during average cold days and believe me, this site will see 
>more cold and snow days (reflections) with temperatures far below average 
>winter temperatures, then it will have hot summer days. The summer average 
>temperature is 83 degrees and I calculate the minimum array voltage is 
>still over 316 volts, on a 480 VAC 3 phase inverter.
>
>Since by both structurally and module count, there is no way we can 
>increase row lengths to provide 13 module strings, and we cannot reduce 
>modules per row to achieve 13 module strings as this would make the system 
>far below what the client expects in system capacity.
>
>Weather data for this location indicates it exceeds our summer design 
>temperature by onlyv60 hours total per year, and no doubt most of these 
>hours occur near the late afternoon at the end of a solar day.
>
>Whats the best way to deal with this less than ideal combination of string 
>sizes since we are getting nailed at both ends - too few modules per 
>string can shut down inverters when hot summer days drop module voltages, 
>and too many modules per string can do some real high voltage damage on a 
>very cold sunny day with a foot of snow on the ground.
>
>What say yea?
>
>Jeff Yago
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.re-wrenches.org/pipermail/re-wrenches-re-wrenches.org/attachments/20091207/5a736d44/attachment-0004.html>