Instead of #8/2, can I use #10/4 using paired conductors

[Mitchell]

Summary
Does the electrical code permit using two smaller-size conductors in parallel instead of one larger conductor.

Detail
I'm in Toronto, Canada and am preparing to buy/run the cable for an OpenEVSE to charge at the full 6.6 kW, and I understand about needing to use #8 wire on a 40-amp 240-volt circuit.

Since the EVSE would be in a garage attached to the house, and in the garage the cable would be exposed, rather than using conduit, I plan on using armoured AC90 (BX) cable all the way from the main panel.

The problem is apparently nobody makes #8/2 (that is, two insulated conductors plus a ground wire) armoured cable, so the usual choice would be to use #8/3 (which is widely available) and simply not use the third conductor.

But I see that #10/4 armoured cable is available, and applying the 80% derating for having more than three conductors in a raceway or cable, each conductor can carry (30 amps x 80% =) 24 amps, so using two conductors in parallel for the L1 and the other two conductors for the L2, this should be allowed to carry up to 48 amps, so the #10/4 can carry more than the 40 amps needed. Not only does this not waste a copper conductor, the cable diameter is less (#10/4 AC90 is only 0.62" diameter, compared to 0.838" for the #8/3).

Another advantage of using #10/4 paired conductors is that #8 solid conductor copper wire has a resistance of 0.6281 ohm/1000', but two #10 in parallel have a total resistance of only 0.499 ohm/1000'. So for my 50' cable run (100' of cable for the loop from the panel to the EVSE and back), the total power lost in the cable due to resistance would be reduced from 47.5 w to 37.8 w, so it would be more efficient too.

But I've been told that while doubling-up conductors is allowed for a service run, it is not allowed on a secondary circuit. I've briefly checked the Canadian and Ontario Electrical Code books (many years ago I knew these well, but only have old versions of them now), and don't see anything about restrictions on using wires in parallel. Does anyone know of electrical code or other reasons for not using two #10 in parallel instead of one #8 conductor.

 

[Nubo]

You wouldn't be doubling-up the ground conductor though?

 

[TonyWilliams]

You wouldn't be doubling-up the ground conductor though?

By code, the ground can be ONE size smaller.

 

[AlanSqB]

Not to dismiss the very valid and interesting technical side of this, my practical side says that it's probably overkill.

The 6.6kw on the LEAF tops out at 27.5 amps. This is so near the de-rated value of the #10 that personally I just set my OpenEVSE to advertise a max of 24A and went with the smaller wire. This saved me $$ and time and I still have a car that safely charges from VLBW in just a few hours.

I did this knowing that I will probably need to upgrade the run for my next EV, but I'm probably going to need a 90A sub panel for that anyway so the difference was moot.

YMMV, but I wanted to share my thought process as I tend to be a "buy the best/most expensive/most future proof solution" type person but I talked myself out of it in this case.

Thanks
Alan

 

[Mitchell]

Thank you all, good point about the ground, I hadn't thought of that. The #10/4 cable's ground is already one size smaller (#12), so that is definately a good technical/code reason not to do this (when I received the advice this "just isn't" done on branch circuits with no other reason, I had thought that was just close-mindedness). I appreciate the quick and knowledgable reply and will abandon the #10/4 idea.

I too had been wondering if I should put in a larger cable now "for my next EV" (and I don't even have my first one yet), but then wondered, how big to go. The suggestion of 90 amps may be too much for me, as my house only has 150 amp service, with a 40 amp circuit to the air conditioner. Maybe I'll run #6/3, as that would allow some future increase, and the 0.915" cable diameter would still fit through the 1" hole I was able to drill in my concrete block wall.

My other choice was to share my 30 amp dryer outlet (accepting the slightly lower charge rate), as the dryer is just on the other side of my concrete-block wall. The interesting challenge was whether this could be done by modifying the OpenEVSE instead of using a manual transfer switch. Perhaps by adding a second current transformer (the OpenEVSE appears to have an unused ADC input available) to sense the dryer current and modifying the OpenEVSE code so it would stop charging if the dryer was started while it was charging. After cutting some holes in my basement ceiling, I decided it would be simpler to just run a new 40 amp circuit.

