wire size for 10kW PV system
 

I am in the process of building a shop upon which I plan to eventually place 10 kW of grid tie PV panels. The choice of 10 kW reflects a lot of things, but this could increase to 15 kW, but not more.

Given that this is the output (10 kW), the maximum single phase current with full sun should be about 41-42 amps. From this shop, back to the house, is about 200 feet. I am guessing number 6 copper inside PVC underground would be sufficient for less than a 5% voltage drop.

I went to one of the handy dandy online voltage drop calculators and found out that my guestimate of 6G wire was correct (only 3.4% drop).

Voltage Drop Calculator - for single and 3 phase ac systems and dc systems

Now, I need to put in footings TOMORROW and am thinking of the PVC size to place said wiring in. You "can" place three 6G wires in 1"conduit, but the 1.25 inch is only a tiny fraction of money more. And a LOT easier to pull cable through.

If I went up to 15 kW, then 6G may not be large enough . . . .

Am thinking of putting in a 4" sleeve under the footing to take almost anything.

Now here is where I am ignorant. I am an engineer NOT an electrician and would like to hear from the electrician end as to recommendations.

The shop will have very little amp demands, no 3 phase welders, no high current stuff. Just lights, a fan and maybe a small AC unit (total of maybe 20 amps). So any wire that backfeeds power at 40 amps (or more) will certainly be large enough for any loads in the shop.

Recommendations before I make a fool of myself (and certainly not the first time!!)?

Thanks in advance,

Steve

The panels are not going to be making 10kW all day, in fact, it only gets there every once in a while. Try using a smaller solar panel and some sort of data logger to get an idea of actual production. You still need to size the wire to safely carry the maximum current, but you can use typical production when oversizing for efficiency reasons, maybe even not have to oversize at all.

Wire is rated in amps for safety reasons. As long as you stay under this limit, the sizing is all dependent on voltage and length.

The same wire drops the same volts per amp per length.

A piece of wire that drops 1.2v at 10A at 120v is perfectly fine.
It wouldn't be appropriate for 12v 10A, because it will still drop 1.2v. (10%)

You didn't say DC side or AC side.
DC side is more dangerous btw.

This is the AC output (10 kW, 42 amps at 240 VAC). Will be using Enphase inverters.

The peak current issue is what I am concerned about with wire sizing, not the average output over 24 hours.

Steve

What kind of A/C? Size and tonnage?

Code permitting, I would go for doubly fed setup if your run is that long. The wire numbers are completely made up, but the setup on right is more desirable:

That inverter is a modern high frequency full conversion gutless power electronics type inverter and have pretty much nothing when it comes to surge power. The starting current will be sourced from your energy provider service. If you were to try to start the air conditioner off grid, it probably won't be able to start it blow up and bellow out the semiconductor smoke.

http://i.imgur.com/8FUiA0m.png

Even if left set is fine as far as safety and capacity, a run that long means your lights will flicker every time the compressor starts and can get VERY annoying. It depends on the compessor's LRA and wire length.

The setup on right would significantly isolate the lighting circuit from flicker. The voltage sag will be there at every start, but it just avoids it from affecting lights.

Probably stating the obvious here but check your mains voltage at the time of day you expect maximum solar output.
Add this to the maximum voltage drop you calculated and check it,s well under the maximum voltage output of the intended inverters.
If it,s close use bigger cables to reduce the voltage drop, the last thing you want is to have the inverters constantly shutting down due to over voltage.

At 240v and you size according to code (NEC in the US) for safety and voltage drop for the given current level, it should be fine.

ANSI standards allow +/- 5% steady state voltage at the service entry, so a non-sucky inverter shouldn't have any problem pushing the power into steady state of 228 to 252v.

Ormstron

I did check the voltage and voltage drop at the input of the system - by the meter. I did this with a load test and put about a 300 amp load on and only got a 2 v drop.

This is the result of putting in large G copper (not aluminum) with 400 A total service. 100 A to two homes. 100 to large livestock barn (including office) and 100 for future shop (now under construction.

So by backfeeding to grid (from PV panels) I don't have to worry about a large resistance causing a significant voltage rise on mains. Excellent point that I did not think of.

I know ANSI speaks of code being 5% drop. Code is the MINIMUM standard and I would prefer to be better than that. The issue is how much better. I think a 3% voltage drop is about half that and is OK. I can reduce it to a 2.5% voltage drop with 5G wire, but with a lot more money.

Engineering is about getting the most for your $, but an electrician will tell you that it is all about safety as well. Trying to balance both.

Steve
ps - only will need a 12,000 btu AC unit - essentially nothing for load

An A/C that small is not going to be any problem. It would be very unlikely for a wire run rated for at least 42A to cause starting issues, especially if you get a 240V mini split and even more so if it's inverter drive. (In the unlikely event it does, add a hard start kit!)

http://enphase.com/wp-uploads/enphas...tions_M215.pdf

Here is an Enphase doc that shows the voltage drops calculations.
Enphase microinverters do not like that big of voltage drop.

Luckily I was about 130’ away not as far as you going.

If there is too much voltage drop the micro inverters will think the grid is unstable and shut off wait 6 min and check again if the grid is back.

Assuming a 250w panel and the slightly older M215 inverters, there is a max of .91A x 60 panels = 54.6A x 1.25 (safety) = 68.25A.

