Regarding Russ's comment:

"When everything worked as intended they functioned well. They did have a few quirks that took some getting used to though. If, for some reason the cold water pump lost its prime the fireman had to crawl out on the running board and open the vent cocks on the pump to get it to pick up again. Not fun at track speed. When the heater was working, the heating chamber was operating at whatever the cylinder back pressure was at the time with the water being heated to something over its boiling point. Any change the engineer made that reduced back pressure, such as reducing the throttle or even hooking the reverse up a notch, would cause the heated water in the heating chamber to flash to steam and the hot water pump would be trying to pump steam rather than water. This caused the pump to race until the heating chamber filled with hot water again which, for some reason, often caused the cold water pump to lose its prime. Not something you wanted to have to deal with while climbing the 3% grade to Apex."

Interesting point about the heating chamber losing pressure and momentarily flashing to steam. Was this an issue with the S-types?

-Erich

Probably, as they were a direct type heater as well. I suspect that the hot water pump would race until the water level in the heating chamber was restored. However, since the cold water pump is a separate unit it would not be affected and would continue to operate normally.

I think the SA style systems were the standard FWH on modern power. Everything from the UP's 844, N&W 611, most of the berks, NYC 4-8-4s all had some version of the SA system.

They might have been the final development of the FWH system. Other than hearing about float valve issues, I've thought the SA's were pretty reliable.

An advantage that an SA has over the BL is that there is a greater vertical distance between the heating chamber and the hot water pump. This would give the hot water pump a little more suction head to keep the water from flashing into steam at the pump. As long as some water remained in the chamber, I would think that the pump would not suffer a loss of suction.

That is not characteristic of the boiler on 60000. You're thinking of the Benson or once-through type of watertube boiler; the 'watertube' part of the Baldwin locomotive boiler uses a waterwall structure of tubes between a mudring-type structure at the bottom and a pair of drums at the top, essentially mimicking the function of a conventional staybolted waterspace between inner and outer wrappers. The elimination of staybolt issues allowed the modest pressure increase (to 350psi) to be practical, but there is no need for either forced or directed circulation. I believe the Emerson fireboxes on B&O, while of different geometry, ran on the same principle and did not have circulation pump requirements, and the McMillin boilers on NH a decade earlier likewise did not require pumping.

It would have been possible to use some variant of a Lamont firebox, which uses circulation pumps rather than a full boiler feed pump. Here the waterwall is provided in discrete tube circuits, through which water is pumped at about 6x steam-generation mass flow, with proper 'scrubbing' of tube walls, lower chance of DNB or Leidenfrost issues at high radiant heat uptake rate, etc. BUT there is no nominal pressure differential seen by the circ pumps; they only have to overcome the flow resistance, so have to be only a few hp at most for 6000hp's worth of steam flow. The flow from the tubes passes directly into relatively tall cyclone-like separators, where the water spirals down and the steam goes up; there is little relative chance for priming. These will not run at all if the circ pump goes out, but as noted only very small head is required and making robust pumps is therefore easier; I do not know if the jet pumps that run a Cunningham circulator are 'good enough' for a multiple-circuit Lamont at usual eight-coupled locomotive scale.