Views: 0 Author: J-VALVES Publish Time: 2026-11-16 Origin: Site
The purchase order said Class 150. The line runs at 12 bar gauge and 300 deg C. The valve passed its shell test at the factory, arrived with a certificate, was installed, and within three weeks the bonnet joint was weeping and the seat was passing. Nobody did anything obviously wrong. The specification was written as it has been written for thirty years, and the valve was built to the standard named in the specification. What failed was an assumption that sits in the middle of the industry and almost never gets spoken aloud: that a pressure class is a pressure.
It is not. A class number is a row index into a table of allowable working pressures that changes with temperature and with body material. Class 150 carbon steel is rated 19.6 bar at 38 deg C and 10.2 bar at 300 deg C, and the 10.2 bar is the number that governs whether the valve may be installed on that 12 bar line. This article works through how the numbers are built, how to read the table without the four mistakes that account for most of the failures, and how to write a specification line that cannot be misread. It applies equally to gate, globe, check, ball and plug valves, because they share the same rating tables - which is exactly why a rating error propagates across a whole valve schedule instead of staying in one line item.
Class 150 means 285 psi at the base temperature, not 150 psi. The ambient rating in bar is 19.6. Class 300 means 740 psi, or 51.1 bar. Class 600 means 1480 psi, or 102.1 bar. The class numbers are legacy labels - they come from an era when flanges were sized for saturated steam at a stated pressure - and the values behind them were rebuilt decades ago around allowable stress. The labels survived the rebuild; the arithmetic did not.
Class | Base rating at 38 deg C (psi) | Same value (bar) | Where the class exists |
|---|---|---|---|
150 | 285 | 19.6 | ASME B16.5 flanges, all valve types |
300 | 740 | 51.1 | ASME B16.5 flanges, all valve types |
400 | 990 | 68.3 | B16.34 and B16.5, rarely stocked outside power and refining |
600 | 1480 | 102.1 | ASME B16.5 flanges, all valve types |
800 | 1975 | 136.2 | B16.34 only, for small forged valves with threaded or socket weld ends |
900 | 2220 | 153.2 | ASME B16.5 flanges, all valve types |
1500 | 3705 | 255.3 | ASME B16.5 flanges, all valve types |
2500 | 6170 | 425.5 | ASME B16.5 flanges up to NPS 12 |
4500 | about 11,100 | about 767 | B16.34 only, ultra-high-pressure service |
Two consequences follow immediately, and both cost money when they are missed. The first is that a buyer who compares "Class 150" against "PN 16" because the numbers look similar is comparing 19.6 bar with 16 bar at room temperature and 10.2 bar with 11.0 bar at 300 deg C - the ranking reverses. The second is that a specification which says only "Class 150" is not a specification of pressure at all. It is a specification of geometry. The pressure has to be looked up.
Every value in the B16.34 table is the lowest of three limits for that class, size and material at that temperature:
Shell wall thickness, calculated from the allowable stress of the body material at temperature, using a thin-wall formula of the form t = P d / (2 S E - 1.2 P) + c, where P is pressure, d is the inside diameter, S the allowable stress, E the casting quality factor and c the corrosion allowance. There is also a tabulated minimum wall that applies regardless of calculation, which is why small valves are often thicker than the pressure requires.
Flange strength and joint tightness, taken from the flange rating tables for the same class and material group. On a flanged valve this is usually the governing element, because flanges are deliberately conservative.
Bolting, meaning the ability of the bonnet and flange bolts to hold the joint closed at temperature, allowing for creep relaxation of the bolts and differential expansion between bolt, flange and gasket.
Because the flange governs in most sizes, a Class 150 valve is not a thin, light shell that happens to have flanges on it. It is a flange-limited assembly. This matters in three practical ways. First, a welded valve of the same class can be built lighter than a flanged one and still be compliant, which is why butt-weld valves can sometimes carry higher ratings - the Special and intermediate ratings discussed in section 9. Second, the rating guarantees the pressure boundary, not the joint's tightness in service: a correctly rated Class 300 flange pair can still weep on a hot line because the bolts relaxed, and that is a bolting and gasket problem, not a class problem. Third, the rating says nothing about the seat. The table rates the shell at metal temperature. Seat tightness is a separate performance question governed by API 598 and the leakage classes in ISO 5208, which is why a valve can be perfectly rated and still fail acceptance - see the notes on valve seat leakage classes and on API 598 pressure testing.
