Mechanical seal failure

Why do most Mechanical Seals fail
A mechanical seal can either wear out, or fail. To determine which one your seals are doing, look at the wearable face. In most instances this will be the face manufactured from some grade of carbon/ graphite.
Since the seal face is the only sacrificial part of the mechanical seal, a worn out seal is identified as one that has no carbon nose piece left at the time it started to leak. A failed seal is identified by the fact that it has substantial carbon remaining at the time it started to leak.

The above illustrations show the difference between a worn out and a new mechanical seal.
Most consumers experience seal failure rates in excess of 85%, and for the most part these seal failures are easily correctable. Seal failures fall into only two broad categories, either the seal faces opened, or one of the seal components was damaged by contact, heat or corrosion. Whenever we try to troubleshoot any mechanical seal it's wise to remember that only three things are visible to a troubleshooter:

• Evidence of rubbing.
• Evidence of damage including corrosion, physical damage, or discoloration of one of the seal component materials. Most mechanical seals are constructed of three materials:
  • Metal parts
  • A face combination
  • Some rubber like parts (called elastomers)
• The product is attaching to a sliding component causing sticking, or coating on the face causing face separation.

Here are some reasons why a mechanical seal face would open:
The dynamic elastomer is not free to slide or move on the rotating shaft or sleeve.

• The shaft is oversize. A tolerance of + 0.000 - 0.002 inches (+ 0,00 - 0,05 mm) would be typical.
• The shaft finish is too rough. Most seal companies want at least a 32 R.M.S. (0,8 micro meters) surface finish in the area of the dynamic (sliding) elastomer.
• The fluid we are pumping is causing the elastomer to stick to the shaft. The dynamic O-ring can generate a lot of heat if the shaft is not perpendicular to the face of the stuffing box. The rapid movement of the elastomer will generate localized heat causing the following to occur at a faster rate:
  • The product is solidifying (glue and paint will do this)
  • It is crystallizing (sugar syrup and caustic are good examples)
  • It is building a coating on the shaft (petroleum products will form varnish or coke at elevated temperatures, or hard water will form a layer of calcium. etc.)
• Dirt or solids are restricting the elastomer from moving.
• Chemicals added to treat water or impurities in the water can collect on the seal sliding surfaces
• A chemical has attacked the elastomer causing it to swell up and restrict the movement of the seal. In some instances a swollen elastomer has been known to open seal faces while the pump was not running in a standby mode.
• The shaft or sleeve has been hardened and the set screws have slipped. Many sleeves were hard coated to resist packing wear. Stock rooms are full of these sleeves.
• The seal has lost its compression.
  • It was installed with the wrong compression.
  • The impeller was adjusted after the seal was attached to the shaft. This is a very common problem with A.N.S.I. or other back pull out pumps.
  • A temperature change has altered the location of the seal. Remember that each inch of stainless steel shaft will grow one thousandth of an inch for each one hundred degree Fahrenheit rise in temperature or 0.001"/1"/100°F . Metric grows 0,001 mm/1 mm of shaft for each 50°C rise in temperature.
  • The open impeller was adjusted to compensate for normal wear. Typical pump specifications allow the impeller and the casing each to wear as much as 0.125 inch (3 mm) and still be adjusted back to the correct pump efficiency. This is important when you realize that the average mechanical seal has a carbon nose that extends only 0.125 inch (3 mm).
• The springs, spring or bellows are not operating properly.
  • A single spring has been installed backwards allowing the faces to stay in contact while the shaft or sleeve rotates within the dynamic elastomer or end fitting.
  • Excessive misalignment is causing rapid flexing of the spring or bellows causing them to fatigue.
  • The drive lugs have failed and the multiple springs are twisted in their holder.
  • The product has clogged the springs.
  • Many times the outside springs of a dual seal have been painted either at the pump company or as part of a normal maintenance routine.

Something is restricting the free movement of the seal.

• The product is viscous. Remember that some products become more viscous with agitation. These products are called dilatants (cream becomes butter with agitation)
• A recirculation line from the discharge of the pump is aimed at the seal and interfering with its movement.
• A foreign object is in the stuffing box.
• A protruding gasket is touching the movable part of the seal

The shaft is being displaced causing the seal to hit something as it rotates, or to cause the rotating face to run off of the stationary face.

• The pump is operating off of its best efficiency point (B.E.P.) causing the shaft to bend.
• The rotating assembly is out of dynamic balance.
• The shaft is bent.
• There is misalignment between the motor and the pump.
• Pipe strain is twisting the pump stuffing box.
• Heat causes expansion and that always opens the possibility for rubbing or wear.
• Cavitation, slip stick, harmonic vibration, bad bearings or some other form of vibration is causing excessive movement of the shaft.
• The shaft sleeve is not concentric with the shaft causing it to run "off center".
• The pump has been designed with sleeve or babbitted bearings and shaft movement is excessive.

The seal face is being distorted by either temperature or pressure.

• Lapped hard faces are especially sensitive to either changes in temperature or pressure excursions.

The product is vaporizing between the seal faces causing the faces to blow apart.

• If boiler feed water vaporizes it leaves behind all of the chemicals that were added to the water to prevent hardness, to control PH, soften boiler scale etc....
• In cryogenic (cold) applications the vaporizing fluid can freeze any lubricant that might have been placed on the seal faces. This frozen lubricant can damage the carbon/ graphite seal face.

