Material selection
Best stainless steel for seawater: 316L, 2205 or 2507
For submerged seawater choose 2507: it is the only one of these grades that reaches a PREN (pitting-resistance score, defined below) above 40, the level Rolled Alloys and Wikipedia describe as typical for seawater, if it is ordered to a PREN minimum. 316L is the floor for sheltered marine air only. 2205 is the strength and chloride-cracking upgrade over 316 and borderline for immersion.
- Submerged or flowing: 2507, after checking seawater temperature, chlorination and cathodic protection. Splash zone: 2507 is the best guess, but no grade has proof, so test a coupon.
- Sheltered marine air only: 316L.
- Unsure: 2507, then test a coupon in your own water. Its downside: we hold no data on sourcing, welding or machining it.
We hold no price, lead-time, welding or machining data for 2507, so ask your fabricator and quote both grades. Pick in three questions · What to send your supplier
- PREN
- pitting resistance equivalent number (also written PRE): %Cr + 3.3 × %Mo + 16 × %N. Higher resists pitting better.
- SCC
- stress corrosion cracking: cracks from tensile stress, chloride and heat together.
- ASTM G48
- a laboratory test in ferric chloride that finds the temperature where pitting or crevice attack starts.
- Crevice
- a tight gap (flange, gasket, deposit, fouling) where seawater stagnates and attacks the metal.
- Super duplex
- duplex steel with more chromium, molybdenum and nitrogen, such as 2507; the PREN is near 43.
- UNS number
- the alloy's catalogue number, such as S32750; the same alloy has other names in EN, DIN and JIS.
- Mill certificate
- the maker's test report for your batch, with its actual composition.
- Cathodic protection
- an applied or sacrificial current that protects submerged steel from corrosion.
- ppm, mg/L
- for chloride in water these are the same unit.
Real heats sit inside the bars, so ask for the mill certificate. Our record holds only the nominal composition of 2507 (25 Cr, 4 Mo, 0.3 N), so it is a point, not a range: the standard's composition limits are wider than nominal, so only an order that specifies a PREN minimum (suppliers print 41 to 42.5) guarantees 40. PREN is a pitting proxy and a ranking, not a specification, and it is not a fair way to compare austenitic with duplex grades.
Data behind this chart
| Grade | PREN low | PREN high | Cr % | Mo % | N % | Basis |
|---|---|---|---|---|---|---|
| 304 | 18 | 20 | 18 to 20 | 0 | no limit in our record | limits |
| 316 | 22.6 | 29.5 | 16 to 18 | 2 to 3 | 0 to 0.1 | limits |
| 316L | 22.6 | 29.5 | 16 to 18 | 2 to 3 | 0 to 0.1 | limits |
| 2304 | 22.5 | 29.7 | 21.5 to 24.5 | 0.05 to 0.6 | 0.05 to 0.2 | limits |
| 2205 | 34.1 | 37.8 | 22 to 23 | 3 to 3.5 | 0.14 to 0.2 | limits |
| 2507 | 43 | 43 | 25 | 4 | 0.3 | nominal |
- ✓316Lfits
- ✓2205fits
- ✓2507fits
- ?316Lunproven
- ?2205unproven
- ?2507unproven: highest PREN
- ✕316Ldoes not fit
- ~2205borderline
- ✓2507fits
- ?316Lunproven
- ?2205unproven
- ~2507borderline: one test only
✓ fits ~ borderline ✕ does not fit ? unproven in our data
- 1Is any part submerged, splashed, flowing, fouled or creviced?Yes2507The strongest candidate: highest PREN of the six and the only seawater crevice result. Splash-zone and crevice performance are still not verified, so ask the supplier for references and qualify it in your water.
- 2Is it a hot chloride process fluid above 60 °C, not immersed?Yes2205 or 2507Not 316: Alleima limits 304 and 316 to 60 °C in dilute oxygen-bearing chloride. Heated seawater is also a pitting and crevice problem, so check the pitting data above too.
