Cooling-water corrosion coupons and a materials decision
An open recirculating cooling tower at a chemical plant circulates 850 m3/h. Make-up water: pH 7.9, chloride 320 mg/L as Cl-, total hardness 640 mg/L as CaCO3, alkalinity 180 mg/L as CaCO3. Treatment is a zinc-phosphonate blend dosed at 40 mg/L into the circulating water.
Two mild-steel coupons, each of exposed area 25.0 cm2 and density 7.86 g/cm3, hung in the return header for 30 days. Coupon A was exposed while the dosing pump was out of service: initial mass 42.3160 g, final mass 41.7040 g. Coupon B was exposed with the pump working: initial mass 41.9820 g, final mass 41.8840 g. Coupon B showed no pitting; coupon A showed scattered pits up to 0.15 mm deep.
The plant standard for mild steel is a general corrosion rate not exceeding 0.10 mm/year. The maintenance engineer must decide whether to restart the same programme or change it, and must also answer the design team's question about using SS 304 tubes in the surface condenser on this water.
1. Calculate the corrosion rate of both coupons in mm/year and the inhibitor efficiency, then judge each against the plant standard.
Read the worked answer
Uniform rate from mass loss: thickness lost = W/(rho x A), then annualise by 365/30.
Coupon A: 0.6120/(7.86 x 25.0) = 0.6120/196.5 = 3.115e-3 cm = 0.03115 mm in 30 days. Annual rate = 0.03115 x 365/30 = 0.379 mm/year.
Coupon B: 0.0980/196.5 = 4.99e-4 cm = 0.00499 mm in 30 days. Annual rate = 0.00499 x 12.17 = 0.0607 mm/year.
Inhibitor efficiency = (0.379 - 0.0607)/0.379 x 100 = 84%.
Judgement: untreated water runs at 3.8 times the 0.10 mm/year limit, so it is not acceptable. The treated result at 0.061 mm/year is inside the limit but with only a 40% margin. The coupons also differ in mode: A pitted and B did not, and a 0.15 mm pit in 30 days extrapolates to roughly 1.8 mm/year of local penetration, which is far more damaging than the uniform figure suggests. A rate quoted from mass loss alone would understate the risk in the untreated case.
2. The plant manager wants to restart the same programme at 40 mg/L and consider the matter closed. Do you agree? Justify your answer and list the changes you would insist on.
Read the worked answer
Partly. The chemistry is adequate, but the incident was a reliability failure, not a chemistry failure, so restarting an unreliable dosing arrangement simply repeats it.
Evidence for keeping the programme: at pH 7.9 with hardness 640 and alkalinity 180 mg/L as CaCO3 the water is mildly scaling, which is protective for mild steel, and 320 mg/L chloride does not restrict mild steel. The dosed coupon achieved 0.061 mm/year, inside the 0.10 mm/year standard, and showed no pitting.
Evidence for changing the arrangements:
- Fit duty and standby dosing pumps with automatic changeover, a low-level alarm on the drum and flow-paced dosing, so a pump failure cannot go unnoticed for 30 days.
- Verify zinc and phosphonate residuals weekly, not monthly, and review after any dosing interruption.
- Add a supervisory corrosion coupon every month for the next quarter.
- On restart, run a short higher-dose passivation period to re-form the film on the already pitted surfaces.
- Check the Langelier Saturation Index and hold it slightly positive.
So the answer is: same chemical programme, different hardware and monitoring discipline.
3. The design team proposes SS 304 tubes for the surface condenser using this same cooling water. Give your recommendation with reasoning.
Read the worked answer
SS 304 is not acceptable here. Its practical chloride limit in cooling service is around 200 mg/L, and this water carries 320 mg/L, before allowing for the 3 to 5 times concentration that occurs across an open tower through evaporation. SS 304 under these conditions is prone to chloride pitting and to chloride stress-corrosion cracking, especially under deposits, in crevices at tube sheets, and in low-flow or stagnant tubes during shutdown.
The condenser also has a heat-transfer wall, so any through-wall pit means cooling water entering the process - a process-safety event, not just a maintenance cost.
Better options: SS 316L, whose molybdenum content raises the chloride tolerance considerably and which is common in fresh-water condensers, provided flow is maintained and the water is treated; or titanium, or a duplex or superferritic grade, if the tube side is this cooling water and long life matters. If 304 must be used, restrict it to a closed, deionised circuit.
This is a materials-selection judgement where the cheapest quote is the wrong answer: the cost of one tube failure exceeds the saving on the tube bundle.