Vent Condenser Cannot Meet a 150 mg/Nm³ Stack Limit at a Solvent Tank Farm
A tank farm stores methanol and toluene in three nitrogen-blanketed tanks. Two tanks hold methanol; the third holds toluene and is the one filled most often. During road-tanker filling, displaced vapour passes through a shell-and-tube condenser cooled by a glycol loop at −10 °C, then a single 6 m³ activated-carbon bed, then the stack. The new operating permit sets a stack limit of 150 mg/Nm³ total VOC, averaged over each unloading campaign; the site has been reporting against a 1,000 mg/Nm³ internal target for years. Last month the continuous FID read 6,500 mg/Nm³ at the condenser outlet and 380 mg/Nm³ after the carbon bed. The condenser vendor proposes a −35 °C refrigerant package and quotes "over 90 % removal". The carbon bed is steam-regenerated every 24 h; working capacity is 9 kg and campaign vents run 18 h at 60 Nm³/h. Operations suggest simply cutting the nitrogen blanket rate. The permit authority has asked for a written technical basis before any change.
1. Determine whether cooling the vent to −35 °C, or to any temperature reachable with a refrigerant, can meet 150 mg/Nm³ for this mixture. Show the calculation and state what role the condenser can realistically play.
Read the worked answer
Saturation sets the floor, and the floor is far above the limit.
For a condensable vapour in a non-condensable carrier, the outlet concentration is capped by the vapour pressure at the condensing temperature:
- Methanol at −10 °C: p* ≈ 2.2 kPa → y = 2.2/101.3 = 0.0217 → c = y·M/22.414 = 0.0217 × 32/22.414 ≈ 31,000 mg/Nm³.
- Methanol at −35 °C: p* ≈ 0.4 kPa → c ≈ 5,900 mg/Nm³.
- Methanol at −50 °C: p* ≈ 0.10 kPa → c ≈ 1,450 mg/Nm³.
- Toluene at −35 °C: p* ≈ 0.08 kPa → c ≈ 3,200 mg/Nm³.
Extrapolating, a condenser would need to run near −80 °C to reach 150 mg/Nm³ for methanol, which is not achievable with a glycol or standard refrigerant package and would freeze any water in the vent.
Implications
- The vendor's "90 % removal" is a percentage of a large number, not proof of compliance; 90 % of 31,000 mg/Nm³ still fails by two orders of magnitude.
- The condenser is a load-reduction device. The carbon bed is the control device, and it must be sized, monitored and permitted as such.
- The measured 6,500 mg/Nm³ is well below the 31,000 mg/Nm³ equilibrium value, so the gas leaving the condenser is not saturated. Before buying anything, check the demister, the sample point temperature and the probe location: a warm sample line or mist carryover will corrupt the number.
2. The carbon bed is currently regenerated on a fixed 24 h timer. Assess whether this is adequate at the new limit and propose the control philosophy you would write into the permit basis.
Read the worked answer
Mass balance on the bed
- Load = 6.5 g/Nm³ × 60 Nm³/h = 390 g/h.
- Per 18 h campaign = 7.0 kg.
- Working capacity 9 kg → margin 2.0 kg ≈ 5 h of operation.
A 5 h margin is thinner than the normal variability in filling rate, vent temperature and drum-connection leakage, so breakthrough before the timer fires is credible. The measured 380 mg/Nm³ is already 2.5× the permit limit, so any further loss of bed performance puts the site in breach.
What a fixed timer gets wrong
A timer assumes a constant load. The load varies with condenser temperature (saturation), filling rate and which tank is being filled (toluene vs methanol). It also ignores slow drift in bed capacity from moisture and heavy ends.
Recommended basis
- Lead-lag beds with automatic changeover, sized so that one bed alone carries a full campaign at the worst credible load.
- Continuous FID on the stack as the primary compliance monitor, calibrated against the actual components. FID response factors for oxygenated compounds differ from a methane calibration by tens of percent, so a methane span gas alone is not defensible.
- Regenerate on a mass-balance counter (integrated flow × concentration) with a fixed timer as a backstop, and trend the outlet as an early warning.
- Treat steam-regenerant condensate as a listed waste stream and confirm the effluent route and consent before the permit application.
- Confirm the permit's averaging basis: "per campaign" allows no credit for dilution or for a clean bed at the start of a campaign.
3. Operations propose halving the nitrogen blanket flow. Evaluate the technical effect on stack concentration, on tank safety and on the permit position.
Read the worked answer
Concentration: no benefit. The condenser outlet concentration is fixed by temperature (vapour pressure), not by carrier flow. Halving the blanket flow halves the mass emitted (390 → 195 g/h) and roughly doubles the time to bed breakthrough, but the instantaneous mg/Nm³ at the condenser outlet and at the stack changes only via the bed's loading history. It is a load-reduction measure, not a concentration measure.
Mass load: real but modest. 195 g/h extends the 9 kg capacity to about 18 h of margin. That buys time; it does not make the design compliant against a 150 mg/Nm³ limit.
Tank safety: this is the dangerous part. The blanket exists to keep the ullage outside the flammable envelope. Methanol (flash point ~11 °C) and toluene (flash point ~4 °C) both require inerting at ambient conditions. Halving the flow risks:
- air ingress during pump-out, when the tank draws in vapour or air;
- a lower margin against the limiting oxygen concentration;
- inadequate inbreathing capacity if the pressure/vacuum vent is already near its rating.
Any change must be demonstrated with an oxygen analyser on the ullage under worst-case pump-out, with a documented margin below the LOC, not a rule of thumb.
Permit position. The vent stream and blanket rate are part of the permitted design basis. Reducing the blanket flow is a management-of-change item and needs the authority's agreement in writing, with the supporting oxygen and emissions data. Present it as a supporting measure alongside the corrected carbon-bed control, never as the compliance solution.