Setting Consolidated 2755B PSVs: ASME I vs VIII Tolerance

The code follows the installation, not the test bench. A Consolidated 2755B mounted on an ASME Section I boiler is a Section I valve for its entire service life, and its set pressure tolerance is the Section I tolerance regardless of what medium the repair shop used to pre-set it. An air bench set is a pre-set, not a code set point. The set pressure of record is the one you verify on the operating unit with the lift-assist device, and that is the number that must land inside the Section I window.

Stop looking for an air-to-steam correlation number. There isn't a published one, and any factor you find in a spreadsheet is somebody's shop lore, not code-acceptable data.

Start Here: Confirm the Code Stamp and the Repair Certificate

Before you argue tolerance bands, settle two things.

1. Read the valve nameplate. A 2700-series safety valve on drum or superheater service carries the Section I capacity certification. If the nameplate says V, you are in Section I. If it says UV, you are in Section VIII. The service connection decides which valve was bought; the nameplate confirms what was certified.
2. Read the repair shop's certificate. Resetting and resealing a code safety valve is repair work under the National Board Inspection Code, and the shop needs a valid VR certificate with the right scope. A shop that cannot test on steam may still hold a VR certificate, but its quality program must define the alternate test medium and the correction it applies, and your Authorized Inspector has to accept it.

If the shop has no VR certificate, the bench work is a mechanical overhaul only. Nothing it produces is a code set point, and the field verification is doing all the work.

Read the Symptom Before You Touch the Adjusting Screw

Most "the valve came back wrong from the shop" calls are one of five things. Diagnose from behavior, not from the paperwork.

What you see	Likely cause	First check
Valve lifts well above the air bench set point on live steam	Temperature effect on spring rate and body/spindle expansion; bench set cold, service hot	OEM temperature correction for the spring code on the nameplate
Valve simmers or leaks at normal operating pressure after a shop set	Set point crowded too close to operating pressure, or seat damaged during bench popping	Operating gap versus set pressure; lap the seat and re-verify
Bench set repeats perfectly, field verification is 30–60 psi off	Air bench gives no sustained flow; the valve never reaches the same force balance it sees on steam	Field verify with the lift-assist device and adjust there
Valve chatters or flutters on the first live lift	Ring settings still in bench position, or inlet pressure loss	Ring positions against the OEM installed values; inlet piping pressure drop
Blowdown out of spec on the unit	Blowdown was never established on the bench because air cannot sustain flow	Ring adjustment on the operating unit, per OEM procedure

A repeatable air set that misses on steam is not a shop error. That is the physics of the test, and it is the reason Section I service leans on live-steam or lift-assist verification.

Understand Why an Air Set Drifts on Steam

Three mechanisms move the set point between the bench and the boiler, and only one of them is about the medium.

• Temperature. This is the big one. A helical compression spring loses stiffness as it heats, and the spindle, disc holder and body grow. Set cold on an ambient bench and the same valve lifts at a different pressure when the bonnet is at steam temperature. This is exactly why the 2700 series uses an open bonnet on steam service: the spring runs at or near steam temperature in service, so testing on live steam removes the correction entirely. Cold-bench testing puts it right back in.
• No sustained flow. An air bench has a finite accumulator. At set point the valve cracks, the bench pressure collapses, and the valve slams shut. You get a pop and a reseat with no flow behind it, so you never establish lift, blowdown, or stable reseat pressure. You also hammer the seat every time you pop it.
• Force balance after lift. Once the disc moves, the medium matters. Steam expanding across the huddling chamber and the ring geometry produces the lift force that holds the valve open; air on a small accumulator does not. That is a post-lift phenomenon, which is why it changes blowdown, stability and reseat far more than it changes the initial crack point.

The pressure at which the disc first moves is set by the spring force divided by the seat area. Medium does not change that. Temperature does, and flow behavior after lift does. Keep those three effects separate when you argue with a test report.

Apply the Right Tolerance Band

Section I and Section VIII use different set pressure tolerance rules, and at boiler pressures Section I is the tighter one. Confirm the exact wording against the edition your jurisdiction has adopted before you sign anything.

≤ 70 psi	±2 psi	±2 psi
71–300 psi	±3%	±3%
301–1000 psi	±10 psi	±3%
> 1000 psi	±1%	±3%

Work an example. At 600 psi set, Section I allows ±10 psi, a 20 psi window. Section VIII allows ±3%, or ±18 psi, a 36 psi window. The Section VIII band is nearly twice as wide at that pressure. So do not shop for the looser band because the shop only has air.

The practical rule for the bench: do not use any tolerance band as your target. Aim the bench set at the nominal cold differential test pressure and record the actual number. You want to arrive at the unit as close to nominal as the bench will let you, because every psi of bench error is a psi you have to chase with the lift-assist device on a hot, pressurized boiler.

Run the Bench Set on Air

Treat this as a pre-set that gets the valve mechanically sound and inside shouting distance of set point.

1. Record the as-found lift pressure on air before you disassemble anything. That number, compared to the last field verification, tells you whether the valve drifted in service or the shop moved it last time.
2. Overhaul: lap the seat and disc, inspect the spindle, guide, disc holder and bushings, replace the spring if it shows set or corrosion, and confirm the spring code matches the nameplate range.
3. Get the cold differential test pressure from the OEM. For elevated-temperature service, the CDTP is not the same as the stamped set pressure. The correction is spring-code and temperature specific and it lives in the manufacturer's maintenance manual, not in a generic table. Do not derive it yourself.
4. Set the ring positions to the manufacturer's bench values, not the installed values. Bench rings are positioned to soften the pop and limit the slam-shut on a bench with no sustained flow.
5. Bring pressure up slowly and hold well below set until the valve stabilizes, then approach set point at a controlled rate. Pop the minimum number of times needed for a repeatable reading. Every extra pop is seat damage you will pay for later as simmer.
6. Perform a seat tightness test at the manufacturer's specified percentage of set pressure. A valve that will not hold on the bench will not hold on the boiler.
7. Reset the rings to the installed positions, wire-seal the adjusting screw and rings, and record the bench set pressure, spring code, ring positions and test medium on the repair form.

Verify Set Pressure on the Operating Unit

This is where the Section I set point is actually established. The lift-assist device applies a measured auxiliary force to the spindle while the system holds a stable pressure below set, and the set pressure is computed from that force and the known effective area.

1. Bring the boiler to a stable pressure with adequate margin below the stamped set point. Instability in system pressure walks straight into your calculated result.
2. Verify the device calibration is current and that the load cell and the effective seat area entered into the instrument match the valve.
3. Run the lift, read the calculated set pressure, and repeat until you get consistent results. Discard the first pull if the valve has been sitting; a first-lift stiction reading runs high.
4. Compare against the Section I band. At 301–1000 psi that is ±10 psi around the stamped value.
5. If you are outside the band, adjust the compression screw, reseal, and re-verify. Log the as-found and as-left values.
6. Confirm the valve reseats and holds after the last lift. A tight valve that goes into simmer after adjustment means the seat is damaged, not that the set point is wrong.

Because the bonnet is open and the valve is at operating temperature during this test, no temperature correction applies to the field number. That is the whole point of verifying hot.

Stop Chasing an Air-to-Steam Correlation Factor

There is no published, generally applicable correlation between an air set point and a steam set point. What OEMs publish are two different things, and they get confused constantly:

• Temperature corrections (cold differential test pressure). Real, published, spring- and temperature-specific, and available from the manufacturer. This is what most "correlation factor" spreadsheets actually contain.
• Capacity correction factors between one compressible gas and another, or between liquids. Real, published, and used for sizing. They do not translate a steam set point into an air set point.

What does not exist is a general factor that converts "popped at X psi on air" into "will pop at Y psi on steam." Manufacturers give recommendations from their own test experience; those are manufacturer-specific and are recommendations, not code data. If you build a correction from your own shop history, it belongs in a documented quality program with a data trail, and your Authorized Inspector or jurisdiction has to accept it.

Nuclear work raises the bar. ASME OM requires safety valves to be tested at system medium and temperature, or that correlation factors be supplied and justified. Regulators do query those factors. When you pull the OEM manuals, expect most to address temperature correlation and few to address test-medium correlation, because the medium correlation is the one that has no technical basis at the crack point. Note also that Section III applies only to primary nuclear systems; non-nuclear systems inside a plant legitimately carry Section I or Section VIII valves, so verify which code stamp is on the specific valve rather than assuming plant-wide.

Avoid the Failures That Repeat on This Class of Valve

• Popping the valve repeatedly on a small air accumulator. The single most common way a shop turns a good valve into a leaker. Minimize pops and use the bench ring settings.
• Shipping the valve with bench rings still installed. The valve will lift near the right pressure and then flutter or chatter on the unit. Verify ring positions and lock screws before the valve leaves the shop, and again before installation.
• Using the Section VIII band because the numbers are friendlier. A Section I installation is judged against Section I. Test medium does not change jurisdiction.
• Guessing the temperature correction. Wrong direction and the valve simmers at operating pressure; wrong magnitude and it lifts high. Pull the number from the manual for the spring code on the nameplate.
• Skipping seat tightness after the last field adjustment. Every adjustment lands the disc on the seat again. Confirm it holds.
• Ignoring inlet piping. Excessive inlet pressure loss produces chatter that looks exactly like a bad set point. Check the inlet run before you blame the shop.

Escalate to the manufacturer when the OEM manual gives no test-medium or temperature guidance for your spring code, when field verification and bench set disagree by more than the Section I band after two adjustment cycles, or when the valve will not hold seat tightness after a correct lap. Those are the cases where a factory recommendation is the only defensible basis for what you do next. If your jurisdiction or Authorized Inspector has questioned a correlation factor, get the written OEM position before you test again, not after.

Frequently Asked Questions