Demineraliser Train Fails After 18 Hours
The demineralisation plant of a 60 kg/cm² drum boiler has a single two-bed train - strong-acid cation followed by strong-base anion - rated at 12 m³/h with a design run length of 40 hours. Within 18 hours of a fresh regeneration these readings appear: cation outlet sodium 45 µg/L (normal below 5), anion outlet silica 0.07 mg/L (limit 0.02), combined outlet conductivity 0.55 µS/cm (limit 0.20), and treated-water pH 9.4 (normally near 7.0). Raw water turbidity has been 15-20 NTU for three days against a normal 2-3 NTU, and the clarifier alum dose has not been changed. Regeneration records show the usual 4% HCl and 4% NaOH doses applied, but only a 10-minute backwash at 8 m³/h, which the operator concedes gives little bed expansion. The boiler is running and the feed tank holds about 6 hours of acceptable water. A standby train exists but is unregenerated.
1. Identify which bed or beds have failed and explain how the chemistry points to the cause.
Read the worked answer
The pH rise to 9.4 is the diagnostic clue. When the cation bed is working, it exchanges calcium and sodium for hydrogen ions, so the water leaving it is acidic and contains carbon dioxide. If the cation bed is exhausted, sodium slips through as sodium salts; after the anion bed these become sodium hydroxide, so the combined effluent turns alkaline and its conductivity rises even though the total dissolved solids are not dramatically high. The 45 µg/L sodium and 0.55 µS/cm conductivity fit that pattern.
Silica leakage at 0.07 mg/L shows the anion bed is also spent: silica is weakly ionised and is only well removed by a strong-base anion resin in the hydroxide form, so short run length and early silica breakthrough go together.
Both beds failing at 18 hours against a 40-hour design, immediately after a turbidity excursion, points to fouling and overload rather than a bad chemical dose: colloidal clay and organics from the raw water coat the resin beads, block exchange sites and are not removed by a 10-minute, low-rate backwash.
2. State your immediate operating decision for the next 6 hours and the regeneration changes you would make.
Read the worked answer
Immediate decision: take the demineralised water outlet off the boiler feed line and recirculate it to the raw water pit or neutralisation sump. Do not blend with good water to dilute the conductivity into range - blending hides the excursion and still delivers sodium and silica to the boiler. Use the 6 hours of stored feed, cut boiler load if necessary, and tell the boiler operator in writing why.
Regeneration changes for the standby train:
- Backwash at a rate that gives at least 40-50% bed expansion, calculated from the vessel diameter, for long enough to clear the turbidity.
- Double the backwash volume and check the effluent for clarity before regenerating.
- Regenerate with the same acid and caustic strength but verify the concentration by titration, not by the drum label.
- Consider an alkaline brine or a surfactant-assisted clean if organic fouling is suspected.
- Check the resin for iron and organic fouling, and inspect the underdrain for channelling.
At the clarifier, raise the coagulant dose and add a polymer so the next water reaches the plant below 5 NTU.
3. A supervisor argues that 0.55 µS/cm with 0.07 mg/L silica is harmless because the boiler has been running for years and the values are still low in absolute terms. How do you respond?
Read the worked answer
Absolute concentration is the wrong yardstick; the relevant fact is that the boiler evaporates this water many times over, so non-volatile impurities concentrate in the drum while silica becomes volatile in high-pressure steam.
Silica carryover deposits on turbine blades as a glassy layer that is very difficult to remove, cuts efficiency and can force rotor rebalancing. Sodium carried into the drum forms sodium hydroxide, which concentrates under deposits and in crevices and causes caustic gouging and stress-corrosion cracking of the drum and tubes. The pH of 9.4 in the treated water also disturbs the feedwater chemistry control that the boiler relies on. Both effects are cumulative: a few hours of low-level excursion does not show up immediately, and that is exactly why it gets tolerated until a tube fails.
So the correct answer is that these values are outside the specified limits, that the limits exist because of the boiler and turbine, and that the water stops going forward until the plant is back within specification.