New Sapeye Lake system: Part 1 CC/Array

[Westbranch]

New System Conundrum:

Ok here is my current situation, we got the walls up and roof on just in time for 2 feet of fresh snow, which has determined that we are not going back for the next 3 – 4 months.  So there's lots of time to think and deliberate about the power system to build/install in what has been designed as a 4 season house.
Due to only getting the walls and roof finished, not being able to access the site for 4 months and not having anywhere warm to stay on site, we are 'behind' schedule and need to modify the building schedule for the next summer season.
We will need lights in order to work at night and I am thinking it best to use part of the old 12 V system, battery /inverters/lights and charge it up with the Genset during the day while I use the cut-off saw and table saw, etc. 
I will try to rig up a couple of 120W panels & MX60 CC on a temp basis but the battery is 25 + feet minimum from the new building so will have to go 110V AC with CFLs and a big thick extension cord.  Things will improve once the basement floor is poured as the battery can be installed in its place of honour...
So the transition period is covered off.  The following are some of the concepts I wish to incorporate:
- Build the system in stages as $ allows,
- Use different inverters for specified loads to minimize 'idle' draw, ie LED lights on small inverter, water pump on large inverter.
- Use load shifting, ie do laundry after float reached, to capture potentially 'lost' solar power or while charging with Gen set.
The plan for the new system should look as follows:
with a starting projected load of 1500 - 2000Wh per day or less as there will be a wood cook stove and a propane fridge and a lot of LED lights and no other major appliances and a good handle on parasitic loads.  Loads are based on assessment of our current house and what we will not have. Once the amount of days >= 6 months possibly an electric fridge (1600Wh/day).  Final Target is <=4000 Wh/day
•   Classic 150
•   Honda Eu 3000iS and Eu 1000i  gensets
•   ~= 1100 Ah  24 V AGM battery, in process of locating/purchasing
•   Phase 1,  12 - 120 W BoRun panels - Vmp 23.4  - Voc 26.9  Imp 5.12 -  Isc  5.54       Nominally 4% charge rate,  40 – 50 A out from CL 150
in 3 panels per string, 4 string array Vmp 70.2 V,  Voc  80.7 V  1440W
I have chosen these as I am in a high wind area and am working towards as small a sail area as possible per pole. 
This array configuration maxes out the 150 at 147.9 Voc at -45 F.  There will be some line loss due to the location of the arrays-to-batt distance.

•   Phase 2,  another 12 – 120W panels

Question 1: is it better to configure at 2 series panels vs 3 panels, to avoid running the CL 150 at max?

[Halfcrazy]

Not sure about the temp adjusted VOC is it really that high? You have a VOC of 80.7 and -45F should be a multiplier of 1.25 or a Temp adjusted VOC of around 101vdc? How often does it get to -45F there?

[Westbranch]

H/C, Not very often now but when we first lived in the vicinity some time back we had 3 weeks of -40 and below , damn cold and beautiful blue sunny skies, so I wanted to be on the safe side of temp correction.

I used a factor of-.32 as I could not find the exact # for these panels, my Mitsubishi 120's are -.346 and the Borun 156w are -.347 which takes me to 153.6V...

there is another thread about designing for +- 80% of max input V ( I think boB might have commented) hence my question about 2 panel stings. 3 will use 30% less wire

HTH

[Halfcrazy]

OK I am going to have to see your math I am totally lost. (It is Monday)

3 of those panels in series will give you a VOC of 80.7vdc and even at -40 I suspect they will never go over 105vdcUnless I am missing something I suspect 4 in series would be just fine.

Also with the Classic I usually am comfortable designing around the average cold if there is a week of record cold the Classic will be in Hyper VOC in the morning for a little while until the sun warms the cells a few degrees. Hyper VOC needs to be used wisely though to make sure you do not spend to much time in Hyper VOC and not making power.

[Westbranch]

Ryan, I used Kent0's spreadsheet for my adjusted values, found here http://midnitesolar.com/smf_forum/index.php?topic=30.0 reply #7
there may be an error in the table.

