Most 304-versus-316L arguments end in the same place: someone says "316L is better" and the quotation goes up. That is not an answer, it is a default. The real question is what touches the wetted surface, at what chloride level, and at what temperature. We are KAVORTE, an independent sourcing and engineering office in Zibo, Shandong. We audit and ship glass-lined, stainless steel and PTFE-lined reactors from 50 to 30,000 L, and we do not own a factory. Here is how we decide between the two grades on a live enquiry.
1. What the two grades actually are
304 is the 18/8 family: roughly 18 % chromium, 8 % nickel, no deliberate molybdenum. 316L adds 2–3 % molybdenum and caps carbon at 0.03 %.
Molybdenum is the whole argument. It stabilises the passive chromium oxide film against chloride attack. The standard way to express that is the pitting resistance equivalent number: PREN = %Cr + 3.3 × %Mo + 16 × %N. 304 lands near 18, 316L near 24. That gap is what you are buying.
The L is not decoration. Carbon at 0.03 % maximum prevents chromium carbide precipitation in the heat-affected zone during welding. Weld 316 without the L and the grain boundaries next to the weld are chromium-depleted — exactly where a chloride duty pits first. For the same reason, on welded pressure equipment ask for 304L rather than plain 304. The low-carbon rule is not a 316-specific feature.
2. The answer by medium
Treat this as a screen, not a datasheet. Confirm it against the final drawing and, where water is involved, against a water analysis.
- 304 or 304L is normally correct for: neutral and alkaline aqueous solutions, caustic soda, nitric acid, organic acids and solvents that carry no chlorides, edible oils and food-contact duties, ethanol, and process water below roughly 50 ppm chloride at near-ambient metal temperature.
- 316L is normally correct for: chlorides from about 50 ppm upward, brines, dilute sulphuric and phosphoric acid, acetic and fatty acids at temperature, and CIP regimes where chlorinated water is possible. It is also the routine choice where a GMP finish and a sterile-grade mechanical seal are specified together.
- Neither grade is correct for: hot concentrated chlorides, hydrochloric acid in any concentration, hydrofluoric acid or fluoride-bearing media, wet chlorine. Those go to a lined vessel or an exotic alloy.
3. The chloride screen, and where it breaks
For a stainless reactor the practical screen is chloride concentration against wall temperature. Below about 50 ppm at ambient, 304 gives long service. Between roughly 50 and 200 ppm, 316L is the normal answer. Above that, or once the metal temperature passes about 60 °C, neither grade is a safe default.
Note that 60 °C is the metal temperature, not the bulk liquid temperature. A jacket at 120 °C drives the inner wall well above the contents. This is the most common error we see on data sheets: a bulk temperature of 40 °C written beside a 140 °C jacket, 316L specified, no further thought given.
The failure mode is chloride stress corrosion cracking. It is cracking, not general thinning, so no corrosion allowance covers it. A vessel can pass its hydro test and crack in service three months later, at a weld, with no visible loss of wall.
4. Where 304 is not the cheap option, it is the correct one
The honest trade-off is that 304 is not a downgrade for mild duties. For several media it is metallurgically the better choice. Caustic soda is the clearest case: molybdenum reduces resistance to caustic over part of the concentration and temperature range, and 304L is the conventional selection for NaOH storage and handling. Concentrated nitric acid is the second: 304 is the workhorse and 316L buys nothing. Food, edible oil and clean low-chloride aqueous service sit in the same group.
Paying for molybdenum you do not need does not buy safety. It buys a false sense of it, while the items that actually decide vessel life — weld quality, passivation, dead corners, seal selection — go unfunded.
5. Grade is half the answer. Finish and welds are the other half
A 316L vessel with unpassivated welds and a rough ground weld cap will pit before a properly built 304 vessel on the same duty. The passive film does the resisting; anything that thins or breaks it starts a pit.
On stainless builds we ask for pickled and passivated wetted welds, ground-flush internal weld caps, full penetration on shell seams, and a stated internal roughness — typically 0.4–0.8 µm Ra for hygienic service, GMP mirror polish where the duty demands it.
Wetted and non-wetted surfaces matter too. On a normal stainless reactor the shell, head, agitator shaft, baffles, dip pipe and nozzle internals are wetted. The jacket, support ring, flanges and drive are not. A 316L shell on a carbon steel or 304 jacket is common and correct. Write which surfaces are 316L into the specification instead of writing "vessel: 316L".
6. Where 316L stops being enough
There is a ceiling, and it is lower than most buyers assume. Hot brine, chlorinated CIP water above 60 °C, hydrochloric acid, fluoride-bearing media — none of these are solved by moving from 304 to 316L. They are solved by changing material family. Our range covers glass-lined for inorganic and organic acids, organic solvents and weak bases within the lining's stated limits; PTFE lining options (PFA, PTFE, ETFE, ECTFE, FEP) for strong alkali, fluoride-bearing and hot chloride media, with PTFE or rubber-encapsulated PTFE gaskets and fluoropolymer-lined bolts; and duplex stainless, titanium or Hastelloy where a lining is not practical.
Specifying 316L for a hot chloride duty is not the conservative choice. It is the wrong material, and the batch will tell you so.
7. What to check when the vessel lands
- Mill certificate to EN 10204 3.1, with the heat number traceable to the plate that was actually formed — not simply to a plate from the same mill.
- Positive material identification on the finished vessel: shell, head, agitator shaft and one nozzle. A magnet is not a grade test. Cold-formed 304 is slightly magnetic and annealed 316L is not, so magnetism proves nothing either way.
- Weld documentation and the passivation record for the wetted surfaces.
- Ferroxyl test on welds where the specification calls for it.
8. What to do next
Send us the medium, its concentration, its chloride level, and your jacket temperature. If chlorides are below roughly 50 ppm and the metal stays near ambient, we will quote 304L and tell you the saving is real rather than a risk. Between 50 and 200 ppm, or with any elevated wall temperature, it is 316L and we will ask you to confirm the L and the wetted-surface list. Above that, we will quote a lined vessel, because a 316L reactor would fail and the quotation would then have been the cheapest part of the project.
Then hold the finish, the weld documentation and the certification to the same standard as the grade. In our experience those three items cause more premature failures than the choice between 304 and 316L ever does.