How Much Better Is 316L Than 304? PREN Differs by 5, Service Life by 3x

Scope of the calculations (what is in and what is not). Three quantities are worked from first principles here: the element-by-element decomposition of PREN, an exponential fit and extrapolation of the chloride threshold concentration, and a cross-source slope for the critical pitting temperature. Not included: stress corrosion cracking, microbiologically influenced corrosion, erosion corrosion, and any quantitative effect of cleaning and maintenance regime on service life. Every service-life multiple quoted is a field observation, not a design value; the only design criteria used are the chloride threshold and the CPT. Where sources disagree, intervals are given rather than single numbers.

Every material meeting produces the same sentence: 316L resists corrosion better than 304. It is true, and it is useless on a drawing, because it does not say by how much.

The pitting resistance equivalent number turns it into arithmetic. PREN = Cr + 3.3*Mo + 16*N gives 19.49 for 304 and 24.93 for 316L, a gap of 5.44 points. What actually decides whether a fabrication lasts is not those 5.44 points but how much environmental capacity they buy. This article converts them into two numbers that can be written straight into a specification: a chloride threshold concentration and a critical pitting temperature.

1. Getting PREN right: what each element is worth

The standard expression:

Cr, Mo and N are mass percentages. Chromium is the film-forming base and is given the coefficient 1. Molybdenum specifically suppresses chloride-induced pitting and carries 3.3. Nitrogen stabilises the passive film and carries 16.

One point that is routinely stated backwards: the largest coefficient belongs to nitrogen at 16, not molybdenum at 3.3. Nitrogen solubility in austenitic grades is however limited to a few hundredths of a percent, so in terms of actual contribution molybdenum does the heavy lifting. The worked numbers below make this explicit.

1.1 Three grades, term by term

1.2 Who actually contributes the 5.44 points

Decomposing the difference between 316L and 304 term by term is more informative than the PREN values themselves:

316L actually carries about one percentage point less chromium than 304 (17.2 versus 18.2 at typical specification levels) and slightly less nitrogen. The entire 5.44 point net gain comes from molybdenum at +6.60. 316L is therefore not an across-the-board upgrade of 304; it is 304 whose chromium shortfall is more than paid back by molybdenum. This also explains why the advantage shows up only where chlorides are present: molybdenum acts on chloride attack, so in dry inland atmospheres or ordinary organic media the two grades behave far more similarly.

2. Why molybdenum works against chlorides

Mechanically, molybdenum enriches the passive film as molybdate at defect sites, blocking chloride ingress both electrostatically and sterically. It raises the pitting potential so local breakdown is harder to initiate, and it accelerates repassivation where the film has already been damaged.

Mechanisms do not go into a specification. Numbers do, so the next step is to convert molybdenum into a chloride concentration.

3. The core question: how much chloride capacity does one PREN point buy

A widely used engineering rule gives chloride limits for the two grades: below 200 ppm for 304 in fresh service and below 1000 ppm for 316L in slightly brackish service. Published supplier ranges put 304 at roughly 200 to 500 ppm and 316L at roughly 1000 to 2000 ppm. This article takes the conservative ends, 200 and 1000 ppm, as design values.

Two points define an exponential relation between threshold concentration and PREN:

Here 0.69897 is the common logarithm of 5. In words: one PREN point multiplies the tolerable chloride concentration by 1.34. Five points give about 4.3 times; 5.44 points give exactly 5.0 times.

3.0 A prerequisite that is easy to miss: which composition set to use

Chloride limits for the same grade differ by a factor of two between sources, 100 ppm and 200 ppm both being quoted for 304. The disagreement is not about test method but about composition basis. A Plymouth Tube technical paper states it plainly: maximum chloride levels for the 300 series today are roughly 50 percent of what was accepted fifteen to twenty years ago, because improved melting practice and competition have pushed typical chemistry down to the ASTM minimum. Twenty years ago a typical 304 carried about 19 percent chromium and a typical 316 about 2.6 percent molybdenum. The paper therefore requires PREN for the 300 series to be calculated from ASTM minimum values.