Eric

PS just got an email fro the factory, they have discontinued the 120 W and only have the 125 W in Mono crystalline now specs are Vmp 17.7 Voc 22.2 Imp 7.06 Isc 7.61

3series x 4 strings gives 144.3Voc
see attached file

Eric



[attachment deleted by admin]

[niel]

westbranch,
now i'm gathering it is an excel file and i can't seem to open it.

in any case 3 pvs with a voc of 22.2v will never add up to 144.3v. that is 66.6voc before temp compensation and i doubt that would ever hit 144v even in an ice age and full sunshine. do you have the temperature info for the pv? for most monos i see about .34%/degree c. from +25 degree c to -43 degrees c (77 to -45 f) that means a change of 23.12% and would make for 3 pvs 66.6voc x 1.2312 or 82voc at -45 degrees f. far cry from 144.3voc i would think. you could do your array as 3 x 4 or 4 x 3 and have no problem as each pv will see a voc at -45 degrees f of 27.33264voc from the original at 77 degrees f and 22.2voc. even 5 in series won't hit 150v at that low temp.

[Halfcrazy]

Easy enough to play with I will run these through the Classics "Wizard" and see what it says but Niels Math is spot on from what I can see. I would go 4 in series to drop the amperage and raise the voltage on that line that you said would have some loss.

[Westbranch]

Niel, it is the new version (2007) of MS Excel, note the x at the end of the file name, the older versions (93 etc.)  of Excel will not open it. 

Eric

add: temp Correction is .3470 % /*C
I sent a PM to Kent as to checking the formulas in his sheet. I agree, after doing a manual check, something is off base,  should be about 82V.

e

[Halfcrazy]

I just ran the 22.2VOC 125 watt panels through the Classics calculator and with 4 in series and -40F I come up with 111vdc as the highest VOC this seems to be exactly in agreement with Niel's calculations of 82vdc for 3 in series.

I can not open that calculator you posted but I can not imagine there is any way possible a panel could more than double in voltage at -45F? My personal array here is Shell 160's they have a VOC of 44.4 or so and run around 96-98vdc when we hover in the -30 to -35F range. Hope this helps I would not be scared to run 5 of those panels in series on a Classic 150 at -40F they would be well under the 150vdc mark

[Westbranch]

thanks Ryan, I ground the #'s by hand and get 111V too. 

An Array of 4 series  x 3 makes things a lot simpler and for array to combiner to CC may use slightly smaller wire too. Off to Niels wire size calculator.

[Kent0]

I ran the numbers for the Borun modules through the spreadsheet and got Voc of 110 V at -45°F for four modules in series.
The reason some of you are getting different numbers with the spreadsheet seems to be because that the spreadsheet expects the temperature coefficient to be entered in volts/°C. Some PV module manufacturers provide the data in volts/°C and others in %/°C. If you have Excel set up to display comments, there is a comment attached to that cell saying the temperature coefficient is generally -0.35%?Voc. Entering = "-0.35%*Voc" will get a reasonable close value to use if the manufacturer hasn't provided a value. I wasn't able to find a datasheet for the module in question, but all of the other modules on the Borun website seem to have a temperature coefficient of -0.347%. Voc is 22.2 volts so the temperature coefficient would be -0.077 volts/°C. The equation to compute Voc is Voc(-42.8°C) = Voc(25°) + (-0.077)?(-42.8-25) = 27.42 volts. Four in series gives 109.7 volts and that agrees with the spreadsheet.

While hand checking this calculation I did run into an error in the spreadsheet If you enter temperature data in °F you get different results than if you enter data in °C. The correct result is when data is entered in °C so please don't use the spreadsheet with temperature data in °F. The conversion between °F and °C is done correctly but someplace a formula doesn't get the correct value when the data is in °F. Ordinarily I'd correct this right away and repost the patched up version. Unfortunately, due to a hard drive failure on the computer, I no longer have the original. So a correction will have to wait until I have time to reproduce the spreadsheet from scratch.

[Robin]

The Classic correction factor for cold temperature is right out of the NEC handbook. That is typically worst case.