8 Inch Class 2500 High Temperature Trunnion Ball Valve, Metal Seated

An 8 inch Class 2500 high temperature trunnion ball valve is a quarter-turn isolation valve in which a 200 mm ball rides on two trunnions and seals against metal seats inside a forged 1.25Cr-0.5Mo body. At 38°C that body holds 425 bar. At 425°C it holds about 310 bar, because a class is a pressure at a temperature and not a number stamped on a nameplate.

J-VALVES manufactures this size in forged ASTM A182 F11, with Stellite 6 or tungsten carbide overlays on ball and seat, RTJ ends to ASME B16.5, and a bolted three-piece body that can be opened without cutting the line. The buyers are HRSG and boiler packagers, refinery revamp engineers and coal chemical plants moving high pressure steam, hydrogen or syngas. This page works through what 425°C does to the design, where the seat material cut-off actually sits, how the valve is proved at 638 bar, and the eight numbers a quotation needs.

What an 8 Inch Class 2500 Valve Has to Survive

Start with the load, because it explains every other decision on the page. A full bore 8 inch valve has a bore area of 0.0314 m². At the Class 2500 rating of 42.55 MPa the closed ball carries 1.34 MN of thrust, about 136 tonnes, pushing it downstream. The same pressure acting on the annular face of a 240 mm seat ring, roughly 0.0138 m², produces about 590 kN of seat load, and that figure is what the seat carrier, the spring set and the gearbox are all built around.

Now add the thermal side. 1.25Cr-0.5Mo expands by about 0.54 percent between 20°C and 425°C, which is 1.1 mm across a 200 mm ball. The ball, the seat ring, the carrier and the body pocket are not the same alloy and are not at the same temperature during a warm-up, so that millimetre does not distribute itself evenly. It is absorbed in two places only: spring preload behind the seat, and clearance designed into the pocket. When neither is sufficient the valve either loses preload and leaks, or closes so hard that the operator stalls on the first cool-down.

Why Trunnion Mounted and Not Floating at 8 Inch

A floating ball seals because line pressure pushes it into the downstream seat. That is cheap and reliable to about 6 inch Class 600, and it stops being either at 8 inch Class 2500, because the stem and the seat would have to carry the 1.34 MN above. A trunnion design takes the thrust off the seat: the ball turns on an upper and a lower shaft running in bearings, the pressure load passes through those bearings into the body wall, and each seat seals on its own spring load.

Three things follow. Torque falls, because the ball is no longer being forced into the seat by pressure. The stem sees torsion and nothing else, so it stays a sensible diameter. And the seats can be given a piston effect, which is what makes self-relief, double block and bleed and double isolation possible in the first place. On this size the trunnion shafts are typically 60 to 80 mm, running in self-lubricating bearings with a steel backing; a plain bush extrudes at these loads within a few hundred cycles.

Below 6 inch the arithmetic reverses and a trunnion body becomes the expensive answer to a problem that does not exist. The trunnion mounted ball valve range starts at 2 inch at J-VALVES. At 8 inch Class 2500 the design is not a preference.

What 425°C Does: Derating, Creep and Galling

Derating is the first thing buyers get wrong

ASME B16.34 puts A182 F11 in Group 1.9, and the allowable stress falls as the metal gets hotter. The figures below are the working set. Treat anything above 400°C as approximate, and confirm against the current edition of B16.34 and the project line class before ordering. The F22 column is indicative, for alloy comparison only.

An 8 inch Class 2500 valve at 425°C is a 310 bar valve. Line classes written without a design temperature are the most common cause of an argument at the witnessed test, and it is a one-line fix at enquiry stage.

Creep sets the alloy, not the class

1.25Cr-0.5Mo is the entry alloy for high temperature steam and it is honest to about 450°C. Above that, wall thickness stops being set by yield and starts being set by creep rupture, and the alloy has to move: 2.25Cr-1Mo (A182 F22) to roughly 540°C, 9Cr-1Mo-V (A182 F91) to roughly 580 to 600°C. The datasheet this page replaces quoted an F11 body for 540 to 580°C steam, which is outside Group 1.9 and would not survive a competent design review. Either the material is upgraded with the temperature, or the temperature comes down.

Galling decides the overlay pairing

A metal seat is two hard surfaces sliding under load with no lubricant film. Cobalt-based Stellite 6 on both ball and seat is the classic pairing and it works, but cobalt on cobalt at high contact stress and low sliding speed is precisely the condition where galling begins, and a galled seat leaks on the next stroke. Three pairings are used in practice: Stellite 6 ball against a nickel-based (Colmonoy type) seat, tungsten carbide at 1,000 to 1,200 HV on the ball against a Stellite seat, or Stellite on both with at least 5 HRC difference engineered in. Whichever is offered, ask for the hardness of each surface and the finish, because Ra 0.1 µm on the ball is what makes a metal seat seal at all.

Metal Seated or Soft Seated: Where the Cut-Off Is

There is no polymer seat at 425°C. PTFE stops at about 200°C, reinforced PTFE at 230°C, PEEK at 250 to 260°C, and a carbon filled PEEK might survive a short excursion to 300°C. Above that the seat is metal, and the question changes from which polymer to how tight a metal seat can be made.

Rate A on a metal seat at 310 bar is not a specification, it is a negotiation at the witness test. Rate D is what a properly lapped metal seat delivers, Rate C is achievable when the differential is moderate and the ball is genuinely spherical, and both are accepted by API 6D for metal seated valves. Writing Rate A into a purchase order for this valve buys a rejected FAT and a month of correspondence.

Tightness is made at the lapping bench, not on the drawing. Ball sphericity within 0.02 mm, seat face flatness within one light band, and a ball and seat lapped as a pair and kept as a pair. That is also why a spare seat cannot simply be dropped into a used valve: the pair has to be re-lapped, and the drawing should say so. The metal seated ball valve page covers the lapping and overlay options in more detail.

Body Construction: Forged Three Piece, and the Wall Behind It

At Class 2500 the wall can be calculated rather than looked up. ASME B16.34 clause 6.1.2 gives t = 1.5 Pc d / (2S − 1.2 Pc). With Pc = 42.55 MPa, d = 200 mm and S = 103 MPa for F11 at 425°C, the minimum wall comes out at 82 mm. An 82 mm section in a casting is a quality risk; the same section in a forged block that is then bored is not. That single sentence explains why 8 inch Class 2500 bodies are forged.

The body joint is a bolted stack of forged flange, spiral wound gasket with a graphite filler and an austenitic winding, and A193 B16 studs with A194 grade 4 nuts. B16 is chosen over the more common B7 because at 425°C it is bolt relaxation, not tensile strength, that sets the preload surviving a thermal cycle. B7 loses a large part of its initial preload after a few hundred hours at that temperature, and a joint buried under 100 mm of lagging cannot be retightened.

Seat Design: SPE, DPE, DIB-1 and DIB-2

The seat is a ring in a carrier, pushed onto the ball by springs, with a piston area that either adds to or subtracts from that force. The direction the pressure is allowed to act decides whether the body cavity can relieve itself, and that is the most consequential line on the datasheet.

SPE self-relieves: cavity overpressure lifts the seat off the ball and vents upstream. DPE seals both ways, so the cavity is trapped, and trapped liquid that warms up has nowhere to go. In dry steam this barely matters. In condensate during a warm-up, or in liquid hydrogen letdown, it matters a great deal, and the cavity needs an external relief set at 1.33 times the rating or a drain connection to atmosphere. DIB-1 gives two barriers and is what double block and bleed actually requires; DIB-2 gives one barrier plus self-relief. The double block and bleed page goes further into which one a specification is really asking for.

Stem, Packing and Fugitive Emissions

Flexible graphite packing with a live loaded gland is the only sensible choice at 425°C. It has no plasticiser to boil out, it does not take a compression set at temperature, and it tolerates a stem that grows about 0.9 mm as it heats. PTFE packing is finished above 200°C, and graphite yarn reinforced with Inconel wire is used where the gland sees vibration.

The stem should be A182 F316 for this duty. 17-4PH is often offered and is a good stem material, but its useful continuous range stops around 300°C, so a 425°C datasheet listing 17-4PH is a flag worth raising. Above about 500°C, or where the stem is long and unsupported, Inconel 718 is the answer.

Fugitive emission performance is quoted to ISO 15848-1, and a certificate means nothing unless it names the temperature class, the cycle count and the tightness class. A CC1 result at 200°C says nothing about 425°C. Ask for the qualification temperature to match the design temperature; a supplier who cannot show that is quoting a test rather than a valve. The gland also needs a stem extension where the line is insulated: not for the packing, which is comfortable hot, but to keep the gearbox grease and the operator's hand inside their limits.

Fire Safety and the Document Package

A metal seated valve still needs API 607 or ISO 10497 qualification, because the parts that burn are the ones that are not metal: the body joint gasket, the packing, and any secondary seal behind the seat ring. The test burns the valve at 750 to 1,000°C for 30 minutes and then requires the seat to re-seal after cool-down. API 6FA is the API 6D equivalent and is what most refinery specifications call for.

The package that makes a valve acceptable on a real project is: EN 10204 3.1 certificates for body, closure, ball and stem, with heat numbers traceable to the NDE reports; the NDE reports themselves, ultrasonic on the forgings and magnetic particle on the body joint faces and the stem; seat and shell test records per serial number; the fire test certificate naming the size and class tested; and a NACE MR0175 hardness survey if the duty is sour. A file containing only a certificate of conformity is rejected on most steam and hydrogen projects.

How the Valve Is Tested Before It Ships

Every unit gets an ambient shell test, an ambient seat test, a torque measurement and a dimensional check. The pressures are not negotiable: shell at 1.5 times the 38°C rating is 638 bar, seat at 1.1 times is 468 bar.

Two details are worth watching at the witness. First, the seat test should be run with the valve closed and pressure applied to each side in turn where the seat design allows it, not once from one side. Second, torque should be measured with a calibrated load cell on the stem rather than inferred from the gearbox ratio, because break-to-open torque on a metal seat is the number that decides whether the actuator you bought will actually open the valve in January.

Hydrotest water needs chloride control on the austenitic trim: below 50 ppm, with immediate draining and drying, or the F316 ball and stem show pitting in the bore within weeks of the test. It is a small instruction that is regularly left out, and it is the most common reason a new valve arrives with rust spots.

Torque and Actuation

A metal seated 8 inch Class 2500 valve breaks open at roughly 8,000 to 14,000 N·m at full differential, cold. At 425°C the figure rises by 20 to 40 percent, and a valve that has sat closed under pressure for a year can double it. Actuators are sized on the hot break figure with a further 1.5 factor, which puts a spring return pneumatic unit in the 20,000 N·m class at the output shaft.

With a 300:1 gearbox a 12,000 N·m break torque needs about 130 N at the rim of a 600 mm handwheel, comfortably inside the 360 N API 6D allows. Size the gearbox on the break torque with a service factor rather than on the running torque, and specify a self-locking worm ratio so the valve stays where it is put under differential.

Where This Valve Is Used

Six duties account for most of the volume, and each one loads a different part of the design:

1. Main steam isolation and bypass in HRSG and industrial boilers: continuous temperature, high cycle count, and a specification that always names API 6D plus a fire test.

2. High pressure steam letdown stations: the highest differential in the plant, and the place where downstream erosion shows first.

3. Hydrogen rich process lines in hydrocrackers and hydrogen plants: here the limit is not B16.34 but the Nelson curve, and above about 200°C at this pressure 1.25Cr is outside it, so the body moves to F22 or F91.

4. Coal gasification and syngas: abrasive particles plus thermal cycling, so the overlay and the purging arrangement matter more than the class.

5. Ammonia and urea synthesis: high pressure with a corrosion duty that usually moves the wetted parts to 316L urea grade rather than F11.

6. Boiler blowdown and high pressure drain: cycled far more often than the process valve beside it, and seat ring fatigue is the failure mode.

Five Field Failures and What Actually Causes Them

1. Seat leaks after the third thermal cycle. The pair was lapped cold and not re-lapped after the first hot cycle. A metal pair beds in once; a seat replaced on site without lapping will leak within a month.

2. The valve will not open when hot. Break torque was sized cold, or the pocket clearance is too tight for the differential expansion. Fix it on the next order by specifying hot break torque, not ambient torque.

3. Body joint weeps under insulation. B7 bolting, or a gasket with a PTFE binder that extruded. The joint cannot be retightened through 100 mm of lagging, so the choice has to be right at build.

4. Stem scoring and packing leak. Gland not live loaded, or the stem was 17-4PH and has overaged. The scoring shows as a spiral, and it is the stem that has to be replaced, not the packing.

5. Gearbox strips a tooth. The gearbox was sized on running torque. On a metal seat break torque can be four times the running torque, and the first tooth to go is on the worm wheel.

How to Specify: Eight Numbers

A quotation for this valve can be accurate or it can be fast, and the difference is eight numbers. Everything else on the datasheet is a preference.

Two things to leave out of the enquiry. Do not write zero leakage: write the ISO 5208 rate. And do not copy the previous project's datasheet without re-reading the design temperature, because that is exactly how an F11 body ends up specified for 540°C steam.

Page notes: prepared by the J-VALVES application engineering group in Wenzhou, Zhejiang, and reviewed against ASME B16.34, API 6D, API 598, ISO 5208:2015 and ISO 15848-1. Ratings above 400°C are approximate and must be confirmed against the current edition and the project line class. Wall thickness, thrust and seat load figures are worked examples for an 8 inch full bore valve at the 38°C rating. Last reviewed October 2026.

8 Inch Class 2500 Ball Valve: Questions Buyers Ask

What pressure does an 8 inch Class 2500 ball valve hold at 425°C?

About 310 bar, not 425 bar. The 425.5 bar figure is the Class 2500 rating at 38°C. A182 F11 sits in ASME B16.34 Group 1.9, where the allowable stress falls with temperature, so at 425°C the rating is roughly 310 bar (4,500 psi). Any quotation that states 425 bar at 425°C has not applied the derating table.

Why is this valve trunnion mounted instead of floating?

Because of thrust. A full bore 8 inch ball at 42.55 MPa carries 1.34 MN, about 136 tonnes, and a floating design would have to pass that through the stem and the downstream seat. The trunnion carries it on two shafts and their bearings, so the seats only see spring load plus a small piston contribution, and the operating torque falls to a level a gear operator can handle.

Can a soft seated ball valve be used at 425°C?

No. PTFE is limited to about 200°C, reinforced PTFE to 230°C and PEEK to 250 to 260°C, with a short excursion to 300°C at most. At 425°C the seat is metal, and the design question becomes how tight a lapped metal pair can be made rather than which polymer to choose.

What leakage rate can a metal seated ball valve achieve?

ISO 5208 Rate D is realistic and is what a properly lapped pair delivers. Rate C is achievable when the differential is moderate and the ball is genuinely spherical. Rate A, which is normal for a soft seat, is not achievable at 310 bar on a metal seat, and specifying it guarantees a rejected factory test.

F11, F22 or F91: which body material for high temperature steam?

F11 (1.25Cr-0.5Mo) to about 450°C. F22 (2.25Cr-1Mo) to about 540°C. F91 (9Cr-1Mo-V) to about 580 to 600°C. Above 450°C creep rupture, not yield, sets the wall thickness, so the material has to be upgraded with the temperature rather than compensated for with a thicker wall.

What is the difference between SPE and DPE seats, and which do I need?

SPE self-relieves: cavity overpressure lifts the seat and vents upstream. DPE seals in both directions, so the cavity is trapped and needs an external relief set at 1.33 times the rating. For dry steam either works; for liquid that can warm up in a trapped cavity, or for double block and bleed, the choice decides whether the body is protected.

How is the valve tested, and at what pressure?

Shell test at 638 bar, which is 1.5 times the 38°C rating, and seat test at 468 bar, which is 1.1 times. Each side of the seat is tested separately where the design allows it, torque is measured with a load cell on the stem, and the gland and body joint are probed. Records are issued per serial number with the material certificates.

Can this valve be used in hydrogen service?

Only after checking the Nelson curve, not the pressure class. Above roughly 200°C at this pressure, 1.25Cr-0.5Mo is outside the acceptable envelope for hydrogen partial pressure, and the body has to move to 2.25Cr-1Mo or 9Cr-1Mo-V. Hydrogen service also needs hardness control and, usually, a specified maximum allowance for surface decarburisation.

What do you need to quote an 8 inch Class 2500 high temperature ball valve?

Size and bore, class and design temperature stated together, fluid and phase, seat arrangement and ISO 5208 leakage class, body and trim material including overlay, end connection and face to face, actuation and fail action, and the test and certification requirement. A line class sheet or datasheet shortens the review to a single exchange.