Preventing seal failure

Preventing premature seal failure 10-4

Here is a three question quiz:

Question: How do you prevent premature mechanical seal failure?

Answer: Find out what is causing the seals to fail and prevent it from happening.

Question: How long should a mechanical seal last?

Answer: Until the sacrificial carbon seal face wears away.

Question: How often do seals wear out?

Answer : Less than ten percent of the time.

If you understand that simple logic you’ll be happy to lean that seals fail for only two reasons:

  • The lapped faces opened.
  • One of the seal components becomes damaged.

Don’t make it too complicated. If you inspect a failed seal and see no evidence of a component damage, then the seal faces must have opened, because seals fail for only two reasons.

Here are the most common reasons that lapped faces open prematurely:

  • The set screws slipped on a hardened shaft sleeve.
  • The dynamic rubber part (the rubber part that moves) stuck to the shaft or sleeve.
  • The spring load on the wedge, v-rings or u-cup used with most pusher type seals was too high. The Crane #9, Durametallic ROTT and Borg Warner “U” are typical of these designs.
  • The shaft or sleeve diameter was too large. +0.000 inches to -0.002 inches (+0.00 mm to -0,05 mm) is a good tolerance for mechanical seals.
  • The sleeve finish was too rough. You want a finish of 32 rms (0,8 microns) or better, unless you are using rubber bellows designs that call for a finish of no better than 40 RMS..
  • The O-ring groove was out of tolerance. There is too much interference on the shaft preventing the O-ring from flexing and rolling.
  • The elastomer swelled up because it was not compatible with the pumping fluid, or a cleaner that was circulated through the lines. In some designs this expansion of the rubber part can force the lapped face out of its holder.
  • Too much axial movement of the shaft:
  • The seal was installed on the shaft and then the semi-open impeller was adjusted to the pump volute for the inital clearance setting, or to compensate for normal impeller wear.
  • Thermal growth of the shaft. Every inch of shaft will grow 0.001″ for each 100°°F (0.001 mm/mm/ °C of temperature increase.
  • Sleeve bearings allow excessive axial movement. Some centrifugal pump designs utilize this type of bearing as a thrust bearing.
  • Too much shaft vibration and no vibration damping. This is a big problem with the metal bellows seals we find in high temperature applications. Vibration damping interferes with the frequency of the vibration. The O-ring found in many seal designs is a natural vibration damper.
  • Harmonic vibration is a major source. The seal is vibrating in harmony with another piece of equipment
  • Cavitation is very common in many applications. Remember that there are five types of cavitation and all of them can cause vibration problems.
  • “Slip stick” vibration can occur between seal faces if the pumping fluid is not a lubricant. Hot water is a good example of a non lubricating liquid. Most gases and dry solids are obvious non-lubricants.
  • Hitting a critical speed is another cause of vibration. Pumps that have this problem are said to have flexible rather than rigid shafts.
  • Installation errors:
    • Not enough load on the lapped seal faces. Do not be tempted to increase the designed load because the additional heat generated can be a cause of seal face damage.
    • The seal was installed at a wrong inital setting.
    • The sleeve moved when the impeller was tightened. Measurements should be taken after the sleeve to shaft gasket is compressed.
    • The mechanic misread the installation print.
    • The mechanic used the old set screw marks as a guide and their location was not correct No print was available at the installation site.
    • The impeller was adjusted after the seal was attached to the shaft. Duriron is the exception because its impeller adjusts towards the back plate causing the seal to over compress.
    • Shaft or sleeve thermal growth. In most seal designs this will unload the seal faces. The seal must be attached to the shaft after all thermal growth has occurred. You are going to need a cartridge seal to do this. Outside mounted, non metallic seals can be an exception. They will over compress with thermal growth.
    • Rotating type mechanical seals need the stationary face installed square to the shaft to prevent excessive axial movement. Stationary designs need the rotating face to be installed square to the rotating shaft. This is much easier to do as long as the seal face is not set screwed to the shaft. It should be butted up against a square shoulder or some other type of “squareness” must be provided.
    • The wrong lubricant was used on the dynamic o-ring causing it to swell up and lock the seal to the shaft.
  • Reversing stuffing box pressure can cause most unbalanced seal designs to open.
  • High shaft speed will cause centrifugal force to drive the rotating face square to the shaft, opening the lapped faces. 5000 fpm (25 meters/sec) is just about the limit for rotating seal designs.
  • The shaft is fretted (grooved from the dynamic elastomer) causing the moveable face to hang up as it tries to compensate for wear.
  • The lapped faces are not flat.
    • The faces never were lapped flat
    • The lapped face was installed backwards. You are running on the non-lapped side.
    • High stuffing box pressure can distort a lapped face.
    • Thermal distortion can distort a seal face.
    • Seal used in cryogenic service (very cold) must be lapped at cryogenic temperatures.
  • The product changed state and is restricting the movement of the seal.
    • The product can crystallize if you change the temperature in the stuffing box.
    • A temperature change or agitation can cause a product to become viscous.
    • Products can solidify with a change in temperature, pressure or agitation.
    • A change in temperature or pressure can cause a product to build a film on the seal sliding surfaces.
    • The product vaporized between the faces and blew them open. This can happen with an increase in temperature or a decrease in stuffing box pressure.
    • Solids clogged the springs or some other part of the seal, restricting seal movement.
    • Solids outboard the seal can restrict axial movement as the seal moves to compensate for carbon wear.
      • Ice can form when some products vaporize, or cold weather can freeze moisture in the air.
      • Crystalls and solids can form outboard because of seal leakage or dirty quench fluid.
      • If you are using a gland quench connection, the quenching fluid must be clean or it will deposit contaminants outboard the seal.
  • Discharge recirculation lines aimed at the moving seal parts can restrict their movement.
  • The seal face hung up in the fretted groove that we find so common in most original equipment seal applications.

The easiest way to tell that you are having seal face opening problems is to inspect the hard face for evidence of wear. Common sense dictates that carbon cannot wear a hard seal face.

If the faces open, it will allow solids to penetrate between the lapped faces and then these solids will embed into the softer carbon when the faces close. The contaminated carbon will then act as a grinding surface making wear marks in the harder face.



  • On February 18, 2018