Why Does a Metal-Seated Ball Valve Leak More After Repeated Cycling? | Coating Wear, Lapping, Thermal Distortion

When a metal-seated ball valve starts leaking more after repeated cycling, the problem is usually at the ball-to-seat contact. The sealing surface may be scratched or worn, the coating may be damaged, the seat may no longer load the ball in the same way, or the ball may no longer return to exactly the same closed position. The leakage pattern often gives the first useful clue: did it rise slowly, appear suddenly, show up only when hot, happen mainly in one pressure direction, or begin right after maintenance?

Leakage Pattern Most Likely Areas to Check First
Slow increase over many cycles Coating wear, abrasion, erosion, loss of seat load
Sudden increase Deep scratch, trapped particle, coating spall, stuck seat
Leakage mainly when hot Thermal distortion, changing seat load, thermal interference
Leakage after overhaul Ball/seat matching, actuator travel, assembly contamination
Leakage mainly in one direction Pressure-assisted seat loading or one-sided damage
Leakage and torque both rise Galling, deposits, rough coating, excessive local load

In this article, leakage means internal seat leakage: fluid is getting through the closed valve between the ball and seat. Leakage from the stem packing, body joints, or flanges is a different problem.

Start With the Leakage Pattern

A cycle count by itself says very little. The same number of cycles can mean mild service in one plant and severe service in another.

Operating Frequency Cycles per Day Approx. Cycles per Year
1 cycle every 8 hours 3 1,095
1 cycle per hour 24 8,760
1 cycle every 10 minutes 144 52,560
1 cycle per minute 1,440 525,600

A valve cycling once per minute can exceed half a million cycles per year. Even a very small amount of wear on each cycle can add up quickly at that frequency.

But high cycle count does not automatically mean high damage. A valve completing 100,000 cycles in clean gas at stable temperature may still see less sealing damage than a valve completing 5,000 cycles in hot fluid carrying catalyst fines under high differential pressure.

When leakage starts to rise, four operating details deserve attention first:

  • differential pressure while the valve is opening and closing;
  • temperature and temperature change during each cycle;
  • solid-particle content and particle hardness;
  • how long the valve spends partly open.

These same points matter when specifying a valve for hot service. The high-temperature metal-seated ball valve guide covers them in more detail.

Check the Seal Band

The shutoff seal is not created by the whole ball surface. What matters most is the narrow circular band where the seat actually touches the ball.

On inspection, look at whether that band is:

  • continuous around the full seat;
  • in the expected position;
  • even in width;
  • free from grooves that cross from one side to the other;
  • free from isolated high-load areas.

A surface can look polished and still leak. Under magnification, metal surfaces are made up of microscopic high points and valleys, so the true contact area is smaller than the area that appears to be touching.[1]

That is why a small scratch can matter more than its appearance suggests. If the groove crosses the full sealing band, it can connect the high-pressure side with the low-pressure side and form a direct leak path.

When the valve is dismantled, compare the scratch direction with the contact band. A large mark outside the sealing area may matter less than a narrow groove running straight across it.

Check the Coating

A hard-coated ball should not be judged only by how much of the surface still looks shiny.

Look for:

  • local coating loss;
  • small pits;
  • microcracks;
  • flakes or chipped areas;
  • exposed base metal;
  • rough zones;
  • uneven polishing;
  • corrosion around damaged areas.

Many tungsten-carbide surfaces are cermets. In simple terms, hard carbide particles sit inside a metallic binder. Research on WC-Co materials shows that the cobalt binder helps hold carbide grains together during sliding contact.[2]

This matters in real service because the carbide does not have to become “soft” for the coating to fail. The binder can wear or corrode, carbide particles can pull out, or the coating can lose support from the base metal underneath.

How the coating was applied and finished matters too. NIST thermal-spray research shows that process conditions can change coating structure and properties such as porosity, hardness, and bonding.[3]

So two balls both described as “tungsten carbide coated” can perform differently if their coating thickness, porosity, bond quality, surface finish, substrate preparation, or grinding quality differs.

For a repair decision, the useful questions are practical:

  • How much coating remains?
  • Is the coating still bonded?
  • Has the base metal been exposed?
  • Is the damage shallow enough to remove without changing geometry?
  • Is the same damage likely to happen again in the same process?

Separate Abrasion, Galling, and Erosion

Abrasion, galling, and erosion can all damage a metal seat, but they do not point to the same root cause.

Damage Type What Usually Causes It What It Often Looks Like
Abrasion Hard particles trapped between ball and seat Directional scratches or grooves
Galling Loaded metal surfaces sticking and tearing during sliding Rough, smeared, transferred metal
Erosion High-velocity fluid or entrained particles Material loss near a jet or throttling area

Abrasive damage depends on more than how much solid material is present. Particle hardness, shape, size, and the load pressing the particle between the ball and seat all affect the result. A small amount of hard angular debris can be more damaging than a larger amount of soft powder.

If loose debris rolls or slides between the ball and seat, the process is often called three-body abrasion: the ball is one surface, the seat is the second, and the particle is the third body.

Galling comes from a different mechanism. NIST describes it as severe sliding wear involving surface roughening, plastic deformation, fracture, and material transfer.[4]

When galling develops, closing torque may rise because the surfaces become rougher and friction goes up.

Erosion often takes over after a leak path already exists. A narrow groove can focus the flow into a small area:

scratch → small leak → fast local flow → erosion → larger leak

The same kind of attack can happen while the valve is partly open. With high differential pressure across a small opening, flow velocity can be high enough to keep hitting the seat edge and nearby ball surface on every cycle.

Decide Whether Lapping Can Still Work

Lapping is useful only when the main geometry is still correct.

Grinding restores the main shape. Lapping refines the final contact.

Before re-lapping, check:

  • remaining coating thickness;
  • ball roundness;
  • seat concentricity;
  • contact-band position;
  • scratch depth;
  • whether the coating is cracked or lifting.

Many metal-seated valves use ball and seat surfaces finished as a matched set. The CARILO forged floating metal-seated valve, for example, uses paired and lapped metal sealing surfaces.

A wider contact pattern after service is not automatically a sign of better sealing.

A deliberately designed wide sealing surface can reduce high contact stress because the seat load and geometry were designed around that width. A band that became wider because of wear or excessive lapping is different. In that case, the original pressure distribution has changed.

Re-lapping is no longer a good option if the material removal needed to clean up the surface would:

  • make the coating too thin;
  • expose the base metal;
  • change the ball diameter beyond tolerance;
  • change the seat geometry;
  • move the contact band away from its intended position.

So a scratch may be physically removable while the part is still no longer repairable.

The metal-seated valve maintenance and leak-testing guide goes deeper into inspection and repair decisions.

Check Thermal Distortion

If leakage changes a lot with temperature, compare the valve in three conditions: cold, during heat-up, and after the temperature has stabilized.

Free thermal expansion can be estimated from:

change in length = original length × thermal expansion coefficient × temperature change

The U.S. Department of Energy uses this basic relationship when describing thermal expansion and thermal stress.[5]

Using a simple example with a 100 mm steel-like dimension and an illustrative coefficient of 12 × 10−6/°C:

Temperature Rise Free Growth of a 100 mm Dimension
100°C About 0.12 mm
200°C About 0.24 mm
300°C About 0.36 mm

The 300°C example is:

100 mm × 12 × 10−6/°C × 300°C ≈ 0.36 mm

This number is not the valve clearance. The ball, seat, body, stem, and support parts are all expanding and constraining each other at the same time.

What matters is the difference in movement between those parts and where the contact load ends up.

Fast heating adds another problem. The outside of the ball can heat before the inside. The process side of the seat can also become hotter than the section held by the body. That temporary temperature difference can slightly change the shape of the parts even though they are made from the same material.

Useful signs to check include:

  • leakage that appears only during heat-up;
  • higher closing torque when hot;
  • contact concentrated on only part of the seat;
  • polishing that is heavier on one side;
  • cold leakage that increased after repeated hot cycles.

If the valve is cycled while the contact is distorted, temporary thermal movement can leave permanent wear behind.

Check Seat Load and Pressure Direction

Pressure can create much larger mechanical forces than the valve size alone may suggest.

For pressure acting over a circular projected area 50 mm in diameter:

Pressure Approx. Force
10 bar 2.0 kN
50 bar 9.8 kN
100 bar 19.6 kN

At 100 bar, 19.6 kN is roughly the force produced by the weight of 2,000 kg under normal gravity.

This is only a pressure-force example. It is not the actual seat load of a specific 50 mm valve. Real seat load depends on effective piston area, ball size, springs, cavity pressure, seat geometry, and pressure direction.

That distinction matters when comparing floating and trunnion valves.

In a floating ball valve, differential pressure pushes the ball toward the downstream seat.

In many trunnion-mounted valves, the ball is supported while movable seats are pushed toward the ball by springs and pressure. The forged metal-seated trunnion valve shows this type of pressure-responsive seat arrangement.

If leakage changes when the pressure direction is reversed, inspect:

  • which seat is pressure energized;
  • upstream and downstream contact patterns;
  • cavity pressure;
  • spring condition;
  • whether one seat has more coating damage.

The ball valve cavity-pressure guide explains how trapped cavity pressure can change seat behavior.

Check Seat Movement, Travel, and Pipe Load

If the ball and seat surfaces do not show enough damage to explain the leak, the next question is whether the moving parts can still reach the position they were designed for.

For movable seats, inspect:

  • debris behind the seat ring;
  • broken or relaxed springs;
  • corrosion in the seat pocket;
  • carrier or guide wear;
  • restricted seat travel.

A seat face may look acceptable and still leak simply because the seat cannot move far enough toward the ball.

The same applies to the ball position. Check:

  • actuator travel stops;
  • coupling backlash;
  • stem keys or splines;
  • bearing clearance;
  • mounting-bracket movement;
  • incorrect setup after actuator replacement.

A CLOSED indicator only shows what the indicator says. It does not prove that the ball has reached the exact intended angle. The seat damage versus actuator travel guide helps separate internal seat damage from incomplete travel.

If the valve passed a bench test and only started leaking after installation, inspect the piping as well.

Using flange bolts to pull badly misaligned pipe into position can load the valve body. That external load can affect the alignment of a precision metal seat.

Use Torque as Supporting Evidence

Torque becomes useful when there is a baseline for comparison and the pressure and temperature are similar.

For example:

  • historical closing torque: 100 N·m;
  • later closing torque: 130 N·m;
  • increase: 30%.

A 30% increase is not a universal alarm limit. It simply tells you that the valve now needs materially more force to close under the same conditions.

Possible causes include:

  • galling;
  • deposits;
  • rough coating;
  • higher seat load;
  • thermal interference.

If leakage goes up while torque goes down, look instead at reduced preload, spring damage, worn seat mechanisms, or loss of contact.

And if torque is normal, do not assume the valve is tight. A narrow leakage path can pass fluid without increasing friction enough to move the torque reading very much.

The ball valve stem torque guide explains the difference between operating torque and allowable stem torque.

Record Data Before Disassembly

Before the valve is removed, capture the operating condition that produced the leak. Once the valve is on a bench, some of that evidence is gone.

Data to Record Why It Matters
Cycle count Shows whether the failure is early-life or progressive
Leakage trend Separates gradual wear from sudden damage
Upstream/downstream pressure Shows differential pressure and pressure direction
Temperature Shows whether the problem is heat-related
Test medium Allows leakage results to be compared correctly
Opening/closing torque Shows changes in friction or seat load
Recent maintenance Identifies assembly or actuator changes
Process solids or upset Identifies likely abrasive or erosion events

The point when leakage first increased can narrow the search quickly:

  • from the first test: check assembly, contamination, geometry, and travel;
  • slow increase: check wear, coating condition, erosion, and preload;
  • sudden increase: check for a particle, new scratch, spall, or stuck seat;
  • hot-only: check thermal distortion and hot seat load;
  • after overhaul: check matching, orientation, cleanliness, and actuator setup.

Measure the Parts

Visual inspection is useful, but it should not carry the whole repair decision on a critical valve.

Ball:

  • roundness;
  • runout;
  • coating thickness;
  • scratch location;
  • surface roughness where specified.

Seat:

  • contact-band width;
  • contact-band position;
  • concentricity;
  • surface damage;
  • free seat movement.

Coating:

  • remaining thickness;
  • cracking;
  • spalling;
  • exposed substrate;
  • local bond failure.

Drive system:

  • actual closed angle;
  • travel-stop position;
  • coupling play;
  • stem wear;
  • bearing clearance.

Do not polish, grind, or lap the surfaces before scratch direction, deposits, coating fragments, and the original contact pattern have been recorded.

Repair or Replace

Finding Typical Decision
Light polishing, correct geometry Inspect; controlled light lapping may be possible
Shallow scratch, enough coating remains Refinish within approved limits
Deep scratch, little coating remains Recoat or replace the trim
Coating spalling Do not treat as a simple polishing problem
Base metal exposed Restore the coating system or replace the part
Ball out of round Restore geometry if permitted; otherwise replace
Seat stuck by debris Repair seat movement before judging the sealing face
Incorrect closed position Correct actuator or drive travel first
Repeated thermal failure Review materials, clearances, loading, and operating sequence
Repeated abrasive failure Review coating, solids, valve speed, and process protection

Fixing the visible damage is only half the job. A re-lapped seat will leak again if hard debris keeps entering the seal, the seat still cannot move freely, or the valve keeps cycling while thermal distortion is pushing the contact into the wrong area.

Compare Leakage Tests Correctly

Leakage numbers only mean something when the test conditions are known.

Record:

  • test standard;
  • test medium;
  • test pressure;
  • pressure direction;
  • holding time;
  • temperature;
  • measurement method;
  • allowable leakage.

ISO 5208:2015 covers pressure testing of metallic industrial valves and verification of closure tightness. ISO reviewed and confirmed the current edition in 2025.[6]

API’s standards plan lists API Standard 598, Valve Inspection and Testing, 11th Edition, dated February 8, 2023.[7]

A gas leakage result should not be compared directly with a liquid leakage result unless the applicable procedure allows that comparison. Gas and liquid behave differently in small leak paths, and the test pressures and acceptance criteria may also differ.

Temperature condition matters too. A reading taken just after rapid heating may not match a reading taken after the ball, seats, and body have all reached a stable temperature.

Prevent the Same Failure

For high-cycle metal-seat service, keep the application data tied to the actual operating condition:

  • cycles per hour or per day;
  • normal and maximum temperature;
  • heat-up and cool-down rate;
  • differential pressure during movement;
  • particle type, size, hardness, and concentration;
  • corrosive process chemistry;
  • opening and closing speed;
  • time spent partly open;
  • required seat leakage test.

If a new valve starts leaking after only a small number of field cycles, first check weld slag, pipe scale, rust, metal chips, actuator setup, and piping load before blaming long-term coating wear.

During overhaul, mark the upstream and downstream seats, ball orientation, and stem position before removing matched parts.

If the service is still being selected, the soft-seat versus metal-seat comparison can help determine whether a metal sealing system is really needed for the actual temperature, solids, and cycling conditions.

Finally

Start with the pattern, not the assumption that the seat simply “wore out.” A valve cycling once per minute can exceed 525,000 cycles per year; a 100 mm steel-like dimension can show about 0.36 mm of free growth over a 300°C temperature rise; and 100 bar acting over a 50 mm projected diameter produces about 19.6 kN of force. These are scale examples, not valve limits. Record leakage, pressure direction, temperature, cycle count, and torque before disassembly. Then check the seal band, coating, seat movement, and actual closed position. Re-lap only when coating thickness and geometry remain inside the manufacturer’s repair limits.

Safety note: Pressure testing, disassembly, coating removal, dimensional acceptance, and return-to-service decisions should follow the valve manufacturer’s procedures, project requirements, plant isolation rules, and qualified engineering review.