Seal installation problems



Seal installation should be “a piece of cake”. All you should have to do is use reasonable care and follow the manufacturer’s instructions. It doesn’t always work out that way, so in the next few paragraphs we will be looking at the installation problems you can have with the various designs of seals

Let’s start with the standard, non-cartridge seal. The following illustration shows a balanced, rotary version running against an “L” shaped stationary face that has been clamped to the face of the pump stuffing box. What sort of problems do we have with this design?

The seal is operating with the wrong compression causing it to either overheat or open up the lapped faces prematurely as the faces wear.

  • The pump “open impeller” clearance setting was adjusted to the volute after the seal was installed.
  • The pump shaft sleeve moved when the impeller was tightened to the shaft shoulder.
  • The installation dimension was measured incorrectly by the mechanic that was installing the seal:
    • The mechanic did not use the print, or could not read the print.
    • The pump has been modified and the print was not revised.
    • The mechanic took his measurement from the wrong location. He measured from the impeller locating shoulder instead of the stuffing box face.
    • The gasket thickness was altered.
    • The wrong measuring tool was used.
  • Temperature growth has changed the initial setting.
  • The pump is fitted with sleeve type bearings, and the axial movement is excessive.
  • The seal was rebuilt to a different axial dimension. Metal bellows seals can change their free length if the bellows assembly was over compressed and “bottomed out” during the assembly process.
  • The mechanic was not wearing his glasses and he could not read the correct measurement on his ruler.
  • The seal was installed using the old set screw mark as the reference. The reference was incorrect.
  • The sleeve or its gasket surface is damaged. This is causing leakage between the seal dynamic elastomer and the shaft sleeve, or between the sleeve and the shaft.
    • This is a common problem in all packing conversions. There are many products available that you can use to repair worn or damaged metal parts to insure a good gasket surface.
  • The seal faces are not flat.
    • The seal was accidentally dropped and no one dares tell the boss.
    • Poor packaging. The seal should be able to survive a 39-inch (one meter) drop.
    • Over tightening of the stationary face against a rough stuffing box in a rotating seal application can distort the lapped face.
    • The stationary face clamping areas are not equal and opposite, causing the faces to go out of flat.
    • The seal was lapped flat at ambient temperature and then put into cryogenic service where it went out of flat at the colder temperature.
  • A recirculation line is aimed at the seal faces. The high velocity recirculating fluid will interfere with the ability of the seal to follow whip, wobble, or run out. Be careful of this configuration when using metal bellows designs. The recirculated fluid may contain abrasives that will wear the thin metal plates.
  • A gasket or some type of fitting is protruding into the stuffing box and touching a flexible rotating seal component.
  • The stationary face is not centered to the shaft. Shaft radial movement can cause the seal faces to separate if the hard face is not wide enough. In some cases the stationary face was broken when the deflecting pump shaft contacted it.
  • The stationary seal face has been installed backwards and you are running on a non-lapped surface. Only one side of a hard face is usually lapped flat. Many manufacturers do a poor job of identifying the un-lapped face.
  • The elastomer (rubber part) was placed on a damaged portion of the shaft or sleeve.
  • The wrong lubricant was used on the elastomer and it is chemically attacking the elastomer. As an example: Any petroleum lubricant will damage an ethylene propylene (EPR) O ring causing it to “swell up”. This failure usually occurs within five days. The flexible seal components will “lock up” and prevent the lapped faces from staying together
  • The seal face was lubricated and then put into cryogenic service. The cold temperature will freeze the lubricant.
  • The application needs some type of environmental control and the connection has not been made. If the connection has been made, see if it is hooked up incorrectly or if the fluid supply has failed.