The table is not one table. B16.34 is organised by material group, and the group number - not the grade name - is what you look up. Two valves with the same class, the same size and different body alloys can legitimately carry different allowable pressures, and the difference grows with temperature.
Group | Typical grades | Temperature character | What changes |
|---|---|---|---|
1.1 | A216 WCB, A105, A350 LF2 | Base table, -29 to 425 deg C in practice | The reference case for most carbon steel valves |
1.2 | A216 WCC, A352 LCC, A350 LF6 | Similar envelope, slightly higher ambient rating | Class 150 base rating 20.0 bar instead of 19.6 |
1.x chromium-molybdenum | A217 WC6, WC9, C12A, A182 F11, F22, F91 | Retains useful pressure well past 425 deg C | The route for steam above the carbon steel ceiling |
2.1 | A351 CF8, A182 F304 | Austenitic, floor far below -29 deg C | Higher rating than carbon steel above roughly 350 deg C |
2.2 | A351 CF8M, A182 F316 | Austenitic, the workhorse corrosion grade | Same shape of curve as 2.1, marginally different values |
2.3 and 2.4 | CF3, CF3M, F304L, F316L | Low-carbon austenitic | Lower allowable stress, so lower ratings at high temperature |
duplex group | A182 F51, F53, F55, A995 grades | Own group entry with its own curve | Higher strength, but the temperature ceiling is lower than austenitic |
3.x nickel alloys | Monel, Inconel, Hastelloy, alloy 20 | Own group entries | Ratings that do not follow either steel family |
The group number is printed nowhere on a standard valve, and this is where substitution damage happens. A supplier offers "equivalent material" to protect a delivery date - CF8M in place of a low-carbon CF3M, or a WCC body in place of WCB - and the paperwork still says Class 300. Sometimes the substitution is genuinely equivalent. Sometimes it moves the valve into a different group, and the allowable pressure at the operating temperature changes without any change to the class, the drawing number or the invoice description. The defence is procedural rather than technical: require the group number to be stated on the datasheet, require the body grade to be verified on receipt, and treat a grade change as a design change even when the class is unchanged. The certification mechanics are covered in the note on EN 10204 material certificates; the point here is simply that the rating follows the group, and the group follows the grade.
Group 1.1 carbon steel covers the great majority of process valves in refinery, water, steam and general industry service. These are the Standard Class ratings in bar gauge, and they are the numbers a valve schedule should be checked against:
Metal temperature | Class 150 | Class 300 | Class 600 | Class 900 | Class 1500 | Class 2500 |
|---|---|---|---|---|---|---|
-29 to 38 deg C | 19.6 | 51.1 | 102.1 | 153.2 | 255.3 | 425.5 |
50 deg C | 19.2 | 50.1 | 100.2 | 150.4 | 250.6 | 417.7 |
100 deg C | 17.7 | 46.6 | 93.2 | 139.8 | 233.0 | 388.3 |
150 deg C | 15.8 | 45.1 | 90.2 | 135.2 | 225.4 | 375.6 |
200 deg C | 13.8 | 43.8 | 87.6 | 131.4 | 219.0 | 365.0 |
250 deg C | 12.1 | 41.9 | 83.9 | 125.8 | 209.7 | 349.5 |
300 deg C | 10.2 | 39.8 | 79.6 | 119.5 | 199.1 | 331.8 |
350 deg C | 8.4 | 37.6 | 75.1 | 112.7 | 187.8 | 313.0 |
400 deg C | 6.5 | 34.7 | 69.4 | 104.2 | 173.6 | 289.3 |
425 deg C | 5.5 | 28.8 | 57.5 | 86.3 | 143.8 | 239.7 |
450 deg C | 4.6 | 23.0 | 46.0 | 69.0 | 115.0 | 191.7 |
500 deg C | 2.8 | 11.8 | 23.5 | 35.3 | 58.8 | 97.9 |
Read the Class 150 row and then read it again. From 38 deg C to 400 deg C the allowable pressure falls by two thirds. A purchaser who selects on the first number is selecting a valve for a service it will not see. And note what happens above 425 deg C: the values continue to fall steeply and the standard adds a caution that prolonged exposure above about 425 deg C may convert the carbide phase of the steel to graphite. The rating is still permissible; it stops being recommended. In practice 425 deg C is the ceiling for carbon steel, and above it the schedule moves to a chromium-molybdenum group - which is a material decision driven by the rating table, not an arbitrary preference for alloy steel.
