I see many electrical safety tips that caution against using Romex wire for extension cords, but never the reason. I'm guessing the reason is that the insulation is not durable to fit the grade, right? If not what is the reason? What type of insulation should an "approved" extension cord be constructed from? many thanks

The short answer is in the name. Extension cords should be made from "cord", not cable.
Since this is usually "hard service" it should be some kind of hard service cord.
Most OSHA inspectors have trouble with any field built extension cords anyway. If you do make one, be sure you have all listed parts, used in the way they are listed (no handy boxes with a Romex connector)

At one time, I was employed to do some tests on sundry connectors that were intended to be used with cords. One of the test was to hang a weight from the cord, to see if the connector did actually prevent the strain from being transmitted to where the wires were connected.

I did test several such connectors on Romex, rather than 'extension cord' material. In every instance, the connector failed to hold on to the romex well enough. That is, the outer jacket would work free of the connector, and you'd be left with the wires holding the weight.

Another consideration is the romex has solid conductors and they are not especialy suited to frequent flexing. The jackets certainly are not meant for rough duty either.

I think that is much the same reason OSHA frowns upon the cord end replacement caps you can put on. After a while of use, they tend to work themselves loose from the cord sheath and expose conductors also.

It is common to run NM for long temp extension cords where the ends are not moved frequently.

The only reason that NM is not approved for extension cords is that is is not tested for that purpose.

On the cord vs cable naming issue: I recently used Bus Drop CABLE for a heavy duty "extension cord" from a PTO generator to my house - the setup had a interlocked breaker installed in the main panel and the whole shebang was passed by the ahj. My point: Bus drop CABLE was ok to plug and cord connect the PTO generator to the house. BTW I agree with the comments on strain releif using NM.

That one's a classic... I just took two extension cords made with German NYM out of service at the local church.

I'd say it's the shape and flexability issues i.e. it pulls out of connectors too easily and is pretty useless as an extention cord because it's too inflexible, plus I doubt that the solid wire has as high a pull-out force from the connector itself, though this is just conjecture. I think that the romex jacket durability v.s. extention cord durability is a fairly minor issue.

-Will

This is one my wife took away from a trade ... wait for it ...
Electricians!

Romex wicks up moisture. That's a real threat to users of temp power -- where you can just bet that it's an issue.

Romex is acceptable as temp power conductors if and only if it is treated as Romex and not SO cord, etc. That means NMC if water is an issue.

Dragging Romex around like it's SOJ... sheesh -- pure trouble.

Figures. But isn't it often true that:

The plumber's house has leaky faucets.....



The mechanic's car/truck is a piece of junk....




The painter's house needs painting...




And the electrician's house has broken/bad devices/fixtures?


And in my case, I have the only home automation system (AMX touchscreen) where the programming still isn't finished (after three years.) But all my client's programs are up and running.

Back to the main topic.. Romex does make lousy extension cord. And somewhere here on ECN there is a thread where someone used a coiled-up box of romex and it melted down!

Finally on one of my last major installs they used 10/3 NM for the temp lights and it was constantly getting cut/damaged by the trades.

We see the same thing here where our equivalent of Romex has been used as an extension cord. The solid conductors just aren't intended to take the frequent flexing and after a while they fracture at the points which have seen the most movement.



You could do that with flexible as well if you pull enough current through the coil that the heat can't dissipate properly.

I guess that every trade has the same habit of doing better work at work than at home.

At one jobsite meeting that I was at, the woman in charge of estimating costs and schedule blew up at her electrician boyfriend yelling:
"You've RUINED me in this business! Now when I get asked how long an electrician takes to install a doorbell button; I have to tell them 6 MONTHS!!!"
The rest of the meeting went much quieter...