- 3Is it only sheltered marine air, no immersion, no splash, no crevices?Yes316LWikipedia calls 316L the typical minimum grade near the seashore. The rest is our reading.
- Otherwise2205 or 2507If strength alone sets the wall thickness: 2205, 450 MPa minimum yield against 205 for 316. If you are still unsure: 2507, qualified in your own water.
316 vs 2205 vs 2507 for seawater: which grade fits?
2205 is borderline for immersion (PREN 34.1 to 37.8); 2507 clears 40; 316L fits sheltered air only. Row by row, the number behind each verdict is in the cell, and the charts below show where it comes from.
| Exposure | 316L | 2205 | 2507 |
|---|---|---|---|
| Sheltered marine air, rain-washed, no crevices | Fits. Wikipedia names it the typical minimum grade near the seashore. | Fits. More than the exposure needs unless you want the strength. | Fits. More than the exposure needs. |
| Splash zone: wet and dry cycles, salt build-up | Unproven. No splash-zone result in our sources; do not extrapolate from clean air. | Unproven. No splash-zone result in our sources. | Unproven. No splash-zone result in our sources, but the highest PREN of the six. No grade has proof here: test a coupon. |
| Submerged or flowing seawater, no crevices | Does not fit. PREN 22.6 to 29.5, where Wikipedia says above 40 is typically specified for submerged seawater and warns against 316L for seawater condenser tubes. | Borderline, for the conditions in our data. PREN 34.1 to 37.8, under the 40 that Rolled Alloys calls typical. Langley lists it for marine shafts, rudders, pumps and valves; whether that is enough depends on crevices, temperature and cathodic protection, which our data does not cover. | Fits. Printed PRE minimum 41 to 42.5, if ordered to that minimum. Used for seawater cooling tubing. Check temperature, chlorination and cathodic protection first: our profile flags hydrogen-induced cracking at high stress under cathodic protection (the risk falls with fine austenite spacing, under 30 µm per Alleima). |
| Crevices, gaskets, fouling or stagnant water | Unproven. No seawater crevice result in our sources. | Unproven. No seawater crevice result in our sources. | Only grade with any data. Crevice attack started at 64 °C in one supplier's accelerated seawater screening test; not a design temperature and not comparable to the ferric chloride numbers. |
| Hot chloride fluids (heated seawater, process lines) | Does not fit above 60 °C in dilute oxygen-bearing chloride. | Better than 316. Low susceptibility in supplier tests; no limit printed. Pitting still governs heated seawater (G48 critical pitting temperature 35 °C). Our source gives 280 °C maximum for pressure vessels (a temperature limit, not a cracking limit). | Conditional. No cracking at 10,000 ppm chloride up to 250 °C in tests. Pitting and crevice limits still apply, and our profile notes lower impact strength after long exposure above 250 °C. |
| Wall thickness set by strength | 170 MPa minimum yield (plate) | 450 MPa minimum yield (plate) | 550 MPa minimum yield (tube; plate not in our record) |
Two grades sit outside the matrix. 304 is not a seawater grade. 2304 has a PREN of 22.5 to 29.7, comparable to 316L, with better cracking resistance (about 125 °C in the supplier test): a chloride-cracking grade, not an immersion grade. Compare directly with the comparison pages below.
Grade facts at a glance
| Grade | PREN | ASTM G48 pitting temperature | Seawater crevice test | Chloride cracking | Minimum yield |
|---|---|---|---|---|---|
| 304 | 18 to 20 (nitrogen not limited) | no data | no data | about 60 °C limit | 205 MPa |
| 316 | 22.6 to 29.5; 24 printed | 10 °C | no data | 60 °C limit | 205 MPa |
| 316L | 22.6 to 29.5; 24 printed | 10 °C | no data | 60 °C limit | 170 MPa |
| 2304 | 22.5 to 29.7; 25 to 26 printed | no data | no data | tested to about 125 °C (dilute chloride) | 400 MPa |
| 2205 | 34.1 to 37.8; 34 to 35.8 printed | 35 °C | no data | low susceptibility, no limit printed | 450 MPa |
| 2507 | 43 nominal; 41 to 42.5 minimum printed | passes at 50 °C or more | 64 °C | none at 10,000 ppm to 250 °C | 550 MPa (tube) |
Different tests, forms and conditions sit in one row each; the charts below give them. "No data" means our sources print no number, not that the grade is untested.