 

[AlanSqB]

Yes my panel is also 150A so it will also include putting in a larger panel. Luckily the service pull is good for way more than the 150 as verified by my power company.

Still going to be a paIn

 

[dhanson865]

You want to know and talk to people that have dealt with it check out http://www.teslamotorsclub.com/showthread.php/12615-FAQ-Home-Tesla-charging-infrastructure-Q-A and post on the teslamotorsclub forum. Plenty of people there that deal with the codes and have cars that charge at 80a.

Keep in mind they think about even bigger wires than you were thinking of

WHAT WIRE SIZES SHOULD BE USED FOR INSTALLING MY HPWC?

Note: All wire sizes assume copper conductors!

For wire-in-conduit, 3 AWG THHN (dry locations) or THWN (wet locations) is sufficient to carry 100 amps, but many electricians will use 2 AWG to ensure there is enough headroom, and it’s carried more widely by supply houses. The ground wire must be a minimum of size 6 AWG.

If these conductors will ride in the same conduit as another circuit, they must be “de-rated” and 2 AWG is required.

2/3 NM cable available in the US (3 conductors of 2 AWG) is insufficient for the 80A charging / 100A breaker setting, because that cable is rated only at 95A per articles 334 and 315, and article 210 requires that an 80A charging load use a conductor with minimum ampacity of 100A. Likewise, I am told that Canadian consumers have the option of type NM cable (3/3+ground) for this installation, but NEC does not permit this (because 3 AWG must be used at the 60 degC rating, only 85 amps).

This tends to generate a lot of confusion, since there is a "next size up" breaker rule, but not a wire/conductor rule. Here is the applicable NEC code:

Section 625.40/625.41 states that vehicle charging loads are always to be considered continuous loads.
Section 210.19(A)(1)(a) says that "Where a branch circuit supplies continuous loads or any combination of continuous and noncontinuous loads, the minimum branch circuit conductor size shall have an allowable ampacity not less than the noncontinuous load plus 125 percent of the continuous load." An HPWC configured for an 80A continuous load, then, requires a conductor with an "allowable ampacity not less than" 100A.
Section 334.80 requires you to use the 60 degree column of 310.15(B)(16) when determining allowable ampacity of the NM cable.
Allowable ampacity of conductors is defined in section 310.15. The 60 degC column of 310.15(B)(16) states that the ampacity of #2 copper conductors is 95 amps.
As 95 amps is not "not less than" 100 amps, 2/3 NM cable would not be usable for this application.
The 240.4 "next size up" rule, for breakers only, doesn't apply to conductors, so it cannot be used.

So if you want to future proof you need to be able to run a circuit that can handle really big charger limits:

80a - Model S dual charger
72a - Model X optional charger
48a - Model X standard charger
40a - Model S single charger

although I figured a 14-50 socket / circuit was enough for me. I don't see the need for >40a charging since I don't drive long distances but if you do keep in mind how big the wires can get for the bigger charging setups.

 

[wwhitney]

By code, the ground can be ONE size smaller.

That's not correct, at least for the NEC; I don't know about the CEC (Canadian Electrical Code). For EGC sizing on a branch circuit or feeder under the NEC, see Table 250.122. The EGC size depends on the rating of the upstream OCPD, not the size of the circuit conductors. The maximum permitted OCPD for a given size EGC is as follows: #14 - 15 amps ; #20 -20 amps; #10 - 60 amps ; #8 - 100 amps, etc. If you compare that to the ampacity tables in 310.15(B), you'll find that depending on the circuit size, the EGC may be required to be the same size as the circuit conductors (for smaller circuits), or may be permitted to be several sizes smaller (for larger circuits).

There is another wrinkle applicable here, which is 250.122(B). It says that if you increase the size of your circuit conductors beyond the minimum required for the OCPD, you must increase the size of your EGC proportionately. This rule makes good sense for the case of upsizing for voltage drop, but has counterintuitive implications when upsizing for futureproofing.

