diy solar

diy solar

Help with 24V system wire size

That ‘kicker’ (or safety margin) is required for DC PV wires and for DC power distribution from batteries,
Hmmmmm..... I knew about the extra margin on PV, but with PV they are dealing with a current limited power source and are calculating wire sizes for no breakers.

I was not aware of the extra 25% kicker on the battery distribution wires. Is that in the NEC? Do you happen to know the section?

BTW: Some DC breakers are rated to run at 100% of trip current and with those, the 1st 25% kicker is not required. (I *think* the breakers that Midnite Solar sells are rated for 100% of trip current)
 
Hmmmmm..... I knew about the extra margin on PV, but with PV they are dealing with a current limited power source and are calculating wire sizes for no breakers.

I was not aware of the extra 25% kicker on the battery distribution wires. Is that in the NEC? Do you happen to know the section?

BTW: Some DC breakers are rated to run at 100% of trip current and with those, the 1st 25% kicker is not required. (I *think* the breakers that Midnite Solar sells are rated for 100% of trip current)
You could be right.

I found this: https://www.fs.fed.us/database/acad/elec/greenbook/3_basicdesigns.pdf

‘Wire size = Load / 0.8 = Ampacity’, so clearly the wire needs to be sized at at least 125% of continuous current.

And ‘maximum loading of any circuit breaker is 80% of any non-motor loads.’

So it seems as though if you have 130A of continuous current, you need wire that will handle 162.5A (2AWG is using ‘chassis wiring’ limits or 2/0 AWG if using power distribution limits (>2hrs)).

And you can also use a fuse/breaker with the same limit (162.5A).

Of course, since 162.5A fuses don’t exist, you’ll need to get a 170A or 175A fuse/breaker meaning that since:

‘Breaker size cannot be larger than ampacity of wire’

You’ll still be OK with the same wire sizes.

So the point is, if you may be drawing currents of 130A for 2 hours continuous or more, you need to be using 2/0 wires, but if your peak current of up to 130A will never last for 2 hours straight and currents will never exceed 90A for more than 2 hours straight, you can get by with 2AWG…
 
So it seems as though if you have 130A of continuous current, you need wire that will handle 162.5A (2AWG is using ‘chassis wiring’ limits or 2/0 AWG if using power distribution limits (>2hrs)).

Very few of my installs ever run the inverter at max and none of them would ever run at max for an hour, let alone 2 hours. In fact for smaller systems, I often have something like a 3K inverter but max load if everything is on would be more like 2KW.

However, someone could plug in a bunch of stuff and run the inverter at 3kW for an extended time. So... does it have to be wired for > 2-hour continuous load? It would be interesting to know what an inspector would say.

The other piece of the puzzle is what type of wire is being used. I almost always use marine grade wire because of it's fleibility.
It is rated for 105C so the amp rating is a lot higher as well:

105C Marine Grade Wire Specs.
1633409930874.png

So my process is to first calculate the inverter current at low battery level. For a 3000W inverter on a 24V nominal battery
3000W/24V=125A. (This is very conservative because I try to design the system such that the battery never gets this low and I try to design the system to never run the inverter at its peak )

I then factor in the inverter efficiency. Let's assume 90%.
125A/.9=139A.

I typically use fuses so I do the 25% margin.
139A x 1.25 = 174A. I would use a 175A fuse.

I pick the Marine-grade wire that can handle 175A. In this case, it is the 2AWG.

If I were to use other types of wire typically use in buildings, it would need to be much larger wire (2/0 or 4/0)
1633410780077.png
 

This is the NEC free air table which is surprising close to the AYBC values.
Especially when you consider its for 90C insulation instead of 105C.

"Table 310.15(B)(17) (formerly Table 310.17) Allowable Ampacities of Single-Insulated Conductors Rated Up to and Including 2000 Volts in Free Air, Based on Ambient Temperature of 30°C (86°F)*"

 
Very few of my installs ever run the inverter at max and none of them would ever run at max for an hour, let alone 2 hours. In fact for smaller systems, I often have something like a 3K inverter but max load if everything is on would be more like 2KW.

