Running control wiring in power conduit for EV Charger amperage control. 725.26(B)(1).

[brycenesbitt]

I'm looking for some help with NEC code interpretation wiring associated with EV charging.
This is for a series of "Wallbox" brand EVSE chargers which have an RS-485 control circuit.
In short they measure the current at the main using CT clamps, and then throttle the vehicle's charging rate to moderate peaks.



The question is where to run the control wires. The two options are:

• Place the CT clamps on the service entrance cable. Create a separate box and conduit channel.
  Drill a hole in the main panel to run the CT clamp sense wires out.
  Thus the low voltage CT clamp wires will exit the main panel as quickly as possible, then rejoin the power supply cables at the Wallbox.
  
  
• Run everything within the same enclosure and conduit. The ct meter DIN rail mounted and requires the low voltage CT sense wires, and access to 240V power.

Wallbox is an Intenational company, and they assume for Europe presumably that all the wiring will happen in the same conduit.
I'd prefer to keep all the wiring in the same enclosure as it's neater, cleaner, cheaper. In my view it's less subject to future damage if integral.
I'm not so sure I can make that fly under the NEC. These wires are clearly "functionally associated"

I see:
Q. I’m working on an electric gate installation. Can I install control wiring in the same raceway with electrical power conductors?

Mike Holt's A: Maybe. According to 725.26(B)(1), Class 1 circuits can be placed in the same cable, enclosure, or raceway with power-supply circuits, if the equipment powered is functionally.

Install location is California, which will switch soon to the 2022 NEC. I can readily source appropriate 600V rated twisted pair wire with shield and 600V sheath.
The control wire is NOT separate from the 240V wiring inside the Wallbox:



Your thoughts on what it takes to make this work safely and satisfy most AJH's?
Can I place a foam sleeve on standard RS-485 twisted pair cable to meet the 1/4" separation requirement for example?

 

[hbiss]

Wallbox is an Intenational company, and they assume for Europe presumably that all the wiring will happen in the same conduit.

So, is that thing even listed for use here? I think not.

-Hal

 

[tom baker]

European wiring is typically not run in conduit, it’s cables in cable tray or similar. Notice how the wiring box has two cable glands. And ditto on Hal’s post. CE means nothing in the US

 

[brycenesbitt]

So, is that thing even listed for use here? I think not.
-Hal

Yes UL listed for USA and Canada, EU/CE and China as well.
The instructions are not well localized for the NEC and USA, but the product is clearly listed.
The photo is the plug in model (40A) there's a hardwire variant set up for conduit on the left side (48A).

All EV charge management systems have to deal with the communication issues. Powerline communication has been tried and abandoned. Others use wifi, but that's a pain as you have to maintain the wifi system over the decades of service life of the equipment.

UL even wrote about their Wallbox contract at https://www.ul.com/resources/powering-north-american-market-access-wallbox-and-ul
Certs listed there are:
UL 2594, the Standard for Electric Vehicle Supply Equipment
CSA C22.2 No. 280, Standard for Electric Vehicle Supply Equipment
UL 2231-1, the Standard for Safety for Personnel Protection Systems for Electric Vehicle (EV) Supply Circuits
CSA C22.2 No. 281.1, the Standard for Safety for Personnel Protection Systems for Electric Vehicle (EV) Supply Circuits
UL 2231-2, the Standard for Personnel Protection Systems for Electric Vehicle (EV) Supply Circuits
CAN/CSAC22.2 No. 281.2-12, the Standard for Personnel Protection Systems for Electric Vehicle (EV) Supply Circuits
UL 1998, the Standard for Software in Programmable Components
UL 991, the Standard for Tests for Safety-Related

 

[jaggedben]

Your description is confusing, but it seems to me you have CTs that go to a meter component which also needs power/voltage reference, and then you have a separate RS-485 comms circuit. The rules for these two things are not the same.

CT leads are not addressed in the code as far as I know, and I see no reason they can't be run together with circuit conductors to the meter.

The RS-485 can only run in the same conduit as the power if you use a Class 1 wiring method. No other circuits can run in the same conduit, only the meter and/or Wallbox power which are 'functionally related'. The RS485 circuit should not pass through the panel.