For example, it is OK to land #8 NM cable on a 40 amp breaker (its ampacity is 40 amps), but it is a violation of 250.122(B) to land #8 NM on a 30 amp breaker: #8 NM has a #10 EGC, as per 250.122 a #10 EGC is good up to 60 amps. But if you land your #8 NM on a 30 amp breaker, you could have used #10 NM for that circuit, and you have therefore upsized the circuit conductors. Since you haven't upsized the EGC from #10 to #8, you have violated 250.122(B).

Cheers, Wayne

 

[wwhitney]

Thank you all, good point about the ground, I hadn't thought of that. The #10/4 cable's ground is already one size smaller (#12)

That's a Canadian-ism. In the US, the NEC requires that cable to have a #10 EGC.

For your original question, the basic NEC rule is that you can only parallel conductors for size #1/0 or larger [310.10(H)]. So no paralleling of #10s. The CEC may be different.

Cheers, Wayne

 

[Mitchell]

That is interesting Wayne, but here's a twist: if I was to use #8/3 armoured cable with a 40 amp breakers, normally the third conductor would be left unused. But if I parallel the one extra conductor with one of the other two (might as well put that extra copper to work, since this would reduce voltage drop and therefore slightly reduce wasted power), would this violate the ground wire sizing requirement (which may not be applicable to me being in Canada anyways). From what you've stated, this would be acceptable for NEC Table 250.122, but I don't follow the logic of NEC 250.122, I would think so long as the ground wire is large enough to trip the breaker, it would be acceptable to use heavier conductors for the other wires (or just one wire).

Also, would this be any worse if instead used #6/3 cable (to allow for a possible future larger EVSE) with 40 amp breakers (until changed out in the future).

 

[MikeD]

Mitchell: I believe that the main concern with using parallel conductors is that past (bad) experience has shown that the possibility of overheating of one of the conductors (due to a variety of reasons) causing insulation breakdown for that conductor and a resulting fire is too high.

 

[Flyct]

Here you go. 50' of 8/2 w/ground
http://www.ebay.com/itm/MC-8-2-with-Ground-50-ft-Copper-Electrical-wire-MEATAL-CLAD-/200577575509?hash=item2eb35aea55:m:mVo-CBkY4seELjsw-aZGJLA

 

[smkettner]

afaik NEC does not allow parallel conductors in this size range.

Get the right stuff. The cost is of materials is minor.

 

[ElectricEddy]

Summary
Does the electrical code permit using two smaller-size conductors in parallel instead of one larger conductor.

Detail
I'm in Toronto, Canada and am preparing to buy/run the cable for an OpenEVSE to charge at the full 6.6 kW, and I understand about needing to use #8 wire on a 40-amp 240-volt circuit.

Since the EVSE would be in a garage attached to the house, and in the garage the cable would be exposed, rather than using conduit, I plan on using armoured AC90 (BX) cable all the way from the main panel.

The problem is apparently nobody makes #8/2 (that is, two insulated conductors plus a ground wire) armoured cable, so the usual choice would be to use #8/3 (which is widely available) and simply not use the third conductor.

But I see that #10/4 armoured cable is available, and applying the 80% derating for having more than three conductors in a raceway or cable, each conductor can carry (30 amps x 80% =) 24 amps, so using two conductors in parallel for the L1 and the other two conductors for the L2, this should be allowed to carry up to 48 amps, so the #10/4 can carry more than the 40 amps needed. Not only does this not waste a copper conductor, the cable diameter is less (#10/4 AC90 is only 0.62" diameter, compared to 0.838" for the #8/3).

Another advantage of using #10/4 paired conductors is that #8 solid conductor copper wire has a resistance of 0.6281 ohm/1000', but two #10 in parallel have a total resistance of only 0.499 ohm/1000'. So for my 50' cable run (100' of cable for the loop from the panel to the EVSE and back), the total power lost in the cable due to resistance would be reduced from 47.5 w to 37.8 w, so it would be more efficient too.

But I've been told that while doubling-up conductors is allowed for a service run, it is not allowed on a secondary circuit. I've briefly checked the Canadian and Ontario Electrical Code books (many years ago I knew these well, but only have old versions of them now), and don't see anything about restrictions on using wires in parallel. Does anyone know of electrical code or other reasons for not using two #10 in parallel instead of one #8 conductor.