Interpolate, do not extrapolate. A valve at 275 deg C sits between the 250 deg C and 300 deg C rows. Linear interpolation gives 12.1 - (25/50) x 1.9 = 11.2 bar for Class 150, which is acceptable for engineering selection. Above the last tabulated temperature there is nothing to interpolate towards, and a valve cannot be rated beyond the material's listed limit by extrapolating the curve.
Use metal temperature, not fluid temperature. The table is indexed on the temperature of the pressure-retaining part. On an uninsulated line that is the fluid temperature, so the two coincide. On an insulated line the metal still runs close to fluid temperature. On a cold line in a hot climate, on a valve downstream of a desuperheater, or on a valve heated by a nearby hot line, the metal can be hotter than the process data suggests. Where the two differ, the metal temperature is the one to use.
The rating carries no allowance for external loads. It assumes pressure alone. Piping reactions, flange misalignment, water hammer and vibration all consume margin that the table does not provide for. A valve chosen exactly at the table value on a line with heavy nozzle loads is overstressed in practice even though it is compliant on paper.
Below minus 29 deg C the rating is frozen, not extended. The standard does not rate Group 1.1 carbon steel below -29 deg C; the value at -29 deg C is the floor, and going colder requires impact-tested material rather than a bigger class. Section 14 deals with the cold end separately because the logic is different from the hot end.
Take a saturated steam header at 12 bar gauge. Saturation temperature at 12 bar gauge is about 191 deg C. Interpolating between the 150 deg C and 200 deg C rows, Class 150 carbon steel is rated about 14.2 bar. The margin is thin but the valve is compliant, and this is why Class 150 carbon steel has been used on medium-pressure steam for a century.
Now superheat the same header to 300 deg C, keeping 12 bar gauge. Class 150 carbon steel is rated 10.2 bar. The valve is now 1.8 bar short of its own service condition - a 15 percent shortfall - and nothing about the order, the invoice or the nameplate has changed. The failure mode is not a burst. It is a slow sequence of gasket weeping, packing extrusion and seat distortion, which is why these errors are usually discovered as maintenance cost rather than as an incident.
Condition | Class 150 WCB rating | Class 300 WCB rating | Verdict for 12 bar gauge |
|---|---|---|---|
12 bar gauge, saturated, 191 deg C | about 14.2 bar | about 44.0 bar | Class 150 compliant, low margin |
12 bar gauge, superheated, 300 deg C | 10.2 bar | 39.8 bar | Class 150 fails by 1.8 bar, Class 300 compliant |
12 bar gauge, 350 deg C | 8.4 bar | 37.6 bar | Class 150 fails by 3.6 bar |
12 bar gauge, 400 deg C | 6.5 bar | 34.7 bar | Class 150 fails by 5.5 bar, consider Cr-Mo above 425 deg C |
The lesson is not "buy Class 300". It is that the temperature column decides, and the temperature column is the one thing a valve schedule usually omits. A schedule that states design pressure, design temperature and class together can be checked in seconds. A schedule that states only class cannot be checked at all.
A vessel has a maximum allowable working pressure of 15 bar gauge and a pressure relief valve set at 15 bar gauge. The isolation valve on the relief line is a gate valve, and the line relieves at 250 deg C. What class does the gate valve need?
Not Class 150. The relief valve is permitted to accumulate 10 percent above set pressure before it passes full flow, so the isolation valve can see 16.5 bar gauge, and it sees it at the relieving temperature. At 250 deg C, Class 150 carbon steel is rated 12.1 bar - 4.4 bar short. Class 300 at 250 deg C is rated 41.9 bar and covers it comfortably.
Two subtleties are worth stating because they are widely missed. First, the block valve must be rated for the relieving condition, not the normal operating condition, and the relieving temperature is often higher than the operating temperature because the relief case is usually a upset. Second, a fire case cannot be solved with a class upgrade. In a fire the metal temperature can exceed the material's tabulated range entirely, and B16.34 offers no rating there; industry practice is to accept local yielding under fire exposure rather than to specify a higher class. If the relief case is fire, the answer is a fire-safe qualified valve under API 607 and a correctly sized relief path, not a Class 2500 body. The sizing and set pressure logic belongs to the relief valve itself, covered in the guide to pressure safety valve selection.
