Best Of

The water is almost certainly leaving the boiler as liquid water either through surging or priming throwing it in to the mains.............

I'll say it again, the water is leaving the boiler as liquid not steam, you need to figure out how and why.

This is correct.

Although the boiler is piped per the installation manual with 2" permitted as the riser and header, the steam velocity becomes 45 ft/sec

This boiler should have a minimum of a 2.5" riser (ideally two risers) and preferably a 4" header to slow the steam and allow the water to drop out of suspension.

Notice there is absolutely no time for the water to drop in the very short distance between the top of the riser and the takeoff for the main.

Also, notice that the mains appear to be considerably larger than the riser.

Water is headed up to the radiators in considerable amounts.

That is a #42 which the "newer" #44 replaced I believe. That I would guess could be 1930s

All those "Mills" boilers with the top and bottom headers were very efficient due to their down draft design. 80% or so would be easy to obtain.

Problem is it is probably massively oversized.

We used to pull the two bottom return headers off clean the muck out and re install them with new nipples. It's quite a job to do that and I did more than I care to admit.

The other problem is the cement and rope between the section would degrade and fall out causing air to leak into the combustion chamber lowering the efficiency.

One advantage they had that Smith used to tout is if one section got cooked or leaked you could cut the supply nipple and two return nipples to the affected section and remove the nipples and put pipe plugs in the headers and continue running the boiler with one less section.

Don't know what size the old steam main coming off the boiler header is but it looks to be 6" screwed i am guessing which also shows the age of the boiler

you may need to add a swing elbow to the header as it drops into the tee feeding the existing header to provide the correct pitch. The new header needs to drain into the tee just as the old header is pitched to drain into the current elbow where the new tee should go. Feeding the steam from above to allow the condensate to drain below with minimal colliding is key in a counterflow system that doesn't have individual drip return lines on each main. That's why Pg. 28 of the boiler manual shows it that way.

Ideally you'd do what @New England SteamWorks shows on one of their diagrams and get a dripped return line for each direction of the main which allows for a header constructed like a typical parallel flow system instead of the fed from above tee configuration shown in my sketches and the boiler manual. NESW's configuration with drips off each main keeps the returning condensate from the mains from having to re-enter the header at all.