However, someone could plug in a bunch of stuff and run the inverter at 3kW for an extended time. So... does it have to be wired for > 2-hour continuous load? It would be interesting to know what an inspector would say.

The other piece of the puzzle is what type of wire is being used. I almost always use marine grade wire because of it's fleibility.
It is rated for 105C so the amp rating is a lot higher as well:

105C Marine Grade Wire Specs.
View attachment 67529

So my process is to first calculate the inverter current at low battery level. For a 3000W inverter on a 24V nominal battery
3000W/24V=125A. (This is very conservative because I try to design the system such that the battery never gets this low and I try to design the system to never run the inverter at its peak )

I then factor in the inverter efficiency. Let's assume 90%.
125A/.9=139A.

I typically use fuses so I do the 25% margin.
139A x 1.25 = 174A. I would use a 175A fuse.

I pick the Marine-grade wire that can handle 175A. In this case, it is the 2AWG.

If I were to use other types of wire typically use in buildings, it would need to be much larger wire (2/0 or 4/0)
View attachment 67530
Good - sounds like our logic is slightly different but we end up at largely the same place - didn’t realize thee that 105C rated was a thing so I’ve generally focused on 90C-rated wire (THWN).

If you’d use a 175A fuse for 3kw@24V and a 2/0 ‘only’ rated for 90C, we’ve pretty much ended up at the e same place ;(though I used a 225A fuse…).

Good reality-check ;).

Your fuse will blow earlier than mine but those 2/0 90C wires rated for 195A are only rated for that amperage in conduit or for over 2 hours continuous use at that current.

Wired through air (chassis wiring), those same 2/0 conductors can handle up to 283A for up to 2 hours continuous, so I should be within spec unless my inverter runs flat-out at 3kW / 120A for over 2 hours straight.
 
This is the NEC free air table which is surprising close to the AYBC values.
Especially when you consider its for 90C insulation instead of 105C.

"Table 310.15(B)(17) (formerly Table 310.17) Allowable Ampacities of Single-Insulated Conductors Rated Up to and Including 2000 Volts in Free Air, Based on Ambient Temperature of 30°C (86°F)*"

Thanks for that reference.

Single conductors in free air up to 86F is clear (300A for 2/0) but does anyone understand what ‘up to 3 single conductors supported on a messenger’ is a reference to? (247A for 2/0).

Or ‘Single copper conductors triplexed in air? (250A for 2/0)

The difference in current ratings between ‘in conduit’ and ‘in free air’ are clear in that document, but there does not seem to be any distinction based on 2 or more hours of continuous use versus under 2 hours of continuous use…
 
Thanks for that reference.

Single conductors in free air up to 86F is clear (300A for 2/0) but does anyone understand what ‘up to 3 single conductors supported on a messenger’ is a reference to? (247A for 2/0).

Or ‘Single copper conductors triplexed in air? (250A for 2/0)

The difference in current ratings between ‘in conduit’ and ‘in free air’ are clear in that document, but there does not seem to be any distinction based on 2 or more hours of continuous use versus under 2 hours of continuous use…
I can't cite it but the de-rate for continuous service is .8.
A 120VAC ceramic heater with a 15amp plug with draw 1440 watts.
120 * 15 = 1800 * .8 = 1440.



Triplex is also 3 insulated wires in one outer jacket.
 
I can't cite it but the de-rate for continuous service is .8.
A 120VAC ceramic heater with a 15amp plug with draw 1440 watts.
120 * 15 = 1800 * .8 = 1440.



Triplex is also 3 insulated wires in one outer jacket.
Thanks. So Triplex is similar to NM in some ways (three conductors being run together, but not in metal conduit).

Any idea what ‘three single conductors supported on a messenger’ is?

Never mind, found it: https://www.electricallicenserenewa...ation-Courses/NEC-Content.php?sectionID=791.0

‘In the 2020 NEC, a messenger wire is defined as a wire that is run along with or integral with a cable or conductor to provide mechanical support for the cable or conductor.’
 
