Weil-McLain P-368 Sight-glass level drop as steam pressure falls?

I made a paper reference scale and placed it behind the sight glass. The lowest mark indicates the LWCO cut-out level, which lines up perfectly with the painted line on the float body (thanks, @109A_5 ). After a blowdown of the float chamber and refill through the automatic make-up valve, the float was satisfied at the LWCI mark, shown slightly above the water level, probably due to cooling. Lastly, the highest mark indicates the normal operating water level.

My experience is that the apparent water level is repeatedly drawn below the LWCO mark until the water level in the boiler eventually reaches the normal operating level. I've deliberately throttled the make-up water supply to a very low flow rate to prevent overfilling during these unusual drawdowns.

Boiler Water Level Experiment – Observations and Results

Following the recommendation of @109A_5, I conducted an experiment to investigate the unusual water-level behavior in my boiler, particularly the drop in the boiler sight glass during pressure decay that occasionally triggers the LWCO.

Experimental Setup

I installed a 0–3 psi pressure gauge on a riser manifold containing two Gorton No. 1 main vents and a fitting connected to the top of a temporary PTFE tubing sight glass. See picture below.

At the bottom of the Hartford Loop, I installed a 3/8" ball valve and compression fitting connected to the bottom of the temporary sight glass. This arrangement allowed me to observe the water level in the external column simultaneously with the boiler sight glass.

I also closed the automatic makeup-water supply valve to ensure that no additional water could enter the boiler during the experiment.

Additional Notes:

I flushed the boiler @patrykrebisz but did not a charge of 8-way treatment @mattmia2 in keeping with good recommendations from the forum.

Observations

1. At rest: The water level in the experimental column was identical to that in the boiler sight glass.
2. During initial steaming: Once air had been vented from the steam main, but with both pressure gauges still reading 0 psi, the boiler sight-glass water level began oscillating around its initial level. Meanwhile, the water in the experimental column dropped approximately 2.5", settling near the midpoint of the top of the Hartford Loop.
3. As boiler pressure increased: The water level in the experimental column dropped an additional 1", for a total displacement of approximately 3.5", while the boiler sight glass continued to oscillate around its operating level.
4. During pressure decay: As boiler pressure decreased from approximately 1.5 psi, the water level in the boiler sight glass dropped to the bottom of the glass, activating the LWCO. This reproduced the original phenomenon I was attempting to investigate. Interestingly, the water level in the experimental column remained relatively steady at approximately 3.5" below its original level. This effectively broke the connection between the boiler water and the wet side of the Hartford loop.
5. Pressure readings: At no time during the experiment was there an observable pressure differential between the two 0–3 psi pressure gauges (predicted by @ethicalpaul.)
6. Recovery: After boiler pressure reached zero, the water levels in both the boiler sight glass and the experimental column gradually returned to their original levels over approximately one minute. Since the makeup-water supply was closed, no additional water entered the system. Once the boiler water level recovered, the LWCO reset and permitted the burner to fire again.

Observations and Possible Conclusions

The experiment demonstrated that the boiler sight glass and the external water column behave quite differently during steaming and pressure decay, even though both return to their original levels once pressure dissipates.

The fact that both water levels recover without the addition of makeup water suggests that the phenomenon involves temporary water displacement or redistribution (suggested by @mattmia2.)

Although both pressure gauges indicated identical pressures throughout the experiment, I recognize that relatively small pressure differences might not be detectable on these gauges. For reference, a 3.5" water-column displacement represents approximately 0.13 psi; these gauges are not that precise.)

I know it's not very scientific to say that I feel like the volume of water is drawing closer to the riser and away from the sight glass. I've long felt this was the behavior but could never test it. @mattmia2 also suggested, "there are some theories about steaming tilting the water line in the boiler" so for now he has the leading theory.

If anyone can draw a more definitive conclusion from these results or recommend additional diagnostic testing, I would appreciate your thoughts.

Thanks!