Depropanizer Floods After a Lighter Feed and a Debottleneck
A depropanizer with 44 valve trays, 2.0 m inside diameter and 600 mm tray spacing uses a total condenser at 15.5 barg top pressure. Design feed is 28 t/h at a reflux ratio of 1.6, with a design column pressure drop of 0.55 bar and top C3 specification of 1.0 mol%. After a debottleneck, feed is 33 t/h at the same reflux ratio. The column now shows 0.95 bar pressure drop, 16.4 barg top pressure, the condenser cooling-water valve 100% open, and the reboiler steam valve at 95%. Top C3 is 2.8 mol% and bottom C3 is 7.5 mol% against a 4.0 mol% specification. The top temperature is 52°C versus a design 49°C, the feed tray is 82°C, and the bottom is 108°C.
A pressure survey shows 0.10 bar across trays 1–8, 0.62 bar across trays 9–36, and 0.23 bar across trays 37–44. The upstream unit now sends a lighter feed: C2 has risen from 1.5 to 4.0 mol%. The panel operator proposes increasing reflux to recover top purity.
1. Using the pressure-drop and temperature data, decide whether the column is flooded, fouled, or damaged. Explain the significance of the distributed pressure drop and the feed change.
Read the worked answer
Flooding is the leading diagnosis. Total pressure drop has risen from 0.55 to 0.95 bar, or from 1.25 to 2.16 kPa per tray. The increase is distributed across the column rather than concentrated in one section, which points to a hydraulic load limit rather than fouling or tray damage. The top temperature has risen 3°C while the condenser valve is fully open and the column pressure is 0.9 bar above design; that is consistent with a flooded rectifying section and reduced reflux. The lighter feed (C2 1.5 to 4.0 mol%) raises vapor load for the same bottoms C3 specification and pushes the column closer to flood.
Tray damage would normally show a lower pressure drop and a sharp loss of separation. Fouling would show a localized pressure-drop increase and a normal hydraulic response. A definitive test is to reduce boilup by 10% for 30 minutes: in flooding, pressure drop should fall sharply and top temperature should drop; in fouling, pressure drop changes little. A gamma scan would confirm the froth height and the flooded section.
2. Evaluate the operator's proposal to increase reflux. Give two better immediate moves.
Read the worked answer
Increasing reflux adds vapor and liquid traffic and moves the column further into flooding; top purity may get worse because flooding reduces tray efficiency. Better immediate moves: first, cut feed to 28 t/h and reduce reboiler duty to restore the design vapor load. Second, check and vent noncondensables from the condenser; if inerts are raising pressure, venting lowers pressure and improves the condenser temperature difference without adding load. Also maximize condenser cooling water and check for fouling. If top specification is urgent, route off-spec C3 to flare or recycle rather than pushing the column.
Do not raise column pressure to suppress flooding unless the hydraulic rating shows it helps; higher pressure reduces volumetric vapor flow but also reduces relative volatility and can make separation worse.
3. For a permanent 33 t/h debottleneck, compare high-capacity trays, structured packing, and a larger column, and state the key constraint.
Read the worked answer
High-capacity valve or sieve trays typically gain 15–25% capacity, but downcomer area, support rings, and tray spacing must be checked; the existing 2.0 m shell may not accommodate the required downcomer. Structured packing gives higher capacity and lower pressure drop, but needs good liquid distribution, a larger sump, and has lower turndown; it may fit if the column is re-headed. A larger-diameter column has the highest CAPEX and longest downtime.
The key constraint is the hydraulic rating at 33 t/h. At 0.95 bar, the column is already near or beyond flood, so a 15% tray upgrade may still be marginal. A feed preflash or upstream C2 removal would reduce vapor load and may be the lowest-cost debottleneck. First verify the tray-by-tray hydraulic model with a gamma scan, then compare lifecycle cost of trays versus feed conditioning.