Condenser Vacuum Creep on a Light-Ends Column
A vacuum condenser on a light-ends column was designed to condense 25,000 lb/h of overhead vapor (latent heat 150 Btu/lb) at a condensing temperature of 200 °F using 300 gpm of cooling water entering at 85 °F and leaving at 110 °F. The exchanger has 1,500 ft² of surface. Over three months the column pressure has crept from 15 to 27 psig, and the overhead vapor now condenses at 212 °F. Cooling-water flow and inlet temperature are unchanged, and the overhead vapor rate is still about 25,000 lb/h, so the condenser duty has not changed. The reflux drum vent goes to the flare; its flow is not measured. Column pressure control is approaching its output limit.
1. Compute the condenser duty and confirm the stated cooling-water flow. Back out the design overall heat-transfer coefficient.
Read the worked answer
Duty
Q = 25,000 lb/h × 150 Btu/lb = 3.75 × 10⁶ Btu/h
Cooling water
W = Q / (cp ΔT) = 3.75 × 10⁶ / (1.0 × 25) = 150,000 lb/h
At 8.34 lb/gal and 60 min/h, 150,000/500 = 300 gpm, which matches the meter.
Design LMTD (condensing on one side)
ΔT₁ = 200 − 85 = 115 °F; ΔT₂ = 200 − 110 = 90 °F
LMTD = (115 − 90) / ln(115/90) = 25 / 0.2451 = 102.0 °F
UA and U
UA = 3.75 × 10⁶ / 102.0 = 36,760 Btu/(h·°F)
U = 36,760 / 1,500 = 24.5 Btu/(h·ft²·°F)
That is a low but reasonable clean coefficient for a vacuum condenser with a noncondensable heel. Note the water ΔT of 25 °F gives a tube-side velocity worth checking - if it is under about 3 ft/s, fouling and scale are expected.
2. Estimate the current UA and the fouling resistance. Does the data support fouling as the cause?
Read the worked answer
New LMTD. Duty and water flow are unchanged (water still 85 → 110 °F), but the condensing temperature is now 212 °F:
ΔT₁ = 212 − 85 = 127 °F; ΔT₂ = 212 − 110 = 102 °F
LMTD = 25 / ln(127/102) = 25 / 0.2193 = 114.0 °F
New UA
UA = 3.75 × 10⁶ / 114.0 = 32,890 Btu/(h·°F)
UA fell by 10.5%, so U is now about 21.9.
Fouling resistance
1/UA_new − 1/UA_design = 3.041 × 10⁻⁵ − 2.720 × 10⁻⁵ = 3.21 × 10⁻⁶ h·°F/Btu
R_f = 3.21 × 10⁻⁶ × 1,500 = 4.8 × 10⁻³ h·ft²·°F/Btu
That is roughly 2–4 times the usual 0.001–0.002 design allowance for treated cooling water, so the fouling is heavy for a three-month interval but not physically impossible.
The key diagnostic is the temperature, not the pressure. If air were leaking into the condenser, the total pressure would rise but the vapor's partial pressure - and therefore the condensing temperature - would stay at 200 °F. Here the condensing temperature itself rose 12 °F, which means the condenser needs a larger LMTD to reject the same duty. That points to a genuine loss of UA (fouling or low water velocity), not to noncondensables.
3. What would you do next, and how would you confirm the diagnosis before committing capital?
Read the worked answer
Confirm before acting
- Trend the water-side ΔT. If duty is truly constant and flow is constant, ΔT must stay at 25 °F. A rising ΔT means the duty is not constant and the vapor rate assumption is wrong (e.g., unmeasured boilup carryover).
- Measure the vent flow and analyze it for inerts. Even though the temperature evidence argues against air, a 12 psi pressure rise in a vacuum system is worth confirming - a flooded vent line or a plugged vent orifice is cheap to find and cheap to fix.
- Check the shell-side and tube-side ΔP. A fouling layer shows up as a rising tube-side ΔP at constant flow; a plugged tube bundle does not.
- Pull a water sample and check the treatment program, the tube velocity, and the last cleaning record.
Actions
- Short term, reduce the pressure burden: verify the column pressure controller is not fighting a saturated output, and confirm the relief device set pressure is not being approached as the column pressure climbs.
- Clean the exchanger at the next planned outage and restore the water velocity above 3 ft/s; if the water is the root cause, the exchanger will foul again.
- If the column cannot tolerate another excursion, a parallel shell (split the surface into two 750 ft² units) lets you clean online. That is the usual answer for a condenser that is the pressure bottleneck.
- Do not simply raise the pressure setpoint - that trades a heat-transfer problem for a relief and product-quality problem.