How well do the six grades resist pitting in chloride?
Seawater attacks stainless steel by pitting and crevice corrosion. PREN ranks grades on composition alone. It is a pitting proxy: it does not predict crevice behaviour or a service life, and Rolled Alloys warns that it is unsuitable for specification. We computed it from the composition limits in each profile (first chart) and set the values that suppliers publish on top.
Three things stand out. The 316 family allows 22.6 to 29.5, so the envelope is wide and suppliers print 24 for typical 316L, far below the 45 that Rolled Alloys calls typical for austenitic grades in seawater. 2205 (UNS S32205) allows 34.1 to 37.8: even its best heat stays under 40. And suppliers disagree on 2205 by under two points (34, 35 and 35.8), which changes nothing: every value stays more than four points under 40. Each grade page lists its sources: 316, 2205, 2507, 2304.
Do pitting and crevice tests agree with the number?
They point the same way, though the methods differ. In ASTM G48, a ferric chloride test, the critical pitting temperature is 10 °C for 316 and 316L and 35 °C for 2205 (Rolled Alloys, as quoted in the 2205 profile); 2507 billet and forgings pass method A at 50 °C or higher. In a seawater crevice test on 2507 tube, corrosion started at 64 °C against 61 °C for 6Mo+N, a comparison alloy in the same test that is not in our database.
The seawater test raised the temperature 4 °C every 24 hours at a fixed potential of 150 mV against a saturated calomel electrode. It is one supplier's accelerated screening test, not a service limit, and real seawater can differ from the test (general engineering knowledge, not in our data). We hold no seawater crevice result for 316 or 2205.
Data behind this chart
| Row | Test | °C |
|---|---|---|
| 316 / 316L | ASTM G48 ferric chloride | 10 |
| 2205 | ASTM G48 ferric chloride | 35 |
| 2507 | ASTM G48 method A, pass level | at least 50 |
| 2507 | natural seawater crevice | 64 |
| 6Mo+N | natural seawater crevice | 61 |
That is why the matrix marks crevice service for 316L and 2205 as unproven instead of guessing. The only grade with a seawater crevice number is 2507, and that number comes from a single accelerated test on tube.
Which grades crack in hot chloride?
Chloride stress corrosion cracking needs tensile stress, chloride and heat. Alleima says 304 and 316 should not be used above 60 °C in dilute oxygen-bearing chloride, and Wikipedia gives about 60 °C for 304. Susceptibility depends on stress, chloride, oxygen and pH, so none of these numbers is a threshold you can design to. Cold seawater is outside this problem. Heated seawater, desalination and offshore process fluids are inside it, and there pitting and crevice limits apply as well.
The chart shows what each supplier tested, not a ranking at equal conditions. Alleima also says 2304 should not be used above 125 °C in oxygen-bearing solutions with more than about 10 ppm chloride. Duplex grades have their own limits: our source gives 280 °C maximum for 2205 in pressure-vessel use (a different limit from cracking), and our 2507 profile notes lower impact strength after long exposure above 250 °C, so 300 °C is a test condition, not a service rating.
Data behind this chart
| Grade | Condition | °C |
|---|---|---|
| 304 / 316 / 316L | limit in oxygen-bearing chloride | 60 |
| 2304 | dilute oxygen-bearing chloride, about 8 ppm oxygen, 1000 h | 125 |
| 2205 | low susceptibility (graph only) | not printed |
| 2507 | 10,000 ppm chloride | 250 |
| 2507 | 1,000 ppm chloride | 300 |
How much stronger are the duplex grades?