Because the ratings are driven by allowable stress, and because austenitic stainless steel loses allowable stress more slowly with temperature than carbon steel does, the ranking of two bodies in the same class reverses as temperature rises. Expressed as the share of the 38 deg C rating that survives at temperature:
Metal temperature | Carbon steel (Group 1.1), share of ambient rating | Austenitic 316 body (Group 2.2), approximate share |
|---|---|---|
38 deg C | 100 percent (19.6 bar) | 100 percent (about 19.0 bar) |
200 deg C | 70 percent (13.8 bar) | about 84 percent |
300 deg C | 52 percent (10.2 bar) | about 66 percent |
400 deg C | 33 percent (6.5 bar) | about 60 percent |
425 deg C | 28 percent (5.5 bar) | about 55 percent |
The stainless figures are indicative, because the exact values differ between CF8 and CF8M and between tabulated editions, and because a low-carbon grade such as CF3M sits slightly lower again. The pattern, though, is robust and it is the engineering point: at 400 deg C a Class 150 stainless valve holds roughly twice the pressure of a Class 150 carbon steel valve of the same nominal class. A schedule that says "Class 150" without a material is therefore not one specification but two, and the difference at temperature is a factor of two.
This cuts both ways in procurement. A stainless body is often selected for corrosion resistance and then assumed to be pressure-equivalent because the class matches. At ambient it is very slightly weaker - about 19.0 bar against 19.6 bar - so "equivalent" is not quite true at the bottom of the table either. The correct statement is that the two bodies have different rating curves that happen to intersect, and the edition of the standard in force is the only authority for the numbers.
B16.34 defines three rating designations, and the difference is not marketing. Special Class allows higher ratings than Standard Class at the same temperature, in exchange for mandatory non-destructive examination of the pressure boundary - volumetric examination by radiography or ultrasonics plus surface examination by penetrant or magnetic particle - and the valve is marked accordingly.
Designation | Rating basis | NDE requirement | Where it can be applied | Marking |
|---|---|---|---|---|
Standard Class | The base table | None beyond the standard's own requirements | Any valve, including flanged | B16.34 STD |
Special Class | Higher table, roughly 20 to 25 percent above Standard at the same temperature, material dependent | Volumetric plus surface examination of body and bonnet | Threaded and welding end valves; not flanged | B16.34 SPL |
Limited Class | Special Class tables applied to small valves | None, but requires design analysis | Threaded or socket weld valves NPS 2 1/2 and smaller | B16.34 LTD |
The practical limit is the one that surprises buyers: you cannot buy a Special Class flanged valve. The flange rating caps the assembly, so the higher table buys nothing on a flanged body, and the designation is restricted to threaded and welding end valves. If a schedule is stuck just above a Standard Class ceiling at temperature, the honest options are to move up a class, change the material group, or switch to a welded valve where Special or intermediate ratings become available - not to ask for a "special class flanged" valve that cannot exist. Suppliers who accept that order are usually delivering Standard Class with a certificate that says otherwise, which is a documentation problem discovered at the worst possible moment.
Three classes sit outside the everyday flange catalogue and cause confusion out of proportion to their number.
Class 800 has no ASME B16.5 flange equivalent. It exists in B16.34 for small forged valves with threaded or socket weld ends, which is why so many NPS 1/2 to 2 forged gate, globe and check valves are marked 800 - 1975 psi at 38 deg C, or 136.2 bar. A buyer who needs to connect a Class 800 valve to a flange must use a welded or threaded transition, because there is no Class 800 flange to bolt to. The forged body route and its dimensional envelope are covered in the comparison of forged versus cast valve bodies.
Class 400 is a real class with a real table - 990 psi at 38 deg C, two thirds of Class 600 - but it is stock-heavy in only a few industries. It is most useful as an intermediate step when Class 300 is short at temperature and Class 600 is a large cost and weight jump, which on a hot hydrocarbon line is a common situation.
Class 4500 is the ultra-high-pressure class for hydrocracker, polyethylene and similar service. It exists in B16.34 and not in B16.5. It is mentioned here only so that it is not mistaken for a typographical error when it appears in a datasheet.