If Wallbox dose not provide any direction on a cable for the RS-485 that is listed for use in the same conduit, you'll need to pull conductors that comply with 725.49(B). If that doesn't meet your functional requirements (e.g. twisted pair), you'll have to run the RS-485 separately. Note it doesn't have to be in conduit.

See 725.48(B)(1), and 725.130(A) Exception 2.

2020 NEC references.

 

[tortuga]

Regardless of code RS-485 needs to be in a separate conduit or you get hard to find packet errors.

 

[texie]

I think we need some clarification here. If the factory supplied CTs have an RS-485 output, then these are not CT secondary conductors. The CTs have onboard electronics to create the RS-485 output. Different rules apply for CT VS RS-485 conductors.

 

[brycenesbitt]

Regardless of code RS-485 needs to be in a separate conduit or you get hard to find packet errors.

@tortuga that's not generally true. The RS-458 wires are twisted pair differential thus the 60hz and other noise is largely cancelled just by the twisted pairs. A foil shield helps also. And a higher level protocol (in this case Modbus) handles packet integrity and retries, much in the manner of TCP/IP and UDP.
The wireless alternatives, such as as WiFi, have far higher packet error rate.
And finally in this application the RS-485 wires terminate inside the EVSE enclosure regardless, so will be exposed to some of the 60Hz and EV DC converter switching noise regardless. I realize that's less than running both wires in parallel but still it's got to be robust regardless.

 

[brycenesbitt]

I think we need some clarification here. If the factory supplied CTs have an RS-485 output, then these are not CT secondary conductors. The CTs have onboard electronics to create the RS-485 output. Different rules apply for CT VS RS-485 conductors.

In this case, the CT conductors are analog. The CT itself could be a CT, a fixed Rogowski coil or a flexible Rogowski coil, but regardless an analog low voltage signal.

An associated electronic device takes the secondary CT conductors and converts them to digital, presenting them via Modbus over RS-485.

A similar format consumer device is the Emporia Vue, which produces a WiFi output signal as an Internet of Things (IoT) device:


The flexible Rogowski coil is used when there's no wire, just a busbar available.

Regardless of the style and wiring of the current sensor, the job is to get those readings reflected in the pilot signal of the J1772 connector at the electric vehicles.

 

[hbiss]

A similar consumer device is the Emporia Vue,

Wow! Looking at that mess in a panel I can't see how that would be allowed. And it probably isn't. CT leads need to go outside. Why do you think they are so long?

-Hal

 

[wwhitney]

An associated electronic device takes the secondary CT conductors and converts them to digital, presenting them via Modbus over RS-485.

Yes, but the question was whether that device is in the enclosure with the CT, so that the conductors in the conduit are RS-485, or at the other end, so that the conductors in the conduit are the secondary CT conductors.

Cheers, Wayne

 

[brycenesbitt]

The RS-485 can only run in the same conduit as the power if you use a Class 1 wiring method.
See 725.48(B)(1), and 725.130(A) Exception 2.
2020 NEC references.

Yes, this is exactly where my code reading skills are hitting the limit.

Article 725.49(B), 2008, says that for Class 1 circuit conductors, "Insulation on conductors shall be suitable for 600 volts. Conductors larger than 16 AWG shall comply with Article 310. Conductors in sizes 18 AWG and 16 AWG shall be type..."
In this case the RS-485 conductors can be 18 or 22 gauge, shielded by a foil, and protected by a 600V jacket and additionally enclosed in a sheath.
-
The other Internet references on Class 1 vs. Class 3 get really muddled and I'm left without clear code interpretation, which is a worry.
-
An example cable is:
HW106 UL Type TC-ER direct burial RS485 cable which would also be NEC Type NPFL and used in fire protective circuits.

 

[hbiss]

The photo is the plug in model (40A) there's a hardwire variant set up for conduit on the left side (48A).

Ok, so what's the problem? Looks like they Ty-Rapped the data cable to the conduit.

-Hal

 

[brycenesbitt]

Yes, but the question was whether that device is in the enclosure with the CT, so that the conductors in the conduit are RS-485, or at the other end, so that the conductors in the conduit are the secondary CT conductors.
Cheers, Wayne

That's part of what needs to be figured out, I think. Let's try a diagram:


The CT clamp must be in the main panel or in the gutter between the main breaker and the meter.
The conductors from the CT clamp to the modbus meter must be short.
The modbus meter is most conveniently located nearby if not right in the gutter, and must be fed 240V electrical power to operate (it need not be the same circuit a powers the EVSE). The RS-485 can be optically isolated if necessary.

