Two questions, and the datasheet answers them with two different numbers

A corrosion figure answers one of two questions. A rate answers how fast the wall thins. A critical temperature answers whether the alloy stays passive in that environment at all. Both appear in this catalogue, they are quoted to prove opposite things, and neither transfers to a service it was not measured in. The 2205 page publishes all three kinds at once — an isocorrosion curve stated at 4 mpy, or 0.1 mm per year, a pitting table in sodium chloride, a critical crevice corrosion temperature in ferric chloride and a general corrosion table in wet process phosphoric acid. The 825 page prints a boiling acid table, a crevice temperature table and a U-bend cracking table side by side. Those three tables are worth an hour before a tonne is ordered.

Mils per year is a distance, not an opinion

One mil is a thousandth of an inch, so one mil per year is 0.0254 mm per year, which is the conversion the 2205 page hands you when it equates 4 mpy with 0.1 mm per year. Read the phosphoric acid table with that in hand. At solution A, 140 degrees Fahrenheit, 2205 is quoted at 3.1 while 316L is above 200 and 904L sits at 47; at solution B, 120 degrees Fahrenheit, the same three read 3.9, above 200 and 6.3. The column heading says inches per year, which cannot be right — nothing loses 200 inches a year — so the figures are mils per year, and that slip is reason enough never to copy a rate without checking its unit. Converted: 316L at above 200 mpy loses more than 5 mm a year, so a six millimetre wall is gone inside eighteen months. 904L at 47 mpy loses 1.19 mm a year, which is real metal on a twenty year life. 2205 at 3.1 mpy loses 0.079 mm a year.

Everything depends on the word boiling

The acid table is where buyers most often fool themselves. The 825 page quotes rates in boiling laboratory sulfuric acid at three concentrations, in mils per year with millimetres per year in brackets: at 10 per cent, 316 reads 636 or 16.2 mm per year against 825 at 20 or 0.5; at 40 per cent, 316 is above 1000 against 825 at 11; at 50 per cent, above 1000 again against 20. The shape matters — austenitic 316 gets worse as concentration rises, and 825 hardly moves. But it is a boiling solution, a laboratory reagent and a clean specimen. A tank at ambient has a different driving force, and quoting 0.5 mm per year from that table to justify a lining is quoting the wrong experiment. Where the duty leaves stainless altogether, the evidence has to change with it: Hastelloy B2 is listed for hydrochloric acid at all concentrations and temperatures, while Monel 400 concedes only some resistance to hydrochloric and sulfuric acids at modest conditions and is, by its own page, seldom the material of choice for them.

The crevice table ranks the same alloys differently

Now take the third table on that page. In ASTM G-48 at 10 per cent ferric chloride, the temperature at which crevice attack begins is 27 degrees Fahrenheit, or minus 2.5 Celsius, for 316; 32 degrees Fahrenheit, or 0 Celsius, for 825; and 113 degrees Fahrenheit, or 45 Celsius, for SSC-6MO and 625. The alloy that beat 316 by thirty times in boiling sulfuric acid is 2.5 degrees better than 316 here. Both are true; they describe different mechanisms, and only one matches what the plant contains — gasket faces, lap joints, deposited solids, every deliberately mechanical joint. In this catalogue the 6MO role falls to 254 SMO, described as used in high chloride process streams and sometimes a more cost effective substitute for high nickel and titanium alloys, and for the oxidising chloride salts generally to C-276, which lists wet chlorine gas, hypochlorite and chlorine dioxide among the few services it survives.

Resistant is not a claim until the test is attached

The U-bend table makes the same point about cracking. Tested as U-bend samples in boiling 42 per cent magnesium chloride, 316 fails, SSC-6MO and 825 are recorded as mixed, and 625 resists; the footnote says mixed means a portion failed inside two thousand hours of testing. Read honestly, mixed and resist are different verdicts, and a claim of cracking resistance without solution, temperature, specimen form and duration carries no information. The same discipline holds for pitting: 2205 quotes a critical pitting temperature in one molar sodium chloride measured in a named cell, a critical crevice temperature in ferric chloride, and the pitting resistance equivalent formula itself, which is how S32760 reaches PREN 40. None of those numbers says anything about general corrosion rate, and 321 or 347 exist to solve yet another mechanism again, intergranular attack rather than pitting.

What the specimen did not carry into your plant

Four things are missing from every laboratory figure, and each belongs on the specification rather than in the margin. Velocity and suspended solids turn general corrosion into erosion corrosion, which no immersion table predicts. Heat flux means the wall runs hotter and more concentrated than the bulk fluid, so a tube fails where steam is generated and not in the beaker. The weld is a different structure: the 904L page notes that a higher alloyed filler can be necessary because of the cast structure of the weld metal, that preheating is unnecessary, and that austenitics carry only 30 per cent of the thermal conductivity of non-alloyed steel, so heat input has to be lower. And every joint drawn is a crevice. Welding, cutting and surface treatment are where those decisions become physical.

Turning a rate into metal on your order

The conversion is arithmetic and should be done before the enquiry rather than after the quotation. Take the rate you believe for your condition, multiply by design life, add it to the wall the pressure calculation requires, and see where it lands. Inside an available size, buy thickness: plate covers 0.3 to 200 mm from Q235 through A36 to Q345. Outside it, change alloy rather than add metal — which is when 2205 becomes cheaper than a heavier austenitic build, when 317LMN earns its nitrogen in sheet, tube or plate and coil, or when duplex is taken as coil and pipe. Above all that sits titanium, quoted by its own page below 0.005 mm per year in seawater. Then fix six lines on the order: medium, concentration and temperature; heat transfer through the wall; crevices and where; solids and velocity; weld detail including filler; and the certificate, since S32760 supplies 3.1 as standard and 3.2 at surcharge. Only the first is a corrosion question, and it is the one most often left blank.