Gauge Glass Says Two-Thirds, Low-Water Cutoff Says Shut Down
A 400 hp, 125 psig firetube boiler firing natural gas supplies steam to a campus loop. Condensate return runs about 60%, makeup is softened, and feedwater enters at roughly 200°F. Over the past week the burner has tripped on low water three or four times per shift, always at high fire. The operator resets the control each time and the boiler relights normally. During each trip the gauge glass shows a level at about two-thirds, well above the low-water cutoff mark, and the feedwater regulator indicates 40% valve position. The low-water cutoff is a float type mounted in its own chamber off the front head; the gauge glass is on a separate water column with its own upper and lower cross-connections. Yesterday the operator blew the glass down for about 20 seconds; the glass stayed empty for nearly half a minute before water slowly climbed back, ending slightly higher than before. Steam header pressure is also hunting about ±10 psi. Nothing has been changed recently on the feedwater or burner controls.
1. Given the tripping pattern and the way the glass behaved during blowdown, what is the most likely fault, and what should the operator do in the next ten minutes?
Read the worked answer
The slow refill is the tell. A healthy gauge glass refills in two or three seconds once the blowdown valve is closed, because water enters through the lower cross-connection almost immediately. Thirty seconds of slow climbing, ending slightly high, means the lower cross-connection is largely plugged with sludge or scale, and the glass is being fed only from the top connection by condensing steam. A glass filled with condensate looks exactly like a glass filled with water, and it will sit at whatever level the condensate accumulates - here, about two-thirds.
That explains the apparent contradiction. The low-water cutoff chamber is a separate, still-open tapping, so the float is reading the true boiler level, which is genuinely falling below the cutoff point. The cutoff is doing its job; the glass is lying.
In the next ten minutes: stop resetting the control. Reduce firing rate and load, confirm feedwater is actually entering (pump running, discharge valve open, check valve not hung, regulator not air-locked), and cross-check level with try-cocks if the boiler has them. Then shut the burner down and isolate the water column for inspection. Do not bypass, jump out, or hold in the low-water cutoff to keep steaming - that is how a boiler gets overheated and ruptured.
2. A colleague argues that because the glass shows water, the low-water cutoff must be faulty and the reset is therefore justified. Evaluate that reasoning.
Read the worked answer
The reasoning inverts the burden of proof. Two independent level indications exist - the glass and the cutoff float chamber - and they disagree. Nothing in the symptoms says which one is wrong. The blowdown performance is what settles it: an unrestricted lower connection cannot take half a minute to refill a glass. So the glass has a demonstrated physical defect, and the cutoff has none.
There is also an asymmetry of consequence that should drive the decision. If the cutoff is right and is overridden, the boiler continues to fire with the tubes or crown sheet uncovered; the metal overheats, loses strength, and can fail catastrophically. If the cutoff is wrong and the boiler is shut down, the loss is a few hours of steam. Between those two outcomes, the operator shuts down and investigates.
The steam pressure hunting is consistent with the same root cause: periodic burner trips and relights, plus a feedwater regulator chasing a false level, produce swings. It is not independent evidence of a control fault.
A final point worth stating: the correct response to any low-water trip is to verify level by a means independent of the device that tripped, then reset per the manufacturer's procedure. Repeated resets with no verified level are a reportable practice problem, not a nuisance.
3. List the work required before return to service and describe exactly how the cutoff is proven afterwards.
Read the worked answer
Mechanical work: isolate and drain the water column and glass; remove the glass, gauge fittings, and the lower cross-connection; rod out both tappings back to the shell - do not rely on blowing through with steam, because a soft plug of sludge will just pack tighter. Inspect and replace the gauge glass valves and gaskets, and check that the glass is the correct length with the lowest visible edge above the lowest permissible water level. Open the low-water cutoff chamber, clean the float and the internal passages, confirm the float is not holed or rubbing, and check that the switch operates freely. While the boiler is open, inspect for scale and sludge in the mud drum and check whether the bottom blowdown valve is passing, since a passing valve both wastes water and can confuse level control. The sludge itself points upstream: review softener performance, hardness tests, and the blowdown schedule.
Proof test after reassembly: with the burner at low fire, slowly open the column drain (evaporative test) and watch the glass fall. The burner must shut off before the water level drops out of the bottom of the visible glass, and it must not relight until level is restored and the control is manually reset. Repeat to confirm repeatability, and test any second low-water cutoff separately. Record the test, the result, and the witness.