Drum Level Swell After a Step Load Increase
A 480 t/h natural-circulation watertube boiler at 8.6 MPa drum pressure is carrying 410 t/h at steady state with three-element control (steam flow, feedwater flow, drum level) in automatic. The dispatcher requests a 40 t/h step increase in steam flow within two minutes. Four minutes after the ramp begins, the operator sees drum level rise 60 mm above normal while indicated steam flow has climbed 30 t/h and feedwater flow has risen only 12 t/h. Drum pressure has dropped 0.35 MPa. The feedwater pump has adequate margin, and the feedwater control valve is 70% open. A high-high level trip is set at 150 mm above normal. The operator must decide whether to intervene on the feedwater valve, reduce the load ramp, or hold.
1. Explain why drum level is rising even though feedwater flow is lagging steam flow, and identify what the level indication is actually responding to.
Read the worked answer
This is classic swell, not a true inventory increase. When steam demand steps up, drum pressure falls (0.35 MPa here), so saturated water in the circuit flashes to steam. The steam bubbles occupy more volume in the risers and drum, pushing the water level up even though the mass of water in the drum is not increasing - it may be falling.
The three-element controller sees steam flow ahead of feedwater flow and will be driving the valve open, but the level signal is contaminated by the swell. The 60 mm rise with only 12 t/h of feedwater increase against 30 t/h of extra steam confirms the level is a density/void-fraction artifact, not inventory.
The judgement: do not chase the level by throttling feedwater. That would starve the circuit and set up a low-level excursion once the swell collapses. The correct read is that the controller is behaving correctly - it is matching steam flow - and the level will recede as pressure stabilizes. The real risk is the opposite swing later.
2. Given the 150 mm high-high trip and current 60 mm excursion, recommend an operating action and state what would change your decision.
Read the worked answer
Hold the ramp rather than continue it. The 40 t/h step in two minutes is aggressive for a natural-circulation unit at 8.6 MPa; the pressure drop and swell are the symptoms. Pausing the ramp lets pressure recover and the void fraction settle, so level recedes without operator valve intervention. Keep the three-element controller in automatic - manual valve manipulation during swell is how operators cause the low-level trip they were trying to avoid.
What would change the decision: if level continued climbing toward roughly 110–120 mm despite holding load, that indicates the swell is not self-limiting, possibly a control valve stuck open or a false steam-flow signal. Then I would verify feedwater valve position against demand and consider transferring to single-element control or manual with a bias, while watching the drum-level trip margin. If pressure instead collapsed further and level fell sharply after the swell, that is the real inventory deficit and feedwater must be added immediately. Also verify the level indication itself - a fouled reference leg can read high.