Much stronger. The minimum yield strength of 2205 is 2.2 times that of 316, plate to plate. 2507 is higher still at 550 MPa, but that figure is for seamless tube; Alleima's SAF 2205 tube is 485 MPa. Design codes also limit tensile strength and temperature and set a corrosion allowance, so the real saving in wall thickness is smaller than the yield ratio suggests; we hold no code allowable-stress data. Values are datasheet minimums for plate, sheet and strip (ASTM A240 where the source cites it), except 2304 (ATI plate) and 2507 (seamless tube).
Different product forms: plate, sheet and strip for most, seamless tube for 2507 (Alleima's SAF 2205 tube is 485 MPa). 2507 forgings to ASTM A182 are 530 MPa.
Data behind this chart
| Grade | Minimum yield MPa | Form |
|---|---|---|
| 304 | 205 | plate, sheet and strip, ASTM A240 |
| 316 | 205 | plate, sheet and strip, ASTM A240 |
| 316L | 170 | plate, sheet and strip |
| 2304 | 400 | ATI plate minimum |
| 2205 | 450 | plate, sheet and strip, ASTM A240 |
| 2507 | 550 | seamless tube and pipe, wall up to 20 mm |
Data behind this chart
| Grade | PREN low | PREN high | Minimum yield MPa |
|---|---|---|---|
| 304 | 18 | 20 | 205 |
| 316 | 22.6 | 29.5 | 205 |
| 316L | 22.6 | 29.5 | 170 |
| 2304 | 22.5 | 29.7 | 400 |
| 2205 | 34.1 | 37.8 | 450 |
| 2507 | 43 | 43 | 550 |
What drives the cost? Alloy content only, no price data here
We hold no price data and will not quote prices. Alloy content is not price, but buyers ask about it. At midpoint or nominal compositions 2205 and 2507 carry roughly 45 to 60% of the nickel of 316 (2304 about 35%) and as much or more molybdenum (3 to 4% against 2 to 3%); 2507 gets its pitting resistance from 25% chromium, 4% molybdenum and 0.3% nitrogen. Full compositions are on the duplex and austenitic pages.
Data behind this chart
| Grade | Ni % low | Ni % high | Mo % low | Mo % high |
|---|---|---|---|---|
| 304 | 8 | 11 | 0 | 0 |
| 316 | 10 | 14 | 2 | 3 |
| 316L | 10 | 14 | 2 | 3 |
| 2304 | 3 | 5.5 | 0.05 | 0.6 |
| 2205 | 4.5 | 6.5 | 3 | 3.5 |
| 2507 | 7 | 7 | 4 | 4 |
What should you send your supplier?
- Copy this line: UNS S32750 (or S32205, S31603); product standard and form named; mill certificate with Cr, Mo and N; PREN at least 41 for 2507; ASTM G48 report if crevices matter.
- Name the grade by UNS number: S31603 for 316L, S32205 for 2205, S32750 for 2507. The profile pages list the EN, DIN and JIS names as well.
- Name the product standard and form: for example ASTM A240 for plate, A276 for bar, A790 for 2205 pipe, A182 Grade F53 for 2507 forgings.
- Ask for the mill certificate with chromium, molybdenum and nitrogen, and compute the PREN yourself with the formula above. For 2507 ask for the printed minimum of 41.
- If crevices matter, ask for an ASTM G48 result at the temperature you need, and name the method.
- For a critical part, test a coupon in your own water before ordering the batch.
Where do the sources disagree?
| Question | What the sources say | How to read it |
|---|---|---|
| PREN of 2205 | 34 (Langley), above 35 (Alleima, on nominal composition), 35 (Rolled Alloys), 35.8 (ATI typical) | Different composition bases. None reaches 40. |
| 316 in seawater | Wikipedia calls 316 the preferred steel for marine environments but not suitable for seawater condenser tubes. AZoM scores sea water 5 of 5 for 316 with no conditions. | Prefer the statement with a stated use. A score without conditions cannot be compared with a test. |
| 316 vs 316L | Same Cr, Ni and Mo limits; carbon 0.08 vs 0.03%; minimum yield 205 vs 170 MPa. | Pitting resistance is the same on paper. The L grade is about welds, not seawater. |
What do we not know yet?