Recomputing with the minimum-basis compositions used in a mill data sheet gives a useful cross-check on section 1:

On the minimum basis 304 reads 18.96 and 316 reads 24.23, a gap of 5.27 against 5.44 from typical values in section 1. The chromium term falls 1.5 points further behind while molybdenum still contributes +6.93, so the net gain still comes entirely from molybdenum. The conclusion holds under both composition bases. Specifications should therefore state a minimum molybdenum content and require PREN to be calculated from the mill certificate, not from a grade name alone.

3.1 Testing the fit on a grade that was not used to build it

The fit uses only 304 and 316L as anchors. 2205 took no part in it. Substituting PREN = 34.95:

in ppm. Natural seawater carries about 19000 ppm of chloride. The extrapolated value differs from the measured seawater figure by 2.0 percent.

That tidy agreement has to be discounted, and the disagreement is set out here rather than hidden. Published chloride limits for 2205 fall into two camps:

The mechanism behind the split is known: the calculated PREN represents the austenite phase of a duplex steel, not the bulk alloy. Roughly half the structure is ferrite, which carries markedly less chromium and molybdenum and is the preferred site for pit initiation. Extending an austenite-calibrated exponential relation to duplex therefore bias results optimistically.

The proper check is a third austenitic grade. Using 317L from the same mill sheet (Cr 18.5, Mo 3.1, N 0.06, PREN 29.7, chloride capability quoted at about 5000 ppm):

The model gives 4102 ppm against about 5000 ppm from the mill sheet, a difference of 18 percent. Within the austenitic family that is the realistic accuracy, not 2 percent.

Conclusion: use this exponential relation to rank austenitic grades (304, 316L, 317L). Extrapolating it to duplex overstates performance. For seawater immersion, crevices or elevated temperature, rely on measured ASTM G48 CPT or CCT.

3.2 Inverting it: what PREN survives 6 percent ferric chloride

The same relation answers a practical question, how severe the standard laboratory solution really is. ASTM G48 specifies 6 percent FeCl3 prepared on an anhydrous basis, that is 60 g per litre:

That is 39.34 g of chloride per litre, or 39344 ppm, and 2.29385 is the common logarithm of 196.72.

A PREN of roughly 37.3 is needed to resist 6 percent FeCl3 at room temperature. Measurements agree: 2205 at PREN 35.0 shows a CPT of 30 to 50 C, just above ambient, while 2507 at PREN 41.9 shows 50 to 80 C with real margin. Trend and magnitude both match.

4. Critical pitting temperature: four independent sources side by side

CPT, determined by ASTM G48, is the lowest temperature at which stable pitting appears in the specified solution. It is a measured value, sensitive to surface finish, heat treatment and inclusion content, so sources differ systematically. Four are listed here rather than one being picked.

Source E is a mill data sheet measured in 6 percent FeCl3 for 72 hours on minimum-basis compositions, 304 at Cr 18.0 with N 0.06 and 316 at Cr 16.5, Mo 2.1, N 0.05. The same sheet gives 317L at PREN 29.7 and 18.9 C, 904L at PREN 36.2 and 40.0 C, and AL-6XN at PREN 44.5 and 65.0 C.

The sources are offset from one another by 10 to 15 C, mainly because of method differences - Method A screens at a fixed temperature while Method E steps in 5 C increments - and surface preparation, since 120 grit wet ground is not equivalent to a 2B mill finish. Their slopes, however, agree closely.

4.1 The slope is the usable part

Taking the 304 to 316L segment:

Source E comes from the mill sheet, using 316 at PREN 24.2 and 15.0 C against 904L at PREN 36.2 and 40.0 C:

2.08 C per PREN point, within 1 percent of the 2.1 adopted here. Of the five independent sources this one has the most complete method statement, and it is the strongest support for the adopted value.

The slope spans 1.67 to 2.78 C per PREN point with a median near 2.18. Taking 2.1 C per PREN point as the design value:

That is the second quantity the 5.44 PREN points buy: about 11 C of critical pitting temperature.

Note the ordering, which is where many summaries go wrong: 304 always has the lowest CPT, 316L sits above it, 2205 above that, and 2507 highest of the four. A CPT of 50 to 60 C for 304 in 1000 ppm chloride, sometimes seen in circulating tables, is not consistent with any of the four sources above and would make 304 outperform 316L.

5. Matching the criterion to the exposure

This is the correction the article most wants to make, and it is the real answer to the question in the title.

5.1 Atmospheric exposure is classified by ISO 9223

Chloride load in air is measured by the wet candle method (ASTM G140 / ISO 9225) and classified into four levels by ISO 9223:

The figure often quoted for sites within 100 m of the shore, 300 to 1500 mg/(m2*d), is precisely class S3 of ISO 9223. One caveat: the wet candle classification is intended for sites closer than 1 km to a seawater source; beyond that, other deposition methods should be used.

Published distance-to-deposition correlations, drawn from atmospheric exposure tests:

The last column is aesthetic life, the time to visible rust staining, not structural life, which in the same environment is typically an order of magnitude longer. It also shows that any life multiple depends heavily on distance from the sea and is meaningless without it.

5.2 Applying CPT to a facade produces an absurd answer

First, how hot does a coastal facade actually get. Using the sol-air approach:

with I at 800 W/m2 for summer and an external surface film coefficient of 25 W/(m2*K). Term by term:

Yet 304 has a CPT of only 0 to 20 C and 316L only 10 to 30 C. Judged by CPT alone, both should pit immediately on a summer facade.

They do not: a great deal of 316L coastal cladding performs for years. The contradiction comes from a mismatch of criteria. CPT is measured under full immersion with a permanent electrolyte, whereas atmospheric exposure is wet-dry cycling with an electrolyte present only part of the time. Under the ISO 9223 time-of-wetness classes tau3 (3 to 30 percent) to tau4 (30 to 60 percent), a South China coastal site sees roughly 110 to 164 effective wet days per year, each wet period lasting only a few hours, often too short for pit incubation to complete.

★ One widely circulated figure also needs correcting. It is often said that the CPT of 316L in 1000 ppm chloride is 20 to 30 C. That misreads the criterion: 20 to 30 C is the measured CPT of 316L in 6 percent FeCl3, about 39344 ppm, not at 1000 ppm. Published measurements give an effective CPT of about 45 C for 316L at 1000 ppm, pH 7, aerated, falling to 15 to 20 C only in seawater at 19000 ppm. Chloride concentration and CPT are strongly coupled; quoting CPT without the concentration is meaningless.

5.3 The answer to the title

So the correct statement of why a 5 point PREN gap becomes a threefold life difference is:

• It is not "corrodes three times slower". In the passive state both grades corrode extremely slowly; the difference is not in rate.
• It is a fivefold shift in threshold. Chloride capacity moves from 200 ppm to 1000 ppm, which decides whether the surface survives until the next wash.
• It is about 11 C of CPT. In immersed, buried or water-trapping locations, that 11 C decides whether the component sits in the safe region all year or in the pitting region all year.
• Crossing the threshold is a step change, not a gradual one. Once a pit initiates it becomes autocatalytic: metal ion hydrolysis drives pit pH down to 1 to 3, chloride migrates in to balance charge, and repassivation fails. The attack then goes deep. Life differences therefore come out multiplicative, which a linear reading of a 5 point PREN gap cannot predict.

As for the factor of three itself: it is a field observation that varies between roughly 2 and 5 depending on surface condition, washing frequency, orientation and sheltering. It must not appear in a design document as a guaranteed value. Published figures for washing put the effect at 30 to 50 percent longer life for washing every six months, 50 to 80 percent for every three months, and more than double for monthly washing. The International Stainless Steel Forum also notes that surfaces washed by natural rain outlast sheltered ones such as the underside of a balcony, the difference being whether the salt is rinsed away. Any statement of service life is meaningless without the cleaning regime next to it. Design may only use the threshold concentration from section 3 and the CPT from section 4.

6. How concentrated is the surface film in an S3 environment

Converting deposition rate into film concentration shows why even 316L has limits. Taking a mid-range S3 deposition of 800 mg/(m2*d):

in mg per g, corresponding to 8000, 40000 and 80000 ppm.

So in an S3 environment the surface film naturally sits between 10^3 and 10^5 ppm of chloride, which exceeds both the 200 ppm threshold of 304 and the 1000 ppm threshold of 316L. What differs is which grade crosses first, how fast a stable pit forms once crossed, and whether repassivation can occur during the dry part of the cycle.

Local concentration makes it worse. Crevices, gasket faces and weld laps are not flushed by rain, and evaporation can concentrate them 10 to 100 times above bulk. Measured concentrations under salt deposits in splash zones exceed 100000 ppm, consistent with the estimate above. That is why a system whose bulk water is within specification still leaks at the flanges: the criterion has to be taken at the crevice concentration, not the bulk one. It also explains why crevice attack, which initiates at a lower chloride level than pitting, is the more common failure mode in coastal work.

7. ASTM A967 passivation: the windows must be quoted method by method

Passivation removes free iron and enriches the surface in chromium. ASTM A967 defines discrete methods in which concentration, temperature and time are bound together. Merging the windows of different methods into one broad range is the most common quotation error.

7.1 Nitric acid methods

7.2 Citric acid methods

Three practical notes:

• "Nitric acid 20 to 50 percent at 21 to 60 C" is wrong. Concentrations above 45 percent occur only in Nitric 4, whose temperature window is 49 to 54 C, not 21 to 60 C. A bath at 30 percent and 40 C belongs to no method in the standard.
• Citric passivation is not "nitric with the time doubled". A 30 C bath falls under Citric 3 at 20 min minimum; where heating is available, Citric 2 at 50 C for 10 min is far more productive.
• Nitric 1 contains hexavalent chromium and is restricted under EU REACH. Use Nitric 3 or a citric method for projects with environmental requirements or EU destinations.

8. Verifying passivation: ferroxyl, copper sulfate and the Cr/Fe ratio

A further quantitative indicator is the chromium to iron ratio of the passive film, measured by XPS or AES to a depth of about 1 to 3 nm against a total film thickness of only 2 to 5 nm. Published typical ranges are 1.4 to 1.6 after nitric passivation and 1.8 to 2.5 after citric passivation, with Cr/Fe of at least 1.0 widely accepted as the lower limit of acceptance.

Quoting 1.0 to 1.5 as the complete acceptance band is therefore incomplete: 1.5 is merely the upper end of the typical nitric range, while citric passivation should reach 1.8 or better to justify the process. Note too that Cr/Fe is an element ratio, not a film thickness. Two surfaces can share the same ratio with film thicknesses differing by a factor of two, so the ratio is a process consistency check, not a life predictor.

9. Weld temper colour: three grades and what to do

At welding temperature chromium oxidises preferentially, leaving a chromium-depleted layer beneath the oxide. The effective PREN there is below that of the parent metal, and pitting usually starts in the temper band. AWS D18.2 sorts the colours into three levels:

Temper colour is thin-film interference from an oxide layer tens to a few hundred nanometres thick. The darker the colour, the greater the heat input, the thicker the oxide and the deeper the chromium depletion beneath it. Only two removal routes work: mechanical grinding to bright metal, or chemical treatment with pickling and passivation paste. Waxing, oiling or painting hides the colour without restoring corrosion resistance, which makes it the most readily accepted and most dangerous site practice.

The acceptable level also depends on service. Food, pharmaceutical and bioprocess equipment normally accepts level 1 only; architectural cladding with specified passivation and periodic washing can accept level 2 after removal.

10. Common mistakes

1. Mistake: using an immersion-service CPT to select coastal cladding. Consequence: it suggests both grades pit so the choice does not matter, or pushes the specification into needless over-alloying. Correct approach: judge atmospheric exposure by the ISO 9223 chloride class together with time of wetness and surface condition; reserve the chloride threshold and CPT for immersed, buried or water-trapping details.
2. Mistake: specifying the grade without the surface condition and passivation requirement. Consequence: within one grade, unpassivated and passivated surfaces sit a class apart, and a level 3 temper band left in place drops the local PREN below that of 304. Correct approach: state grade, surface roughness (Ra below 0.5 micrometre is suggested), the ASTM A967 method and the acceptance test together.
3. Mistake: selecting against bulk water chloride. Consequence: bulk compliant, flanges leak first. Correct approach: take the criterion at the crevice concentration, allowing 10 to 100 times concentration.
4. Mistake: merging temperature and time windows from different passivation methods. Consequence: the bath parameters match no standard method and acceptance has no basis. Correct approach: quote Nitric 1 to 4 or Citric 1 to 3 individually with concentration, temperature and time together.
5. Mistake: separating 304 from 316L with a magnet. Consequence: both are non-magnetic when annealed and both become weakly magnetic at cold-worked bends through deformation-induced martensite, so the test decides nothing. Correct approach: read the mill certificate, or use a portable analyser for molybdenum (2.0 percent or more) and nickel.
6. Mistake: writing "316L lasts three times longer than 304" into a design document. Consequence: the multiple is a field observation varying between 2 and 5 with washing and orientation, and cannot be guaranteed. Correct approach: specify only the two measurable criteria, threshold concentration and CPT.

In one sentence: in PREN = Cr + 3.3*Mo + 16*N, the 5.44 points that 316L holds over 304 come entirely from molybdenum, and they buy not a slower corrosion rate but a chloride threshold moved from 200 ppm to 1000 ppm - 1.3443 times per PREN point, extrapolating to 19384 ppm for 2205 against 19000 ppm in seawater, a 2.0 percent deviation - plus about 11.4 C of critical pitting temperature. Crossing a threshold is a step change, so the life difference is multiplicative rather than a linear share of five points.

FAQ

What are the PREN values of 316L and 304, and where does the difference come from?

Using PREN = Cr + 3.3*Mo + 16*N, 304 with Cr 18.2, Mo 0.1 and N 0.06 gives 19.49, and 316L with Cr 17.2, Mo 2.1 and N 0.05 gives 24.93, a gap of 5.44. Broken down, the chromium term is 1.00 lower and the nitrogen term 0.16 lower in 316L; all of the gain comes from molybdenum at +6.60. The advantage therefore shows up only in chloride-bearing service.

How much more chloride does 316L tolerate than 304?

On the conservative design figures, 304 is limited to about 200 ppm and 316L to about 1000 ppm, a factor of 5.0, which is 1.3443 times per PREN point. Anchored on those two points, the fit extrapolates to 19384 ppm for 2205, which was not used in fitting, against about 19000 ppm in natural seawater, a 2.0 percent deviation. Use it for ranking and selection only; it does not cover sensitised welds or crevice concentration.

Which CPT is higher, 304 or 316L, and how much higher?

316L is always higher. Across four independent sources the CPT of 304 sits between below 0 and 20 C, while 316L sits between 10 and 30 C. The slope of CPT against PREN runs from 1.67 to 2.78 C per point with a median near 2.18, so a 5.44 point PREN gap gives roughly 11.4 C, within a range of 9.1 to 15.1 C. Any table that puts 304 above 316L in CPT is inconsistent with all four sources.

What are the correct ASTM A967 passivation parameters and acceptance criteria?

Concentration, temperature and time must be quoted as a set and never merged. Nitric 1 is 20 to 25 % v/v nitric acid with 2.5 % w/w sodium dichromate at 49 to 54 C for 20 min; Nitric 2 is 20 to 45 % at 21 to 32 C for 30 min; Nitric 3 is 20 to 25 % at 49 to 60 C for 20 min; Nitric 4 is 45 to 55 % at 49 to 54 C for 30 min. Citric 1, 2 and 3 are all 4 to 10 % w/w, at 60 to 71 C for 4 min, 49 to 60 C for 10 min and 21 to 49 C for 20 min respectively. Acceptance is commonly Practice F free iron with no blue spot in 30 seconds, or Practice D copper sulfate with no copper deposit in 6 minutes. The film Cr/Fe ratio should be at least 1.0, with 1.4 to 1.6 typical after nitric and 1.8 to 2.5 after citric passivation.