    • Environmental controls are used to control the temperature, pressure and cleanliness of the product in the stuffing box. These controls include flushing, quenching, heating, cooling, recirculating, venting etc. Make sure these controls are connected to the correct stuffing box, or seal gland port. Flushing, as an example, should come into the bottom of the stuffing box or seal gland.
  • The shaft/ sleeve tolerance or finish is not within specifications. Unbalanced seals are sensitive to the shaft/sleeve diameter and surface finish. Most balanced seals have an internal stepped sleeve, so they are not sensitive to small variances in shaft sleeve diameter.
  • Vertical pump applications have to be vented. Horizontal pump stuffing boxes can be drilled to insure good venting also.
  • The shaft sleeve was removed to make additional room in the stuffing box. The sleeve was positioning the impeller or the sleeve was providing corrosion resistance to the steel shaft.
  • The stationary face is not perpendicular to the shaft in a rotating seal application. This will cause excessive axial movement of the rotating components.
    • The use of a stuffing box facing tool can just about eliminate this problem.
    • Stationary seal designs and self-aligning designs can compensate for this problem.
  • The rotating face is not perpendicular to the shaft in a stationary seal application. This will create the same problem as mentioned above, except that the stationary face will do the moving. This is also the main reason you should not cartridge mount stationary seals unless they have some type of self-aligning feature.
  • The stationary face was not centered to the shaft.
    • This is one of the reasons we use cartridge seals because they have centering clips to solve this problem. Seal designs that utilize slotted glands should use some type of centering shims.
    • If you have not provided centering, the seal faces can separate if you are using “out of balance” or bent shafts.
    • Non-centering can cause a pumping action in cartridge seals that could cause a convection tank to run backwards and the barrier fluid to overheat.
  • The API (American Petroleum Institute) gland connections are hooked up incorrectly. Be sure that you understand the difference between flushing, quenching, drain and venting.
  • No suction recirculation line was hooked from the bottom of the stuffing box to the pump suction. Use a discharge recirculation line if you are pumping a fluid close to its vapor point.
  • The set screws can be a problem
    • The seal was set screwed to a hardened shaft. Seal set screws are manufactured from corrosion resistant materials. They are, therefore, softer than conventional setscrews and can vibrate loose. Most packing sleeves were hardened to resist packing wear.
    • The setscrews were over tightened and stripped.
    • The Allen wrench, used to tighten the setscrews has rounded corners and is not providing the proper torque.
    • In isolated cases the setscrews have loosened because the sleeve was made of too soft a material.
  • The gland flushing ports are machined off center. Shaft rotation will either help or hinder the flushing, depending upon their location. The flush rate changes with shaft rotation.
  • The shaft was machined down to accommodate a smaller diameter seal, or the shaft was machined for repair. In either case the shaft has been weakened and is now more liable to deflect with radial loading.
  • A severely cocked stationary will break the springs in the rotating unit because of the excessive back and forth movement. Excessive lug wear will also be visible.
  • The stuffing box often has product attached to its inside surface. This product can interfere with the free movement of the flexible seal components.
  • Horizontally split, double ended pumps have there own special problems when it comes to a seal installation:
    • In a conversion from packing to a mechanical seal it is common for the impeller to be positioned by sleeves that are not sealed at the impeller or inside diameter causing leakage between the sleeve and the shaft.
    • A new gasket between the pump halves can extrude into the stuffing box and touch the seal, stopping it from moving freely.
    • A gasket between the top and bottom halve must seal all the way to the stuffing box face or you will get a leak path on the side of the stuffing box.