- Prices and lead times, and weldability and fabrication data for the duplex grades (ferrite balance, weld-metal PREN, pickling).
- Seawater crevice temperatures for 316 and 2205 (we only have them for 2507), and any splash-zone results.
- Long-term corrosion rates in mm per year in natural seawater.
- Cathodic protection, coatings and galvanic coupling, which change what a given grade can do. Our 2507 profile flags hydrogen-induced cracking under cathodic protection.
- Biofouling, flow velocity and chlorination, which change seawater behaviour a great deal.
- The alternatives engineers also use: 6Mo alloys, 904L, titanium and copper-nickel are not in the database yet.
Questions people ask
Is 316 stainless steel good for seawater?
316 is not suitable for submerged seawater; it is acceptable only in sheltered marine air, in our reading. Wikipedia calls 316L the typical minimum grade near the seashore and says 316 and 316L are not suitable for condenser tubes in seawater. Its PREN is 22.6 to 29.5; Rolled Alloys calls a minimum PREN of 45 typical for austenitic grades in seawater (40 for duplex), and warns that PREN is a ranking, not a specification.
Is 2205 duplex good for seawater?
2205 (UNS S32205) is borderline for immersion. It beats 316 on strength (450 MPa minimum yield against 205 MPa) and on PREN (34.1 to 37.8 against 22.6 to 29.5), but stays below the minimum of 40 that Rolled Alloys calls typical for duplex in seawater. Langley lists it for marine propellers, shafts, rudders, pumps and valves; for submerged or crevice service the documented choice in our data is 2507.
What PREN is needed for seawater?
Rolled Alloys, a supplier, calls a minimum PREN of 40 typical for duplex stainless steels in seawater (45 for austenitic, 35 for ferritic) and warns that PREN is a ranking, not a specification. Wikipedia says a PREN above 40 is typically specified. Among the six grades here only 2507, a super duplex, clears 40: suppliers print a minimum PRE of 41 to 42.5.
What is the difference between duplex and super duplex in marine service?
Super duplex (2507) carries 25% chromium, 4% molybdenum and 0.3% nitrogen, which gives a PREN near 43 against 34.1 to 37.8 for duplex 2205. In our data 2507 is the only one of the two with a seawater crevice test result: 64 °C in a supplier's accelerated test, which is not a service limit.
What is the difference between 316 and 316L in seawater?
Almost none for pitting: both have 2 to 3% molybdenum and the PREN formula has no carbon term. The L grade limits carbon to 0.03% so welds are less likely to sensitise. Minimum yield differs: 205 and 170 MPa.
At what temperature does chloride stress corrosion cracking start?
Alleima says 304 and 316 should not be used above 60 °C in dilute, oxygen-bearing chloride solutions. In Alleima's tests 2304 showed no cracking risk up to about 125 °C (about 8 ppm oxygen, 1,000 hours), and 2507 showed no cracking at 1,000 ppm chloride up to 300 °C or 10,000 ppm up to 250 °C. Those are test conditions, not service ratings.
How this was made
Each grade is a profile in the Space Mushroom database. The profiles are drafted by AI assistants from the cited pages, and a script checks that every value appears in the source it cites; each profile page lists its sources. Every number in this article is either computed from those profiles or checked against them when the page is built, so it cannot drift from the data. A named reviewer has not yet signed this article. PREN is computed from composition limits with the formula above; where a source prints a PREN we show that too. We name the product form with each strength because annealed plate, sheet and tube are not the same table. Related reading: strength of metals compared, all stainless steels, standards.
Last checked 24 October 2026. Next review when a profile changes.