Materials and Surface Finishes

316L contains about two to three percent molybdenum, which raises it's resistance to chloride pitting and crevice attack; 304L has none and costs less. Both are limited by the same chloride stress-corrosion cracking mechanism at elevated temperature, so moving from 304L to 316L buys pitting resistance, not immunity from cracking. The typical mistake is specifying 316L for a chlorinated cooling-tower loop and expecting it to solve a cracking problem; it delays the failure, and the bundle still cracks at the U-bends where residual stress is highest. For clean steam and softened water, 304L tube is a sound and cheaper choice.

When chloride level, temperature and tensile stress together put a 300-series tube in the cracking zone, or when the chemistry is acidic, reducing or sour enough to attack the passive film. Each step up answers a named problem: duplex 2205 for chloride cracking at a cost not far above 316L, Alloy 20 for sulfuric acid, Hastelloy C-276 for mixed acids with chlorides, titanium for seawater. The decision point is the U-bend, because cold bending leaves residual stress exactly where chlorides concentrate. Send the chloride content and the hottest wall temperature, and the alloy choice is usually a short conversation.

Yes. Duplex and the nickel alloys work-harden faster than 304L or 316L, and titanium has less ductility and springs back more, so the bender needs a larger minimum centerline radius or the outside wall thins beyond what the code allows. For the same shell this costs tubes in the innermost rows, which is why an alloy upgrade on a replacement bundle can mean slightly less surface than the original. Order the alloy and the bend radius together so the rating absorbs the difference.

Often, yes: the shell sees only the shell-side fluid, so a carbon steel shell with a stated corrosion allowance and a 316L bundle puts the expensive alloy only where the process touches it. The pairing fails when the shell-side fluid is itself corrosive, or when the two metals are electrically coupled in a conductive liquid and the carbon steel becomes the anode, wasting away at the tubesheet gasket face. Check which fluid is on which side before accepting the saving. On steam shells it is the normal choice; on tower water the galvanic question needs answering first.

Ra is the average surface roughness in microinches, and a lower number leaves fewer grooves for soil and biofilm to anchor in, so it sets how well a clean-in-place cycle works. The mistake is specifying Ra alone: a mechanically polished tube can meet the number yet carry folded metal and embedded grit that an electropolished surface at the same Ra does not. State the Ra and the method together, and say whether the tube interior or the exterior is the product side, because polishing the inside of a U-tube is a different operation from polishing the outside of a bundle.