Special installation problems with cartridge seals

  • The centering or installation clips were not removed prior to starting the pump.
  • The cartridge was installed by pushing on the gland. The friction on the sleeve static elastomer is causing the seal to over compress. Be sure to push the sleeve towards the gland before you tighten the set screws
  • The dual seal cartridge was installed by pushing on the gland. The friction on the sleeve static elastomer is causing the inner seal to over compress and the outboard seal to unload. Be sure to push the sleeve towards the gland before you tighten the set screws
  • The cartridge was installed on the stuffing box and then the stuffing box was placed over the shaft. This is causing the seal to be non-concentric with the shaft and could cause the shaft to hit the close fitting bushing in an API (American Petroleum Institute) type gland.
  • Check proper centering by using a spacer (normally a centering clip) to insure that the gland is centered to the shaft.
  • In some double seal cartridge applications, a pumping action will occur if the sleeve is not concentric to the gland. This can cause a convection tank to run backwards.

Special installation problems with rubber bellows seals

  • Silicon grease was used as a lubricant on the bellows and the rubber bellows is not sticking to the shaft. The proper lubricant is designed to cause the bellows to “swell up” and attach its self to the rotating shaft.
  • The shaft or sleeve is too smooth. It should be no better than 40 rms.
  • The carbon has been installed backwards. In most designs it will fit either way, so be sure you are running against the lapped side of the face.
  • The seal installation must be completed in fifteen to thirty minutes, or the rubber bellows will vulcanize to the shaft in the wrong location.
  • Be sure that none of the pump or seal dimensions have been altered. The seal must be installed at a fixed length that is not easily changed.
  • Most of the rubber bellows designs use a “buna N” rubber boot that is sensitive to ozone attack. Buna N rubber has a shelf life of only one year.
    • It should be noted that the normal mode of failure for this type of seal is for the rubber boot to experience a rupture or “blowout” causing a massive seal failure. It is important to use back up protection if you use this type of seal.

Special problems with metal bellows seals:

This picture describes a rotating version of a metal bellows seal used in the chemical industry. The bellows seal is sealed to the shaft by a static O-ring

  • If you are using the rotating version of the metal bellows seal, the close fitting anti-vibration lugs can hang up on the sleeve. The tolerance of the sleeve outside diameter is very critical with this type of seal.
  • In high temperature applications the rotating face holder is made from a low expansion material that has little to no corrosion resistance. If you lose cooling, the resultant shaft thermal expansion can engage the “damping lugs” in the low expansion steel face holder and “drag open” the seal faces. These “damping lugs” are not shown in the illustration. You can see them in the alphabetical section.
  • The seal hydraulic balance line shifts with stuffing box pressure fluctuations.
  • A pump discharge recirculation line can wear into the thin bellows section.
  • The chemical version of these seals is built from austenitic metals that change their spring load when the bellows is over compressed. It is real easy to over compress a seal during the installation process. There is a great temptation to push on the bellows face to feel the compression.
  • At elevated temperatures the carbon face can loosen in the metal holder. You should also watch out for glued in faces that can come loose if the product you are sealing attacks the glue.
  • Good designs will rotate the fluid with the seal to prevent a slurry from wearing the thin metal plates.
  • Be careful of stationary seals in high temperature applications. The flush connection will produce uneven cooling that can cause face distortion.
  • High temperature applications utilize heat-treated bellows materials. Be aware that this metal can anneal and loose its spring rate at temperatures as low as 300° F (150°C)
  • If the bellows seal is over-compressed during the assembly process the free length can change.

Here are some problems associated with split mechanical seals:

  • Some designs use glued elastomers (O-rings). This causes a hard spot in the elastomer that will not seat properly, causing continual leakage.
  • Many split seal designs cannot be used in an alternating pressure/vacuum application. This can be a problem when you have a mixer application to seal. Check to see if your design has this pressure/ vacuum capability.
  • Some elastomers are not available in split O-rings, so watch out for material availability.
  • Many outside designs move towards the product as the faces wear, making them ineffective in slurry applications. The above installation shows a rotating, split seal installed in a gland that is acting as an extension of the stuffing box.

Other problems that can occur shortly after any seal is installed:

  • The cooling jacket on the pump builds up a layer of calcium on the inside, interfering with the heat transfer. Condensate, steam, or heat transfer oil is a good substitute for hard well water, or dirty shop water.
  • Teflon® comes off of the metal parts. Sometimes a Teflon® coating is put on a seal part to lower hysteresis problems and to stop solids from attaching to the part. The product is penetrating the porous Teflon® coating and attacking the base material. In some cases you may have to remove the baked on coating with a fine emery paper.
  • A pressure drop at the seal faces and across the elastomer can cause some products to change state. This can cause the fluid to :
    • Solidify.
    • Become viscous.
    • Build a film on the faces.
    • Crystallize.
    • Blow out pieces of the elastomer
    • Cryogenic temperatures can freeze the elastomer or any lubricant placed on the seal faces.
  • At elevated temperatures, the differential expansion between the metal holder and the seal face can cause the seal face to become loose in its metal holder. Metal has an expansion rate approximately three times that of carbon and most other seal face materials.
  • The convection tank is running backwards or is not running at all. Please look a the following illustration:

  • Check to see if the seal gland is concentric with the shaft.
  • Be sure that the piping is connected to the correct gland ports.
  • Check the recommended height and distance specifications.
  • Be sure you have a proper liquid level in the tank.
  • If there is a pumping ring, make sure it is running with the correct rotation.