This is the single most common rating error in water and HVAC work, and it is made by experienced engineers because the vocabulary is shared while the standards are not.
A resilient-seated wafer or lug butterfly valve is typically qualified to MSS SP-67 and API 609, and its rating is stated in psi cold working pressure: commonly 150 psi, sometimes 200 psi in the smaller sizes. That is 10.3 bar, not 19.6 bar. Meanwhile the same valve is described in catalogues as a "Class 150 butterfly valve", which refers to the flange drilling - it will bolt between ASME Class 125 or 150 flanges - and not to the pressure class of the body. A water line designed at 16 bar with a resilient-seated butterfly selected on the words "Class 150" is 5.7 bar outside the valve's rating.
Valve type | Rating basis | Typical statement on the nameplate | What "Class 150" means there |
|---|---|---|---|
Gate, globe, check, ball, plug to B16.34 | B16.34 pressure-temperature table | Class 150, with material and STD/SPL/LTD marking | The full B16.34 rating, 19.6 bar at 38 deg C |
High-performance butterfly to API 609 Category B | Class rating per B16.34, often metal or PTFE seat | Class 150 or 300, with temperature curve | Usually the B16.34 rating, but check the seat and shaft limits |
Resilient-seated butterfly to MSS SP-67 | Cold working pressure, typically 150 psi | 150 psi CWP | Flange compatibility only |
Butterfly to AWWA C504 | Class 150B, 150 psi bidirectional shutoff, 300 psi shell test | AWWA C504 Class 150B | A waterworks class, not an ASME class |
Two further butterfly specifics matter. A wafer valve is clamped between flanges and cannot be used as an end-of-line isolator unless the manufacturer rates it for dead-end service; a lugged body can. And the disc and seat materials usually set the real temperature limit well before the body rating does, because the elastomer seat is the component that softens, swells or takes a compression set. On a butterfly, the pressure question and the temperature question are answered by different components.
European projects use PN designations. There is no exact equivalence, and the cross-reference tables that circulate are approximations built for cost estimating, not for design.
Temperature | Class 150, carbon steel (bar) | PN 16, carbon steel group 3E0 (bar) | PN 40, carbon steel group 3E0 (bar) |
|---|---|---|---|
Room temperature | 19.6 | 16.0 | 40.0 |
100 deg C | 17.7 | 14.8 | 37.1 |
200 deg C | 13.8 | 13.3 | 33.3 |
300 deg C | 10.2 | 11.0 | 27.6 |
400 deg C | 6.5 | 9.5 | 23.8 |
The crossover is the interesting part. At room temperature the ASME Class 150 body is the stronger of the two by 22 percent. By 300 deg C the PN 16 body is ahead, and by 400 deg C it is ahead by nearly 50 percent - because EN 1092-1 derates carbon steel on a shallower curve than B16.34 does. So the familiar claim that "Class 150 is roughly PN 20" is true only at the bottom of the temperature range and only as an approximation of nominal size.
Mechanically, the two are not interchangeable at all. A B16.5 Class 150 flange and an EN 1092-1 PN 16 flange have different outside diameters, bolt circles, bolt sizes and facing details, and they do not bolt together. Face-to-face dimensions come from different standards as well - B16.10 against EN 558 - so a replacement changes the pipe spool. For a mixed-standard plant the datasheet must state which rating system the valve is built to, and a PN valve ordered to replace a Class valve is a piping modification, not a like-for-like swap.
B16.34 and API 598 require a shell test at 1.5 times the 38 deg C rating, and a seat test at 1.1 times it. For Class 150 that means 29.4 bar on the shell and 21.6 bar on the seat, both with water at ambient temperature.
Look at those numbers against the service condition in section 6. The valve that will run at 12 bar gauge and 300 deg C is tested at 29.4 bar cold - nearly three times its hot rating - and its seat is proven at 21.6 bar, twice the pressure it will ever hold in service. The test is a manufacturing quality check, not a demonstration that the valve is suitable for the line. It proves the casting is sound and the seat geometry is right. It says nothing about the joint at 300 deg C, nothing about packing extrusion at temperature, and nothing about the gasket's ability to hold after fifty thermal cycles.
