TSA vs PSA: Two Different Regeneration Philosophies
Temperature Swing Adsorption (TSA) heats the bed to 200-280°C to drive off adsorbed water. Higher temperature = more complete regeneration = higher working capacity, but uses more energy.
Pressure Swing Adsorption (PSA) drops the bed pressure from 7-10 bar to 0.5-1 bar. Faster cycles, lower energy per cycle, but lower working capacity per pass.
- TSA typical: 220-260°C, 4-8 hour cycle, 10-14 wt% working capacity
- PSA typical: 0.5-1 bar desorption, 5-15 min cycle, 2-4 wt% working capacity
- Hybrid TSA+PSA: combines high capacity of TSA with speed of PSA, used in LNG and large air separation
The Temperature Window That Matters
Below 180°C: regeneration is incomplete, working capacity drops 30-40% within weeks, moisture breakthrough happens prematurely.
200-260°C: optimal range for most industrial applications. 220°C is the sweet spot for natural gas dehydration.
Above 300°C: hydrothermal damage to the zeolite crystal structure. Each 50°C over 280 shortens sieve life by ~30%.
Regeneration Gas Purity: The Hidden Killer
Regeneration gas must be dry (< 100 ppmv water) and oil-free. Using wet or oil-contaminated regen gas poisons the very bed you're trying to regenerate.
Common contamination sources:
1) Compressed air without desiccant pre-drying, 2) Off-gas from a process stream containing hydrocarbons, 3) Fuel gas that contains H2S or mercaptans.
Solution: always pre-dry regen gas with a smaller molecular sieve guard bed (10-15% of main bed volume) to protect your main bed.
Cycle Time vs Capacity Tradeoff
Longer cycle = higher working capacity (sieve adsorbs more before breakthrough) but lower availability. Shorter cycle = lower working capacity but higher uptime.
Standard practice: design for 50-70% of theoretical saturation at breakthrough, then regenerate. This gives you 2-3x margin before any moisture slip.
How to Tell When Your Sieve Is Dying
Three warning signs that regeneration is no longer saving the bed:
1) Regeneration temperature needs to be raised 20-30°C to maintain dew point, 2) Cycle time is being shortened to keep breakthrough in spec, 3) Pressure drop across the bed is increasing (sieve dust plugging flow paths).
If you see all three, the sieve is at end of life. Plan replacement, don't wait for breakthrough.