Number of bends between pull points

[kenaslan]

There has been some discussion here and there on the maximum number of bends in any conduit run. I believe that we all agree that it is 360 degrees or 4 quarter bends. The code states "There shall not be more than the equivalent of four quarter bends (360 degrees total) between pull points, for example, conduit bodies and boxes." IMO The definition of a "pull point" is a box or conduit body with a removable cover. No removable cover, no pull point. There are some that suggest the idea that you can unscrew a FLMC connector and pull the wire then put it back together again meets the intent of the code. I believe that is wrong. The amount of bends in the FLMC MUST be added to the conduit run in order to get the correct number of bends. Once a box or conduit body is encountered we restart the count toward 360 degrees again. Has anyone found a NEC definition for "Pull Point"? Anyone for the Code Panel have any insights? I have a "constructability manager" with the responsibility is to build it as cheap as possible. They believe that because "pull point" is not defined in the code, they can define it any way they want.

 

[infinity]

There are some that suggest the idea that you can unscrew a FLMC connector and pull the wire then put it back together again meets the intent of the code.

I can loosen an EMT coupling and pull it apart too but that doesn't qualify as a pull point (been there done that).

I think that you've properly answered your own question.

 

[don_resqcapt19]

There has been some discussion here and there on the maximum number of bends in any conduit run. I believe that we all agree that it is 360 degrees or 4 quarter bends. The code states "There shall not be more than the equivalent of four quarter bends (360 degrees total) between pull points, for example, conduit bodies and boxes." IMO The definition of a "pull point" is a box or conduit body with a removable cover. No removable cover, no pull point. There are some that suggest the idea that you can unscrew a FLMC connector and pull the wire then put it back together again meets the intent of the code. I believe that is wrong. The amount of bends in the FLMC MUST be added to the conduit run in order to get the correct number of bends. Once a box or conduit body is encountered we restart the count toward 360 degrees again. Has anyone found a NEC definition for "Pull Point"? Anyone for the Code Panel have any insights? I have a "constructability manager" with the responsibility is to build it as cheap as possible. They believe that because "pull point" is not defined in the code, they can define it any way they want.

First, because of incorrect code language, the degrees of bend reset every time you change the raceway type. This is not how the code is, or even should be enforced, but just an effect of the code language. As a result of that poor language will appear in the 2026 code.

300.24 Bends.
The total degree of bends in any combination of wiring methods used for a single run shall not exceed 360 degrees between pull points.

(below is the substantiation for that change)

Having too many bends between pull points could make pulling wires through a raceway difficult, if not impossible. It could also lead to damage of the conductor insulation or cable jacket. There is presently no code rule limiting the number of bends between pull points where multiple types of raceways are joined together between those pull points. A run of RMC is not permitted to have 1080 degrees between pull points. A run of IMC is not permitted to have 1080 degrees between pull points. A run of PVC conduit is not permitted to have 1080 degrees between pull points. Each one of these raceways is individually limited to no more than 360 degrees of bends between pull points, but a run of RMC having 360 degrees of bends joined to a run of IMC having another 360 degrees of bends joined to a run of PVC conduit having another 360 degrees of bends IS permitted and would have 1080 degrees between pull points! Joining different wiring methods together is not a violation. A pull point such as a box or conduit body is not required at these transition points. A coupling could be used here in accordance with 300.15(F). To protect conductors and cables from being damaged, a limit needs to be placed on having too many bends between pull points when different wiring methods are joined together.

That being said, I know of no one that adds the bends of LFMC to that of the raceway that the flex connects to. In fact, if a 90 degree connector is used on the flex, it is almost impossible to pull through the flex and its connectors.
I have never pulled through any type of flex to the final connection point unless the run was very short and the flex only had straight connectors. Often the wire has been pulled in the raceway before the flex is even installed.

 

[Seven-Delta-FortyOne]

I think that because 300.18(a) requires the conduit to be completely assembled before pulling conductors, that would not be a permitted definition of a pull point.

While that section is commonly violated, especially when using flexible conduit, I would think that would not be an acceptable definition.

 

[don_resqcapt19]

I think that because 300.18(a) requires the conduit to be completely assembled before pulling conductors, that would not be a permitted definition of a pull point.

While that section is commonly violated, especially when using flexible conduit, I would think that would not be an acceptable definition.

Some times the code has to be changed to match the real world installation practices. If it is not changed, the rules are just ignored.

 

[wwhitney]

In fact, if a 90 degree connector is used on the flex, it is almost impossible to pull through the flex and its connectors.

I seem to recall some text in the NEC that was explicit that 90 degree connectors to be used only at exposed terminations, and that they may be used by pulling the conductors into the rest of the run, and then putting the 90 degree connector on. If that's correct, you can argue that the run-side of the 90 degree connector is a pull point, so you could have a box to box run that has 90 degree connectors on each end and 360 degrees of bend in between the connectors.

Cheers, Wayne

 

[LarryFine]

I agree with Wayne here. They can't require angle connectors to be accessible in case they need to be re-wired, then say that you can't depend on them being dis-assemble-able to be wire-able.