DC Source Output Conductors

Status
Not open for further replies.
C

Canton

Senior Member
Location
Virginia
Occupation
Electrician
Hello all,

I have taken over a PV install that was done by another contractor. The work was done by a "qualified contractor", I use that term loosely and it was not Permitted. That is why am there. This is residential also.

A few things that jump out at me while going through the job on the first sight visit. I have not done a PV install in about 5 years. At that time we had panels in series that added up to about 50VDC, this job has two grid-tied inverters one being fed with 2 PV stringers at 432 VDC 8 amps each and 1 stringer at 360 VDC 8 amps. (about 9,900 Watt) Why is the DC voltage so high and is it legal?

Next, there is no combiner box for all the PV panels on the roof but they are some how all series connected for the added voltage. I have yet to get on the roof to see. The DC Source Output conductors come into the attic from the roof and are changed over to Romex.....???? and ran in the exterior walls two floors down to the inverters in the basement. I thought DC source Output conductors had to be in an approved raceway...like conduit....?

Any thoughts would be appreciated....?
Thanks!
 
Canton said:
this job has two grid-tied inverters one being fed with 2 PV stringers at 432 VDC 8 amps each and 1 stringer at 360 VDC 8 amps. (about 9,900 Watt) Why is the DC voltage so high and is it legal?
Click to expand...

It is legal and it is generally desirable to run the DC voltage as high as possible to cut down on voltage drop and conductor sizes.


The DC Source Output conductors come into the attic from the roof and are changed over to Romex.....???? and ran in the exterior walls two floors down to the inverters in the basement. I thought DC source Output conductors had to be in an approved raceway...like conduit....?
Click to expand...

As far as I know you are correct unless something has changed again.

It would help if you told us what year NEC this job is under.
 
Canton said:
...
...
Next, there is no combiner box for all the PV panels on the roof but they are some how all series connected for the added voltage. I have yet to get on the roof to see. The DC Source Output conductors come into the attic from the roof and are changed over to Romex.....???? and ran in the exterior walls two floors down to the inverters in the basement. I thought DC source Output conductors had to be in an approved raceway...like conduit....?
Click to expand...
Under [2011] the outside conductors on the roof need to be in raceway or else be approved for weather and sunlight, such as USE-2 or PV wire (a new NEC/UL wire type)
Under most circumstances there needs to be a disconnect on the roof before the DC enters the house. The exception is 690.31(E), and using this exception requires that the conductors up to the remote disconnect be in metal (conduit or MC basically.) So check that out when you are on the roof.
 
I would say that the romex is only legal if it is on the inverter side of a readily accessible disconnect either outside or "inside nearest the point of entry". A roof disconnect is not typically readily accessible. An attic disconnect is unlikely to be.

Personally I think that running DC inside a building in romex is a bad practice and I would never do it, especially through an attic. MC cable maybe, romex no.

GoldDigger said:
Under most circumstances there needs to be a disconnect on the roof before the DC enters the house. The exception is 690.31(E), and using this exception requires that the conductors up to the remote disconnect be in metal (conduit or MC basically.)
Click to expand...

The NEC doesn't care if there's a rooftop disconnect, although many jurisdictions require it based on their fire code or some other ordinance. As far as the NEC is concerned, conductors inside a building must comply with 690.31(E) as far as the first readily accessible disconnect.
 
Canton said:
A few things that jump out at me while going through the job on the first sight visit. I have not done a PV install in about 5 years. At that time we had panels in series that added up to about 50VDC, this job has two grid-tied inverters one being fed with 2 PV stringers at 432 VDC 8 amps each and 1 stringer at 360 VDC 8 amps. (about 9,900 Watt) Why is the DC voltage so high and is it legal?
Click to expand...
All string Inverters use higher voltage of 250 to 550 volts. Yes it is legal and is how it is done.

Reason is real simple What size conductor would you need on a 9000 watt 50 volt source for a distance of 50 feet one-way with 3% or less voltage drop?

Was your answer 250 MCM copper cable?

Step the voltage up to say 400 volts only requires a 10 AWG wire.
 
dereckbc said:
All string Inverters use higher voltage of 250 to 550 volts. Yes it is legal and is how it is done.
Click to expand...
There are 1000V string inverters now, although one probably cannot use them for a rooftop system on a residence. DC in Romex through an attic, though... I don't think so.
 
ggunn said:
There are 1000V string inverters now, although one probably cannot use them for a rooftop system on a residence. DC in Romex through an attic, though... I don't think so.
Click to expand...

Why can Romex not be used for the DC output sources "After" the DC fused disconnect going into the house?
 
jaggedben said:
I would say that the romex is only legal if it is on the inverter side of a readily accessible disconnect either outside or "inside nearest the point of entry". A roof disconnect is not typically readily accessible. An attic disconnect is unlikely to be.

Personally I think that running DC inside a building in romex is a bad practice and I would never do it, especially through an attic. MC cable maybe, romex no.



The NEC doesn't care if there's a rooftop disconnect, although many jurisdictions require it based on their fire code or some other ordinance. As far as the NEC is concerned, conductors inside a building must comply with 690.31(E) as far as the first readily accessible disconnect.
Click to expand...

DC in Romex cable....bad practice...why?
 
DC in Romex cable....bad practice...why?
Click to expand...

See 690.31(E) in NEC 2011:

"Where dc photovoltaic source or output circuits...are run inside a building or structure, they shall be contained in metal raceways, Type MC metal-clad cables that complies with 250.118(10), or metal enclosures from the point of penetration of the surface of the building or structure to the first readily accessible disconnecting means."