An environmental control has failed. There are many types used with Mechanical Seals, here are a few of the common environmental controls:

• Flushing is used for cooling and to wash away solids.
• Quenching is used for temperature control and vapor removal.
• Barrier fluids are used to keep air away from a fluid and to provide temperature control.
• Cooling/ heating jackets are used to keep products in a liquid state and at the proper temperature.
• A suction recirculation line is installed from the bottom of the stuffing box to the suction side of the pump. This is done to remove stuffing box solids in the pumping fluid and to provide cooling to the seal components.
• A line can be installed from the discharge of the pump to the stuffing box to increase stuffing box pressure whenever you pump a fluid close to its vapor point. It is also wise to install a carbon restriction bushing in the bottom of the stuffing box with a clearance of approximately 0.005" to 0.007" (0,13 mm to 0,018 mm) on the inside diameter.
• Dual seals can be installed to prevent a pressure drop across the inside seal face and to control the temperature at the seal face.

Unbalanced seals can open their lapped faces in vacuum applications.

• Those pumps that run under vacuum include: condensate pumps, heater drain pumps, pumps that lift liquid and any pump that takes its suction from a condenser or evaporator. Remember to use O-ring elastomers in vacuum applications as this shape elastomer will seal either vacuum or pressure.
• The product has built up on one of the seal faces causing the faces to separate. This is a common problem with petroleum products or any product that can build a film on a surface. Since this coating is not dense enough to provide good sealing, it can cause the faces to leak at shutdown.

When a seal face opens it allows solids to penetrate between the lapped surfaces. The solids imbed themselves into the softer carbon/graphite face causing it to act like a grinding wheel. This grinding action will cause severe wear in the hard face. It should be noted that seal face opening accounts for the largest majority of mechanical seal failures.
The second major cause of seal failure is when one of the seal components is attacked by the sealing fluid or a chemical being used to clean or flush the lines. Chemical attack is easy to see:

• The Carbon will appear to have a sponge like appearance
• Plated materials will have their hard coating peel off when the base material is attacked. This same thing happens when you allow rust to penetrate behind automobile paint and you then notice that the paint is peeling off in sheets.
• The elastomer will usually swell up and get soft. When an elastomer shrinks and gets hard it is almost always evidence of excessive heat. Prior to failure caused by excessive heat, most elastomers will take a compression set (the round O-Ring becomes square)
• Metal components will develop pits and an overall dull appearance. The color of the metal is often an indication of the amount of heat it was subjected to:

FAHRENHEIT	COLOR OF THE METAL	CENTIGRADE
700 - 800	Straw Yellow	370 - 425
900 - 1000	Brown	480 - 540
1100 - 1200	Blue	600 -650
> 1200	Black	> 650

Here are a few things to consider when you suspect corrosion is the problem :
The corrosion rate of almost all chemicals doubles with each 18 degree Fahrenheit (10 C.) rise in temperature.

• Be sure to vent vertical pumps. Air trapped in the stuffing box is a good insulator.
• See if the operator is running the pump with a restricted discharge. In addition to deflecting the shaft it can cause a severe heat rise in the pump. The control valve may be stuck in the throttled position.
• Try to use a recirculating line from the bottom of the stuffing box to the suction side of the pump. This is practical in almost any application other than when we are pumping a product close to its vapor point and there would be a danger of vaporizing the product in the stuffing box.
• When ever possible bore out the packing stuffing box or install a large seal chamber in place of the packing stuffing box. This extra room will allow centrifugal force to centrifuge and clean the fluid in the seal chamber as well as provide extra cooling in the seal area.
• It is normal to dead end the fluid in the stuffing box when a cooling or heating jacket is being used. If a recirculation line is installed in the stuffing box along with the cooling jacket, the jacket will become inoperative because the circulating hot fluid will not be in the stuffing box long enough to be cooled by the jacket.
• Be sure to check that the cooling jacket is functioning. A layer of calcium inside the jacket, can just about stop heat transfer. If the water is too hard in your area, consider condensate as an alternative cooling fluid.
• More than one stuffing box jacket has frozen in cold weather, be sure to use non freezing cooling fluids at lower temperatures
• If a convection tank is being used with dual seals make sure it is operating. Every design has limits, make sure you are not exceeding them. Also check that the fluid is flowing from the top of the stuffing box to the convection tank and returning to the bottom of the stuffing box. I have seen many of these applications running backwards.
• Use only balanced seals. They generate less heat than unbalanced seals.
• If there is a bypass line installed from the discharge piping to the suction side of the pump, it may be heating up the incoming fluid.
• Check to see if the cooling jacket has been isolated and drained. This often occurs when a metal bellows seal is used in hot oil applications. An empty cooling jacket will act as an insulation to the stuffing box fluid.
• Remember that the cooling jacket is also there to cool down the shaft and protect the bearings. Do not disconnect it.

When you look for corrosion be sure to check out any cleaners or solvents that are used to flush out the system or clean the lines. Many grades of Viton® can be attacked by cleaning the lines with steam or caustic. It is important to identify all of the materials used in the seal components.

• Carbon fillers can be attacked by heat and chemicals
• Plated materials can crack due to differential expansion.
• Stainless Steel springs can break due to Chloride Stress Corrosion.
• Hardened set screws can corrode and vibrate loose.
• Some elastomers can be attacked by steam. Be careful of using petroleum grease on elastomers as some compounds can be attacked by any petroleum product.

Some hard coatings have very little flexibility and will crack with a small differential temperature. Be careful of tungsten carbide with a cobalt binder; nickel binder would be a much better choice.