I had a conversation about steam purge over a hot dog grill this past summer with a neighbor who does I&E contract work at a process plant. Steam pressure wasn’t as high (20 Bar?) but same deal - how to purge DP transmitter impulse lines with steam (crystals form and block in the impulse lines).

We both thought that the 'condensate level remains constant because of the cooling effect of the wet/drop legs' premise is called into question when the impulse lines are no longer dead-headed, but now have live steam flowing through them. There must be some amount of condensate flashing at the water-steam interface. Does the system 'equilibrate' at some stable wet leg condensate level that is reliably constant? Don't know.

And there's an issue of the availability of a purge steam supply pressure higher than the line pressure.

Then there’s the issue of a relatively low-flow constant flow regulator. If a Brooks style integrated constant flow-rotameter is not available at that pressure/temperature (I don't whether one is or is not), then we talked a constant pressure regulator with an restriction orifice or a low-flow (needle?) valve, again, assuming supply steam pressure greater than line pressure.

Some months later he told me that they did not try the steam purge, they went with one of those auto-rodder units to clear the impulse lines. So all I ended up with was a picnic discussion, no answers. Hopefully someone here on the forum knows and can fill you in.

Purge flows in impulse lines must be low and kept constant so as not to affect the readings. I do not see this happening with steam except under very special conditions where there will be no condensation occurring. The lines must be kept full of liquid or be kept clear of liquid.

Looking at my steam tables it's obviously superheated steam.

I would just treat it as a normal blow-down, drop off the line and go down a few feet then tee off and slope back up into the transmitter with a blow-down valve at the bottom.
When you blow it down the vertical pipe is going to grow in length so the slope up section should be tubing with enough bends in it to compensate for the growth otherwise you might see the tubing pull out of the fittings which would not be nice. You could even replicate a pigtail syphon in the tubing to ensure that the transmitter is not subject to steam temperature,

I would also include a condensate pot to assist in refilling the blow-down with condensate.

Hope this helps