Seawater cooler: is 2205 duplex the right upgrade?
A coastal refinery's seawater cooling circuit uses a fixed-tubesheet exchanger with 316L tubes. The tubes have perforated at the rolled tube-to-tubesheet joints after four years. Seawater is 19,500 mg/L chloride with 8 mg/L dissolved oxygen, sand-filtered upstream, dosed continuously with 0.3 mg/L sodium hypochlorite; inlet is 31 °C, outlet 42 °C, tube velocity 1.8 m/s. Inspection found crevice attack in the tube holes and behind the water-box partition plate. The turnaround plan is a like-for-like bundle in 2205 duplex, retaining the existing epoxy-coated carbon steel water boxes with aluminium sacrificial anodes. The budget for this cycle covers only one of the two identical exchangers in the circuit.
1. Using pitting/crevice resistance numbers, judge whether 2205 duplex is adequate for the tube-to-tubesheet joint at these temperatures and this chlorination regime.
Read the worked answer
Rank the alloys. PREN = Cr + 3.3Mo + 16N gives roughly 26 for 316L (17 Cr, 2.5 Mo, 0.05 N), about 35 for 2205 (22 Cr, 3.1 Mo, 0.17 N), about 43 for 2507 super duplex and about 43–44 for the 6% Mo austenitic grades.
Temperature matters more than the alloy label. In natural seawater the critical crevice temperature is roughly 15–25 °C for 316L and around 45–50 °C for 2205. That already explains the 316L failures at 31–42 °C. But this system is continuously chlorinated at 0.3 mg/L: hypochlorite raises the steel potential by roughly 200–400 mV, which typically pulls the critical crevice temperature down by 15–20 °C. The realistic 2205 threshold here is nearer 30–35 °C.
Judgement. The outlet is 42 °C and the crevice is the tube hole, the worst geometry in the exchanger. 2205 is marginal at best and may repeat the failure. I would not accept rolled 2205 joints on this duty without either (a) a higher grade - 2507 or 6% Mo - or (b) eliminating the crevice by seal-welding the tubes to the tubesheet, plus a tubesheet coating.
Relabelling the alloy without changing the joint design is the error being proposed.
2. Evaluate keeping the existing epoxy-coated carbon steel water boxes with aluminium anodes behind a duplex tube bundle. What new risks does the mix create?
Read the worked answer
The couple is the problem. Duplex is cathodic to carbon steel, so any coating holiday in the water box becomes a small anode facing a very large duplex cathode. That is the classic bad area ratio: rapid, deep pitting at holidays, faster than in the original all-316L design where the potential difference was smaller.
The anodes are now shared. The aluminium anodes must protect both the water box and, unavoidably, part of the duplex surface. Current demand rises, anode life falls, and the protection actually reaching the box depends on the coating's condition, which is unknown for an older lining.
Overprotection risk. Driving the box to very negative potentials to compensate causes hydrogen evolution, lining disbondment and blistering, and the disbonded area then shields itself from the anodes. Aluminium anode output also falls in low-flow or heavily chlorinated water.
What must be done
- Survey and repair the lining before assembly; do not assume it is intact.
- Verify anode mass and current output against the combined wetted area, and confirm the anode alloy suits chlorinated seawater.
- Fit reference cells or at least test coupons so potentials can be checked after start-up, and set a potential window rather than 'as negative as possible'.
3. With budget for only one exchanger, propose a decision and a plan for the second unit. Justify the trade-offs.
Read the worked answer
Do not buy the same bundle twice. Repeating the 2205 rolled-joint design in unit one would leave unit two with a known failure mode and no data. The one available budget should buy information as well as hardware.
Unit 1 (funded): install the upgraded bundle - 2507 or 6% Mo tubes, seal-welded to the tubesheet, with a coated tubesheet face and verified water-box lining. This is the design that should stop crevice attack at 42 °C under chlorination.
Unit 2 (deferred): do not simply retube in 316L. Retube with the existing grade where tubes are sound, plug the failed tubes within the exchanger's tube-plug limit (typically about 10%), coat the tubesheet, confirm anode capacity, and accept the derated duty. Record the duty loss so the process side sees the consequence.
Make the deferral evidence-based: shorten the inspection interval on unit 2, add tube-to-tubesheet joint inspection (eddy current plus visual on pulled tubes), and trend outlet temperature as a fouling/plugging indicator.
Also check the chemistry first. If the chlorination can be moderated or intermittent, both exchangers become easier to protect. Treatment change may be cheaper than alloy change and should be evaluated in the same decision.