Conductors in parallel ( both grounded or ungrounded) only permitted in sizes No 1/0 and larger , unless a control circuit; CeC C22.1-15 , Rule 12-108, sub-rule (1) items (a) thru (f) and sub-rule(5). Minimum size conductors for bonding to ground determined by the size of the largest ungrounded conductor ( new in 2015- used to be based on O/C size) Table 16A; #8 copper ungrounded requires a #10 copper bonding conductor. Ampacity of the #8 conductor determined by table 2 (not more than 3 copper conductors in a raceway or cable) 75 degree column = 50 amps.( terminating on equipment marked with termination temperature rating of 75 degrees C ). The codes have changed quite a bit particularly the 2012 CEC and the 2015 CEC. Hope that helps.

 

[Mitchell]

Thank you all for your wisdom to make sure I do this right. No parallel conductors, I've here been authoritatively told.

I just picked-up the car a few hours ago, it is on trickle charge, which definately encourages me to get out to buy the (right) cable and get it installed.

The OpenEVSE has been ordered (sadly, it appears that it does not qualify for Ontario's EVSE subsidy program), I've posted my Leaf purchase details in "Re: 2016 Leaf Only - Pls Post Lease or Purchase deals."

 

[GlennD]

Once a socket appears NEC no longer applies. It only covers the wiring up to the socket. After that UL applies. That said, Paired conductors are quite safe since the ground is just used as a signalling path but the UL requires the conductor to be size same or one size less that the current conductors

That said, a cable like Tesla's or Quick Charge Power's does not meet the UL specifications. In practice, The EVSE will shut down when the ground current exceeds 20Ma( GFCI). This negates the large size requirement.

 

[Valdemar]

In practice, The EVSE will shut down when the ground current exceeds 20Ma( GFCI). This negates the large size requirement.

What if the GFCI is faulty?

 

[Mitchell]

Here you go. 50' of 8/2 w/ground
http://www.ebay.com/itm/MC-8-2-with-Ground-50-ft-Copper-Electrical-wire-MEATAL-CLAD-/200577575509?hash=item2eb35aea55:m:mVo-CBkY4seELjsw-aZGJLA

That would be perfect, but I'm in Canada and the shipping cost would be more than the cost. Good to know #8/2 armoured cable exists though.

 

[Mitchell]

Trying to decide whether to install heavier cable than needed to the EVSE ...

Since #8 solid copper wire has a resistance of 0.6281 ohm/1000' and #6 is 0.3952 ohm/1000', and my 50' run from the main panel to the EVSE would have a total of 100' of cable for the "electrons" round-trip, installing #6 would reduce the loop resistance from 0.06281 ohm to 0.03952 ohm. At 27.5 amps, this would reduce the power lost in the cable from 47.5 watts to 29.9 watts, eliminating 17.6 watts wasted during charging, which would reduce charging loss by (17.6/6600 =) 0.27%, which barely insignificant.

So the answer is it is only worth installing heavier cable for possible future requirements, it is not worth it just to reduce voltage drop and power lost in the feeder cable.

 

[ElectricEddy]

Trying to decide whether to install heavier cable than needed to the EVSE ...

Since #8 solid copper wire has a resistance of 0.6281 ohm/1000' and #6 is 0.3952 ohm/1000', and my 50' run from the main panel to the EVSE would have a total of 100' of cable for the "electrons" round-trip, installing #6 would reduce the loop resistance from 0.06281 ohm to 0.03952 ohm. At 27.5 amps, this would reduce the power lost in the cable from 47.5 watts to 29.9 watts, eliminating 17.6 watts wasted during charging, which would reduce charging loss by (17.6/6600 =) 0.27%, which barely insignificant.

So the answer is it is only worth installing heavier cable for possible future requirements, it is not worth it just to reduce voltage drop and power lost in the feeder cable.

The CM area of #8- 7 strand copper is 16,510 more than sufficient to carry the required 27.5 A over your required distance of 16 meters without even coming close to the 3% allowable voltage drop. The question is... what size are the lugs on the EVSE , will they accept #6 :?: