String sizing tool

[Halfcrazy]

We have been working behind the scenes to bring a Classic string sizing tool to the users. I am happy to say we think we are almost there and we would like everybody to play with it and see what they think. Please look for any odd math or anything else that seems out of place. Also just let us know if you see something you would like to see done different.

I will sticky this thread for know and we can all use it to comment or post pictures of the tool etc.

www.midnitesolar.com/sizingTool/

Thank you
Ryan

[Kent0]

Ryan,

A minor point here that can cause trouble if the user isn't paying attention. PV module datasheets show the voltage temperature coefficients as negative values; the voltage goes down as the temperature goes up. The string sizing tool expects the temperature coefficient to be entered as a positive value. To avoid confusion and prevent mistakes, the negative sign could be put outside the editable portion of the form and the sizing tool could ignore any sign the user enters. At present if the user enters -0.35%, the voltage computed at -30°C is less than at 25°C. Since all PV modules have negative temperature voltage coefficients, there's no harm in treating the data correctly regardless of what the user enters.

[Halfcrazy]

Kent
Thank you that is a great point. I will get that taken care of asap.

Ryan

[Kent0]

Ryan,

Just noticed a similar point regarding the % symbol. The sizing won't accept an input of "0.35%" only "0.35". A % symbol that isn't editable would help prevent someone from entering 0.0035 instead of the equivalent 0.35%.

[niel]

i must be doing something goofy or the calculator is because with a vmp of 18.4v and 2 of them in series the thing says i have insufficient voltage for the hottest day of 103 degrees f. i did put the - sign in front of the temp coef.

[Halfcrazy]

Neil
You found a coding bug I will forward this to Rob and we will resolve this.

Thank you
Ryan

[Kent0]

Ryan,

It's nice to see that this tool is available. I'm impressed how quickly you have addressed issues with it.

One feature I really like is that it tells you the module temperature that will cause the Classic to enter HyperVOC mode. I ran into an issue with the HyperVOC temperature. Modify the default data so there are 3 modules in series, the HyperVOC temperature is -75°C, that agrees with my calculations. Now put 4 modules in series. The Max VOC line indicates the design is marginal and that HyperVOC will be activated. The HyperVOC temperature line indicates "N/A"; it should be 26°C. Not only will HyperVOC be activated, it'll be activated most of the time. This is a design example that shows HyperVOC being seriously abused. I'm not sure how N/A applies to this case.

One thing I'd like to add to the user input is a field for the user to enter the PV module name or model. The only purpose this serves is to identify the page if it is printed.

[Halfcrazy]

Kent
I will look into the Hyper VOC thing. On the module manufacturer and Model # that is a great Idea. I will run this by Rob. He will be out of town for a week so I will make a list. We are working on a database of modules behind the scenes and this should show up in a few weeks.

Ryan

[niel]

its looking pretty good. i must say the charge current would be very optimistic seeing as how the classic itself will draw power too and the charge current is also allowed to be rounded up. i've been guilty of this myself, but it isn't very accurate as far as reality goes. maybe cover your butt by saying max charge current?

all in all i really like it and hats off to who all worked on it.

[niel]

i forgot to ask if you placed some safeguards or red flags like when the watts doesn't jive with the vmp x imp +/-??%.......? range safeguards are good too like putting -272 degrees f is unrealistic, but is interesting. maybe these could still be allowed to be calculated, but with notices of something possibly out of whack?

[Halfcrazy]

Niel
We did generate some warnings but not that particular one. How about "Watts should equal VMP*IMP + or - 10%" I am not sure how much wiggle room to give it but I will look at a bunch of data sheets and see what I find. It could be a soft error that lets you continue but says"Warning: the VMP, IMP and Watts do not agree with each other" ?

Here are a few Errors we do have in there that I can remember off the top of my head:

VOC must be greater than VMP
ISC must be greater than IMP
Coldest ambient must be lower than Hottest Ambient.
VMP of the array on the hottest day must be 125% or more of Nominal Battery voltage.
We ignore - signs in all fields besides temp so if some one adds a _ sign or not the temp coefficient always stays correct.

Of course there is the VOC reality check and the Wattage reality check as well in the "Design Check"

I welcome suggestions for other errors I should have as well as suggestions about everything from cosmetics to language etc.

Ryan

[Halfcrazy]

Based on some suggestions I have mad some quick hen scratching on a screen shot. Please look at the photo here: https://www.dropbox.com/gallery/31975941/1/New%20folder?h=b5ea3c

I added "Max" to battery charging current as well as "STC" to the values based on Nameplate. I also shortened Maximum to Max under modules in series.

Now my big questions:

Does the "Design Check" section labeled "Array Power (Wattage)" section make sense? I had one opinion that including (Wattage) and saying "Excessive" was confusing and I should get rid of (Watts)?

Also the same person mentioned the "Number of Classics required" section was confusing because we left it in decimal? My decision there was based on the thought you could look at it and say "Oh I need 1.4 Classics I can add more panels" Or "I need to take out a panel or two" it was sort of like a little design help. Keep in mind this tool is truly designed for a single Classic and anything over 1.2 will stay Red and the wattage will show "Excessive"

Ryan

[Kent0]

Ryan,

To address the decimal number of Classics required: how about putting the number in percent and change the label to "% of Classic capacity". The note could be changed to say, Midnite Solar recommends that the rated PV array power not exceed 120% of capacity. With grid connected inverters it is common practice to install an array that is 120% to 125% of the array capacity, so a 120% recommendation shouldn't look alarming or strange to anyone.

In the design check I would omit "(wattage)" because the row doesn't have numeric data or units. In fact I would avoid the term wattage entirely as it is a colloquialism. The appropriate words are either "power," or if talking about power units, "watts."

There is a misspelling/typo: "Maximun number"

Maybe I'm archaic but the temperature display units look backwards to me. Shouldn't it be °C and °F instead of C° and F°?

I took a closer look at the anticipated array power at maximum temperature value today and think it is misrepresented. The array power appears to be computed based on the PV module temperature being equal to the maximum air temperature. The actual PV module temperature depends on more variables than you probably need or want to consider. However, if there has been sunlight on the module for very long, its temperature will always be above the ambient temperature. A useful number for this purpose is the NOCT value on the manufacturers' datasheets. It is normally in the 45°C to 48°C range for an ambient temperature of 20°C, irradiance of 800 Watts per square meter, and no wind. At 1000 Watts per square meter the module temperature will be a little hotter yet. Short of having the user enter even more data, I'd estimate the module temperature as 25°C more than the maximum ambient temperature and maybe include a note that this value is just an approximation - "your actual mileage will vary."

[niel]

ryan,
looking good, but there's another range thingy needing addressed as the temp to enter hyper voc can go lower than absolute 0 which is -459.67° f (-273.15° on the Celsius scale). when it hits that point or lower you can have it revert to just saying absolute 0 or never.

as to +/-% for calculated watts that % may be too lax as we need to go by the stc ratings for all pvs. i'm thinking of somebody hitting say 18.4v vmp when they may have wanted 18.3v vmp. this could translate into as low as .5% in error so i would tighten it up to at least +/-1%. what a pv may actually output with +/-xx% we can't determine and we must go by the actual stc ratings on spec sheets. sometimes spec sheets are rounded a bit in their numbers so i'm not too sure what you may need to up the % range to to accommodate this rounding, but i'm sure it is a tight tolerance. on a 200w pv this would be +/-2w for +/-1% error allowance and in my example of entering 18.4v vmp instead of 18.3v vmp by mistake would still be allowed at +/-1% calculator tolerance and may be more indicative of another pv with a higher vmp rating (and higher imp and wmp to come up with the usual 5w pv power increment). follow what i mean? it may not be realistic to catch all errors, but i think the spec sheets would be tighter than 1% in their rounding and i think you can tighten up the the error % as well. i suppose whatever error message chosen may have to indicate the +/-xx% chosen. of course this is before temp compensation which could skew things.

the warnings and soft error messages are good and you should still allow the calculations where applicable.

i think kent o meant 'f and c instead of c and f'. he put 'c and f instead of c and f'.

[Lya72]

Hi,

Very good tool for newbies like me.

A little thing :
On my Ipad, there is a formatting problem on the temperature block. the second radio button is going on the next line near the C° label.

Yann