[In NEC 2014, this moves to 690.31(G).]
 
Canton said:
Why can Romex not be used for the DC output sources "After" the DC fused disconnect going into the house?
Click to expand...

A fuse won't help; a dead short on a DC line feeding an inverter will never blow a correctly sized fuse in a disco.
 
SolarPro said:
See 690.31(E) in NEC 2011:

"Where dc photovoltaic source or output circuits...are run inside a building or structure, they shall be contained in metal raceways, Type MC metal-clad cables that complies with 250.118(10), or metal enclosures from the point of penetration of the surface of the building or structure to the first readily accessible disconnecting means."


[In NEC 2014, this moves to 690.31(G).]
Click to expand...

Read the article a little closer...690.31(E).....from the point of penetration of the surface of the building or structure to the first readily accessible disconnecting means......

So I would have my DC output circuits in conduit on the exterior of the exterior wall of the building....penetrate the building in conduit into the back of my DC readily accessible disconnect......then change over to romex and fish it through the walls to my inverter one floor down. How is this not compliant?


Also....690.31(A) would not be applicable here....."Where photovoltaic source or output circuits operating at maximum system voltages greater than 30 volts are installed in readily accessible locations, circuit conductors shall be installed in a raceway." Fished romex through a wall is not readily accessible.
 
Canton said:
You will have to explain this to me....?
Click to expand...

Simple. A PV panel or string of panels is a current source, with the maximum available current depending on the amount of light hitting the panel. Limited even into a short circuit.

The short circuit current is roughly 120% of the normal operating current. So if you size the fuse or breaker for 125% of the normal current, or 125% x 125% of the normal current, depending on your code cycle, then 120% of normal current will never activate OCPD. It might trigger either DC AFCI or DC GF detection, which are both part of the latest inverter requirements. Just not at the panels though, so it does you no good in protecting the DC wiring in the attic. :(
 
GoldDigger said:
Simple. A PV panel or string of panels is a current source, with the maximum available current depending on the amount of light hitting the panel. Limited even into a short circuit.

The short circuit current is roughly 120% of the normal operating current. So if you size the fuse or breaker for 125% of the normal current, or 125% x 125% of the normal current, depending on your code cycle, then 120% of normal current will never activate OCPD. It might trigger either DC AFCI or DC GF detection, which are both part of the latest inverter requirements. Just not at the panels though, so it does you no good in protecting the DC wiring in the attic. :(
Click to expand...

I understand this..... i did not include the whole quote. My reply was directed to another response, about DC voltage on romex cable, my response was directed to that previous post.
 
ggunn said:
A fuse won't help; a dead short on a DC line feeding an inverter will never blow a correctly sized fuse in a disco.
Click to expand...

I will clarify...there is nothing that prohibits Romex from being used to supply DC power...as long as installed as I described....also I have received approval from the AHJ.
 
Canton said:
I will clarify...there is nothing that prohibits Romex from being used to supply DC power...as long as installed as I described....also I have received approval from the AHJ.
Click to expand...

Well, if the PV System is of the ungrounded type, the white wire in 2-wire non-metallic sheathed cable, if used for the output circuits would be in violation of the code.

As for romex being bad practice as jaggedben said, IMO using romex for a high voltage DC run thru a dwelling is a disaster waiting to happen in that a lot of homeowners will attempt do their own electrical work. If renovations were done in the future and someone comes across that romex run, they just might try to add branch circuit wiring to that cable.

Your AHJ may have approved that installation. But if it were my installation, I would not be comfortable with the use of romex for the high voltage DC output circuits.

With that said, let this be a warning to you that using romex for the installation you have described could be hazardous to persons and property.
 
Make sure you review the "identification and grouping" and "circuit routing" requirements in 690.4(B) and 690.4(F), as well as the "marking and labeling" requirements in 690.31(E). Also, the requirements in 690.31(E)(2) are illuminating.

PV power source conductors are energized whenever the modules are illuminated. Further, the PV power source is current limited, and its output power is proportional to irradiance. So there are any number of scenarios where the PV circuit conductors could be damaged and exposed and still energized.

This is why the Code wants PV power circuits in metal conduit or similar when inside a building, and it requires that all the PV power circuits are identified and routed where they are unlikely to be damaged. They Code-Making Panel doesn't want fire fighters cutting into a building and contacting these conductors, and it doesn't want anyone else to inadvertently damage these conductors either. With an ac circuit, if you put a nail through Romex, you might trip the circuit breaker before you burn down the building. With a PV power circuit, that same nail could result in an uncontrolled arcing fault?and the fire and shock hazard can persist even if the inverter's arc-fault or ground-fault circuitry does its job.

Romex just doesn't provide enough physical protection to be suitable for PV power circuits. (Plus homeowners will assume it contains 120 Vac, as shortcircuit suggests.) If I inherited a job and found Romex used for PV power circuits, I wouldn't touch anything until I had the homeowner's permission to strip out that wiring and replace it.
 
Last edited:
Canton said:
You will have to explain this to me....?
Click to expand...
SolarPro already did, but perhaps a simpler way to put it is that the maximum current that the combiner can deliver to a short is less than the size fuse you would have to put in the disco. A dead positive to negative short in that Romex would just arc and burn and never blow the fuse, and nothing the inverter could do would stop it. I would never run DC through an attic without putting it in conduit. Actually, I would never run DC through an attic at all if I could avoid it.
 
Status
Not open for further replies.
Top