So you would need a #4 wire as #6 is only approved up to 60A. Note that you are on the low side of the capacity for a #4 (at a minimum) so without doing much more calculation, I suspect you would be quite good for voltage drop too. #2 is more typical for a 100A service (if you use a 90C wire).

If you buy Enphase M215 you will use .9 amps for the current but if you end up using the new M250 you will use 1 amps for the current. (use these number in your calc just like in the doc @ 240 Volts)
M215 actually max out at 225 watts where the M250 max out at 250 watts.

So the 250 watt panels go well with the M215.

Here is a good sizing doc's

http://enphase.com/global/files/Enph...ightsizing.pdf

Enphase has put out some really good documentation
http://enphase.com/global/resources/geo_us/
They have module Compatibility list and much more.


If you are comparing solar modules go here California actually test the modules to show how they perform.

Incentive Eligible Photovoltaic Modules in Compliance with SB1 Guidelines - Go Solar California

My Canadian Solar modules are 230 watts and they tested out at 211 watts
My Light way 240 watts tested out 215 watts.

This helps you see what you are paying for really works. I have on occasion max out the M215 @ 225watts not very often at all 250 watt is a better match.

When I called around all the sales people told me the M215 maxed out @215 watts this would have made the 230-240 a good match. They we giving out bad info that is why I am bring this up now.

I would not use M250 with 250 watt modules because you would not make more power and it would put you using 1 amp in the calc. this would just make you spend more money on wiring with no real additional power being made.

I am using M215 with 54 panels 54 X .9 = 48.6 X 1.25 = 60.75 amps
If I was using the M250 54 X 1 = 54 X 1.25 = 67.5 amps
I would make no more power but have to buy bigger wire.

Now if I would have bought 270 watt panels that would be better for the M250. It seems it is always the smaller wattage panels that go on sale. if space is a issue M250 might be better I don't think that is the case for you.

The reason I stated using the M215 was, as you stated. I have installed lots of systems like this because the 250w panel was the best price point and the M215 has max performance that is seldom seen in a 250w panel. You may lose a few watts here and there but very few. As you say, if there was a 260w+ panel, that is different.

This is very interesting as my main service panel is actually even farther away than the 200 feet I stated. The 200 feet length brings me back to a sub-panel where I have a larger diameter wire (4G) going to the main point of contact.

I may have to reconsider putting the PV panels on the shop and use the big livestock barn instead. The main service panel is locate in that barn.

I had no idea that the Enphase inverters were so sensitive to voltage drop (or increase depending which way current is flowing). A 1% voltage drop is a very small amount.

So my question: Is the stated 1% voltage drop an actual necessity or is it something that is really not needed?

I can understand a too small wire size causing a voltage increase given that the PV panel will backfeed across the wires and this current multiplied by the wire resistance will cause an inappropriate increase in voltage at the Enphase end of things (Ormond's comment).

But 1%?

Nowhere in the Enphase documentation does it say why 1% is needed.

I bet most main wiring for Enphase installations do not meet this rigor.

Steve

IIRC, the code is quite lax and may even be 3 or 5%. I think that they are saying 1% so the most power gets out of the system. Remember that power is transmitted based on voltage differential between the grid and the inverter and to some extent, the higher the voltage the grid sees, the more transfer will take place.

I really think it is the voltage rise that becomes a issue.


They are back feeding the grid so they monitor the voltage. The grid voltage goes up they need to go higher to back feed the grid. If the grid has loose tolerances and you have too much voltage rise you can get out of the window that the micro inverter can go. They can adjust these tolerances some the window they call grid profile some. I had them adjust it on my array. They told me any more adjust needed I would have to call my power company to do a line study and get their power more in speck. I am one of the last two houses on this line a new power company provides just two house away.


"
Utility voltage: The utility strives to maintain voltage at the PCC within +/- 5% of nominal. The protective functions of the microinverters are set to +10%/-12% by default. The high voltage end of the tolerance is of most concern because the inverters are a SOURCE and not a LOAD. If the utility is consistently 5% high, that leaves less than 5% for all wiring and interconnection losses and inverter measurement accuracy. If you are concerned about the utility’s voltage, you may request that your utility place a data logger at the PCC and make a record of the voltages available to you at the site. “

This is the doc I used for my calculations.
http://enphase.com/wp-uploads/enphas...Vdrop_M215.pdf
page 4 talks about line volatage rise. I center feed all my runs and ran bigger wire because of my distance.

I found this to be a very useful doc.

Just call Enphase support they were really good with pre sale support. I called a couple times before I purchased my system. As long as it is not a NEC code question they are good at answering questions.
Every time I have called support they knew what they were doing or very quickly got me to someone that knew the answer. I have not found this in most other companies support. Usually first level cant tie their shoes.


Keep in mind that I am going by the documents Mike has lots of installs under his belt. Book smart does not ever take the place of actual field experience.

The other thing most people just install on their roof none of this stuff applies because the cable runs are pretty short. 95% of the install never have to even think about these voltage rise issues. It is just when you add distance that makes the difference then it matters.

Maybe Mike has done some installs with long distances and can share his experience.

The video I watched with one install with distance similar to mine they center taped and use bigger wire just like my install.