 

[brycenesbitt]

Ok, so what's the problem? Looks like they Ty-Rapped the data cable to the conduit.
-Hal

Running the twisted pair external to the conduit is an option. This will however create a small potential for an electrical fault within the EVSE enclosure to place 240VAC on the data cable. If the data cable is inside the conduit, nobody can mess with it, nobody can touch it, nobody can cut it, nobody will paint it or Ty-Rap other crap to it. If the data cable is inside a conduit in the conduit (multiduct or tube in tube) even better :=)

 

[LarryFine]

View attachment 2568100

Shouldn't those modules be outside the enclosure?

 

[brycenesbitt]

Ok, so what's the problem? Looks like they Ty-Rapped the data cable to the conduit.
-Hal

Sucks when the parking spot is supplied by conduit under a concrete driveway....

 

[hbiss]

The CT clamp must be in the main panel or in the gutter between the main breaker and the meter.
The conductors from the CT clamp to the modbus meter must be short.
The modbus meter is most conveniently located nearby if not right in the gutter, and must be fed 240V

See, that's the difference between the US and European countries. With their little plastic panels (consumer units) and the way they are allowed to wire things that's easy to do over there. Here, our equipment and material is much more "robust", and our rules are much more stringent.

I read through the UL paper in the link you posted. All I see is concern with component selection and safety (rightly so) but, unless I missed it, nothing regarding installation for the US market- which is what you are asking about. That should have been addressed by UL also.

This shouldn't be something that you need to "engineer" in order to install it. That issue should have been addressed by the manufacturer.

As a footnote, the lighting industry solved the problem of running low voltage (0-10V dimming) with the line voltage by getting the cable manufacturers to make Luminary cable. Luminary cable has a pair of low voltage conductors isolated from the line voltage conductors under one jacket. It's available in NM and MC versions.

That said, I know they make CAT-5e and CAT-6 cable with a 600V rating for industrial control panel wiring. So if you can find a (single pair shielded?) RS-485 cable with a 600V rating you could run that with your power conductors.

I can readily source appropriate 600V rated twisted pair wire cable with shield and 600V sheath.

-Hal

 

[brycenesbitt]

As the OP here: I am seeking ways to keep everything in the enclosure.
Meaning seek solutions for the raised basic electrical safety issues, code issues, confusion issues for future maintenance, any operational issues and risks.
I expect this to be hard.
-
My thesis is that going in the conduit is more sustainable long term: less to maintain, less potential for damage to the communication wires over a 10-15 year service life, easier routing when conduit is pre-existing or under concrete. Just UV damage damages cables and cable ties even the UV resistant type. So I'm willing to invest 5-6x the wire cost in low voltage cables, if they can be pulled rather than Ty-wrapped. If needed I would be willing to pull a separate tube then pull communication cable to meet the NEC's 1/4" spacing requirement: though that's not needed from the telecom noise argument.
-
The CT clamps are unfortunately not negotiable. The local meters here are not all capable of sharing load data. Long term it may be possible to use Zigbee communication from the meter rather than an added CT clamp. But this makes the install specific to the meter or panel, and requires IT skills to set up. The advantage of the CT clamp approach is I could leave a circuit diagram which would clearly communicate to future maintenance personnel how the system works -- no app or Bluetooth or tech required. Any future maintenance person can whip out a clamp meter and verify the system's integrity.
-
This system is way better than what others do with the Dryerbox and Stove sharing setups, which sense the dryer or stove is operating, and cut power to the EVSE. The CT clamp approach promises to be a lot more elegant, maybe even cheaper, and greatly reduces the needed wire lengths.
-
So what would it take to get 18 gauge communication wire in the same existing conduit as 40A 240V power delivery? What novel safety issues does this raise, that don't exist with dual conduit between the same locations?

 

[hbiss]

So what would it take to get 18 gauge communication wire in the same existing conduit as 40A 240V power delivery?

I just told you. A cable rated at 600V! You said you had a source for it didn't you?

-Hal