That is for the ampacity, which is only half of what drives the size of your wire. The other is the loss on the wire due to resistance. For that you need to look at a different NEC table. I don't remember for sure, but I thought it was NEC chapter 4 table 8 (but who knows?). Here's a pic of it:
View attachment 67495
So let's say you have your batteries 40 feet from your inverter, and you expect a maximum current of 100A (40 feet is extreme, but stay with me). Based on the ampacity, you decide 4AWG should be close enough (Ampacity of 95A with 90°C rated insulation). Now if you look at the table above, you see that 4AWG has a nominal resistance (at 25°C) of about 0.2533 ohms per thousand feet. In 40 ft, the resistance is 40 * 0.2533 / 1000 = 0.010132 ohms. Now multiply that resistance times the current of 100A, and you get a loss in 40 ft of 40*0.010132 = 0.405V. But wait! You have current going down two wires (red and black) for that 40 feet. So the total loss is actually 2 * 0.405V = 0.81V. In a 12V system, loosing 0.81V in the wire is a killer. It's only a little better in a 24V system.
I assumed that the reader would apply the Ampacity chart to the AC side of the inverter, which is what it's mainly intended for, not the DC (low voltage) side. On the DC side, I keep, as I think I mentioned, the cables short (to limit cost, weight, and, of course, voltage drop). As a guide to the lower limit of cable size, I like the SAE table which BatteryCablesUSA thoughtfully includes in their cable descriptions (I copied the relevant part, below). I've used 4AWG from my batteries to my 2000W inverter and a couple of other things, the OP might get away with 4AWG for a 3000W inverter, if it's short, or go to 2AWG. This morning, when I turned on a 1000W electric kettle, I looked at the voltmeter, voltage sagged 0.9V immediately, and then declined more slowly as my mix of LFP and NCA batteries dealt with about a good part of 1C load for 5 minutes or so. A couple of minutes, maybe a little more, after the kettle turned off, voltage was back to perhaps 0.1V or 0.2V lower than at start. That's with cable lengths under 2 ft. I expect voltage will sag less when I parallel a 2nd 55AH LFP battery.

SAE J-378 Table of Allowable Amperage for conductors 50V or less with 105C rating Gauge Amps Outside Amps Inside AWG of Engine Space of Engine Space 8 80 68 6 120 102 4 160 136 2 210 178 1 245 208 1/0 285 242 2/0 330 280 4/0 445 378
 
Last edited:
I assumed that the reader would apply the Ampacity chart to the AC side of the inverter, which is what it's mainly intended for, not the DC (low voltage) side.
I wouldn't say ampacity is of no import on the DC side. Because the amps tend to be much higher on the lower-voltage DC side, it is pretty easy to exceed the ampacity and have some melting wires. If there is any distance at all (say more that 10 feet), the ampacity will be much less significant than the wire resistance and the losses. If you size the wire to keep the losses low enough, you will be way under the ampacity limits of that wire. On the other hand, if the runs are very short, the losses will be low and it may be tempting to use a smaller wire than is suitable for the amperage that it will carry.
 
I wouldn't say ampacity is of no import on the DC side. Because the amps tend to be much higher on the lower-voltage DC side, it is pretty easy to exceed the ampacity and have some melting wires. If there is any distance at all (say more that 10 feet), the ampacity will be much less significant than the wire resistance and the losses. If you size the wire to keep the losses low enough, you will be way under the ampacity limits of that wire. On the other hand, if the runs are very short, the losses will be low and it may be tempting to use a smaller wire than is suitable for the amperage that it will carry.
Neither would I. In fact, I didn't. I said it's not mainly intended (perhaps I should have said applied) there.

Also I said I prefer:

SAE J-378 Table of Allowable Amperage for conductors 50V or less with 105C rating Gauge Amps Outside Amps Inside AWG of Engine Space of Engine Space 8 80 68 6 120 102 4 160 136 2 210 178 1 245 208 1/0 285 242 2/0 330 280 4/0 445 378
 
Back
Top