This is why the factory test certificate and the field leak can both be true. The hot joint leaks because the bolts relaxed and the gasket unloaded as the flange pair expanded differentially; the seat passes because the seat materials were at 20 deg C during the test. Run the arithmetic on any high-temperature leak and the test pressure is almost never the issue. The related mechanical question - whether the body is thick enough in the first place, given casting quality factors and examination coverage - is a separate argument, set out in the note on forged and cast bodies linked above.
The cold end does not behave like the hot end. Below the tabulated minimum, the rating does not fall away gradually; it stops, and a different requirement takes over.
For Group 1.1 carbon steel the table begins at -29 deg C, and the rating at -29 deg C is the floor. Nothing in B16.34 extends below it, and no class upgrade helps. What changes instead is the material: low-temperature carbon steel castings to A352 LCB or LCC are impact tested to -46 deg C as part of the grade, 3.5 percent nickel steel to A352 LC3 to about -101 deg C, and austenitic stainless such as CF8M is accepted down to about -196 deg C. The pressure capability at low temperature is not usually the limiting factor - yield strength rises as temperature falls - so the design question becomes toughness, and the acceptance evidence becomes Charpy impact test results at the minimum design metal temperature rather than a table lookup.
Two distinctions are worth keeping separate, because they are routinely merged. The B16.34 table floor is a rating limit. The material's own qualification temperature is a specification limit, and the two differ: an A350 LF2 forging is commonly impact qualified to -46 deg C while its B16.34 rating table starts at -29 deg C. Neither of those tells you the minimum temperature of the assembled valve, because the packing, the seat insert and the body gasket usually reach their own limits first. The full assembly governs, and the material guide for duplex and super duplex seawater service makes the same point from the corrosion side: the body is rarely the only material decision on the valve.
The small-size bolt pattern overlap. From NPS 1/2 to NPS 3, ASME Class 300 and Class 600 flanges share the same outside diameter and bolt circle. They will bolt together, and the joint is then rated only to the lower class. This means an upgrade from Class 300 to Class 600 in small sizes buys a heavier valve that still cannot exceed the Class 300 joint rating if it is mated to an existing Class 300 flange. The class is a property of the pair, not of the valve alone.
The nameplate that says only "150". A compliant B16.34 nameplate carries the class, the body material, the rating designation and the size. A valve marked only with a class number cannot be audited, because the rating depends on the material group. On receipt, the practical check is a positive material identification test against the certificate - and note that a hand-held X-ray analyser reads alloying elements, not carbon, so CF8 and CF3 can look identical to it unless a carbon-specific method is used.
The twin rating on a soft-seated valve. A ball valve may be rated Class 300 at temperature and still be limited by its seat material well below that, because PTFE softens and creeps. The body rating and the seat rating are two different numbers and both belong on the datasheet. The mechanical background is in the guide to ball valve seat design.
The "equivalent" grade substitution. Already covered in section 3, and worth repeating because it is the most common way a compliant valve becomes non-compliant: the class stays, the drawing stays, the group changes, the rating at temperature changes.
Most rating failures are specification failures, and they are cheap to prevent. These clauses, added to a valve datasheet or a purchase order, close the gaps:
Clause | Wording to use | What it prevents |
|---|---|---|
Rating reference | "Rated per ASME B16.34, current edition, at the design temperature stated below" | Selection on the ambient value |
Design condition | "Design pressure 12 bar gauge, design temperature 300 deg C, metal temperature" | A schedule that states only class |
Material group | "Body to A216 WCB, B16.34 Group 1.1" | Silent grade substitution |
Rating designation | "Standard Class (STD)" or "Special Class (SPL)" explicitly | Assuming a premium that was never ordered |
Seat rating | "Seat materials rated for continuous service at 300 deg C at the stated differential pressure" | Body rating masking a soft-seat limit |
Leakage | "Seat test to API 598, acceptance Rate D liquid" or the class actually needed | Acceptance arguments at commissioning |
Minimum temperature | "Minimum design metal temperature -29 deg C, impact tested per B16.34 where required" | A hot-end schedule with no cold-end check |
Flange standard | "Flanges to ASME B16.5 Class 150 raised face" and, for EU projects, "EN 1092-1 PN 16 type B1" - never both as equivalent | Non-mating flanges on site |
Marking | "Nameplate to include class, body material, rating designation and size" | Unverifiable valves on receipt |
Certification | "EN 10204 type 3.1 certificate, heat number marked on the body, PMI on receipt" | Grade substitution discovered in service |
Pull the valve schedule and add two columns: design temperature and required rating at that temperature. Most schedules already have the pressure. Almost none have the temperature beside the class, and that omission is the root of the problem this article describes. Then, for each line, look up the number in the group table for the body material actually being supplied and compare it with the design pressure.
