Use of B31.4 for temperatures belwo -30 degrees 1

You use A106 to at least -48 deg C if you can use Figure 323.2.2B and get the weld metal and HAZ to meet a Charpy test.

It is less hassle to use a low temp material like A333 instead of A106 grade B.

"Was not developed" are the key words.
Use A333 as bimr suggests.
I have done the same for liquid anhydrous ammonia pipeline at -33C using A333.

It seems that you go into process piping area with the temperature.

If you are building a pipeline, you don't have the choice of changing codes (will you drop the pipeline markers, will you drop the burial depth provisions, etc), although you could incorporate various aspects of B31.3 if appropriate, which doesn't mean you are stretching the B31.4 code all that much by simply lowering the temperature, as long as the material is used within its limits.

Good luck at moving oil at -33C; light ends (gasoline) no problem. Are you misinterpreting the low ambient temperature as the MDMT?

I posted a reply which got lost, but my point is where do you stop? -40, -50 ??

Also B 31.4 has another section - 401.2.3.7 which strongly implies that outside the -30C / 120C range the design stress or design factor would need to be modified. Sure you can get material to work down to temperatures outside those limits, but maybe not at a DF of 0.72??

31.3 would result in a higher wall thickness, but work outside the written boundaries of a code and you are in gradually increasing murky waters.

The real issue is is that a temperature the pipe will actually see.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

Cold isn't as dangerous as hot, as axial loads are usually tension, as are hoop stress, so both being on the same side of Mohr's circle means lesser total shear stress. The tendency for buckling is also greatly reduced. It definitely stops at the material's range, or wherever the analysis gets fuzzy.