The lines that fail will cluster in a predictable way: steam and hot oil above 250 deg C, hot water above 150 deg C, and anything where the design temperature was written as a round number for margin and then never revisited. The corrections are usually cheap if they are made at the enquiry stage - a class step on a handful of sizes, or a stainless body where the chemistry allows it - and expensive if they are made after the valves arrive, because the flange rating change propagates into the pipe spools on both sides.
For the valves themselves, the gate valve range, the ball valve range and the check valve range are quoted against a stated design temperature rather than a bare class number, which is the only way the rating can be confirmed before the order is placed. Related articles on test pressures, leakage classes, body manufacturing routes and material certification are collected in the technical article library.
What pressure is a Class 150 valve actually rated for?
285 psi at 38 deg C, which is 19.6 bar - not 150 psi and not 20 bar. That is the base value for carbon steel in ASME B16.34 material Group 1.1. The number falls with temperature: 17.7 bar at 100 deg C, 13.8 bar at 200 deg C, 10.2 bar at 300 deg C, 6.5 bar at 400 deg C and 5.5 bar at 425 deg C, which is the practical ceiling for carbon steel. A stainless body in the same class starts slightly lower, at about 19.0 bar, but holds about 11 to 12 bar at 400 deg C where carbon steel has fallen to 6.5 bar. So the answer to the question is not one number - it is a column in a table, read at the metal temperature of the installation.
How do I find the B16.34 material group for my valve?
From the standard itself, using the body grade. Group 1.1 covers A216 WCB castings, A105 forgings and A350 LF2; Group 1.2 covers WCC and low-temperature LCC; austenitic grades sit in Groups 2.1 to 2.4 depending on carbon level, with CF8M typically in 2.2; chromium-molybdenum grades and nickel alloys have their own entries. Do not take the group number from an online comparison chart, because the group assignments are tied to the edition of the standard. Take it from the edition in force on the project, or from the valve manufacturer's rating sheet, and put the group number on the datasheet next to the grade. A grade substitution that changes the group changes the allowable pressure without changing the class on the drawing.
Can I interpolate between the temperatures listed in the table?
Yes, between tabulated rows, and linear interpolation is normal engineering practice: a Class 150 carbon steel valve at 275 deg C sits halfway between 12.1 bar at 250 deg C and 10.2 bar at 300 deg C, so about 11.2 bar. What is not permitted is extrapolation past the last tabulated temperature, or below the minimum. Some project specifications require the next lower tabulated row instead of an interpolated value, which is a conservative project rule rather than a requirement of the standard, and it is worth knowing which convention the specification uses before a rating calculation is submitted for approval.
Is a Class 300 valve always stronger than a Class 150 valve?
In pressure terms, yes for the same material and size, and by a wide margin - 51.1 bar against 19.6 bar at 38 deg C. In practice there are three exceptions worth knowing. A Class 150 stainless body out-rates a Class 150 carbon steel body above roughly 350 deg C, so the comparison depends on material as much as class. A Class 300 valve still carries a seat rating set by its seat material, which can be the real limit on a soft-seated ball valve. And from NPS 1/2 to NPS 3, Class 300 and Class 600 flanges share the same bolt pattern, so a joint between a Class 600 valve and a Class 300 flange is rated only to Class 300 - the class belongs to the pair, not to the valve alone.
Does a higher pressure class solve a fire case or a thermal relief problem?
No, and treating it that way creates a new problem instead of solving the old one. In a fire the metal temperature can go beyond the material's tabulated range altogether, so there is no B16.34 rating to move up to; the accepted practice is to qualify the valve for fire performance under API 607 and to size the relief path correctly, allowing local yielding during the event. Trapped liquid between two closed seats is a different case again: the pressure rise comes from thermal expansion of the liquid, can reach tens of bar for a ten degree rise, and is independent of the class of the valve. The fix is a cavity relief path or an external relief device, not a heavier body.