API Standard, its flaws

What is the problem with this API specification as a potential standard for the Chemical Process Industry? There are a lot of things I do not like about it in its present form. If combining with the CPI means a complete re-writing of the API specification that will be fine, depending upon the final result.

  • 2.1.1 Some seal designs do not lend themselves to a cartridge design. Split seals as an example. You could mount a split seal on a split cartridge, but that would be “over kill” in most cases.
  • 2.1.2 I do not like the definition of pusher seal in this standard. The term “pusher seal” is emotionally charged and misleading. It is used to describe a reliable O-ring seal in the same category as spring loaded Teflon® wedges, or chevrons, and non-elastomer “U” cup designs. The implication is that the “non-pusher” metal bellows seal is a better choice. The fact is that O-ring seals are usually a better choice because of their ability to flex and roll and the O-ring provides a built in vibration damper that eliminates the need for letting a bellows metal face holder bounce off the shaft or sleeve.
  • 2.1.5 The dual seal specification recognizes only tandem or series mounted rotating seals. It ignores concentric and “face to face” designs that make sense in some applications where space is not available for tandem configurations. Over the years the API has failed to recognize that there are four ways to install dual seals in a pump. They have played with the terminology over the years but have never got it simplified. It should be:
    • Face to face
    • Tandem or series
    • Back to back
    • Concentric, or one inside of the other.
  • 2.1.6 The specification calls for the inner seal of a dual seal to be either balanced or reverse balanced depending upon whether high pressure barrier fluid or lower pressure buffer fluid is circulated between the dual seals. It totally ignores two way balance of the inner seal that would allow the consumer his choice between barrier or buffer fluid.
  • 2.1.6 The specification call for the dual seals to be mounted in series (tandem), but almost all gas dual seals supplied to refineries to date have been supplied in the “back to back” configuration which is the worst possible installation method for slurry and abrasive service, because it places the slurry at the inner seal inside diameter.
  • 2.1.7 The specification approves rotating seals only and recommends stationary seals for speeds above 5000 fpm (25 m/sec). The fact is that stationary seals are almost always a better choice for leak free and the more severe fugitive emission sealing.
  • 2.1.7 Stationary seals (the spring or springs do not rotate with the shaft) can be cartridge mounted if you take precautions to insure that the rotating face stays square to the shaft when the cartridge sleeve is set screwed or tightened to the shaft. It is not an easy problem to solve, but there are several solutions to the problem. Please see “stationary cartridge seals”.
  • 2.2.6 The specification calls for O-ring grooves with a larger groove dimension than normally used to accommodate perfluoroelastomer O-rings.
  • Slotted glands make sense for all pumps not just horizontally split pumps.
  • The specification assumes all pump manufacturers have provided a machined diameter concentric to the pump shaft so that the seal gland can be machined to register on an inside or outside diameter. The fact is that most pumps were manufactured for packing and do not have these concentric machined surfaces available to the seal manufacturer. In the CPI industry, shaft centering makes the most sense.
  • 2.3.10 Maintaining a seal chamber 50 psi (3.5) bar above vapor pressure does not make any sense in the majority of balanced seal applications.
  • 2.4.1 The specification calls for a shaft sleeve and allows the manufacturer to reduce the diameter of the solid shaft to accommodate the sleeve. This increasing of the pump shaft L3/D4 adversely affects the pump and seal performance.
  • 2.4.1 The specification calls for sealing the sleeve on one end, but fails to specify the impeller end except in the case of O-ring seals. If the seal is on the outboard end, the space between the sleeve and shaft can fill with solids and hamper the removal of the sleeve. This can be a major concern in hot oil type applications where “coking” is always a problem.
  • 2.4.3 A shoulder to locate the rotating element will not allow a re-positioning of the seal if an open impeller has to be adjusted. API (American Petroleum Institute) pumps are normally closed impeller designs, but we are talking about the possibility of combining standards with the ANSI design.
  • 2.4.9 A shaft to sleeve diametral clearance of 0.001 inch to 0.003 inch is not practical. You will never be able to remove the sleeve once some solids get between the sleeve and shaft, and they will get there!
  • The majority of hard set screws are not corrosion resistant. If the set-screws are located out of the stuffing box this could be OK, but many designs have the set-screws positioned in the sealing fluid.
  • 2.6.1 The standard seal is equipped with multiple springs, but the standard does not specify the springs must be located outside the fluid. If located in the fluid they can easily clog with solids.
  • 3.2.3 Reaction bonded silicon carbide is specified as the standard hard face even though it is sensitive to caustic and other high pH chemicals frequently used to clean lines and systems. In most cases alpha sintered would be a much better choice.
  • 4.2.1 The term “flush” is misleading. Over the years the API has failed to recognize the differences in bringing liquid to the pump stuffing box area and lumped them all under the common term “Flush”. There is better terminology:
    • Discharge recirculation connects the discharge of the pump to the stuffing box to raise stuffing box pressure.
    • Suction recirculation connects the bottom of the stuffing box to the suction side of the pump usually allowing clean fluid to circulate from behind the impeller into the stuffing box.
    • Barrier fluid describes a higher-pressure fluid that is circulated between dual seals.
    • Buffer fluid describes a low-pressure fluid circulating between dual seals.
    • Quenching fluid is introduced into the seal gland outboard the seal to wash away leakage and control the environment outboard the seal.
    • Jacketing fluid circulates around the outside the stuffing box to control stuffing box temperature.
    • Flushing fluid is fluid from an outside source introduced into the stuffing box that dilutes the pumpage. It is seldom desirable, but sometimes necessary.
  • The specification allows spring-loaded elastomers (O-rings) that do not have the ability to flex and roll.
  • The specification allows a single spring seal design even if it is sensitive to the direction of rotation.
  • The specification does not prohibit the use of mechanical seals that frett (damage) shafts and sleeves.
  • The specification should call for the seal’s dynamic O-ring to move towards a clean surface to prevent “hang up”.
  • The standard does not recognize seal designs with both faces spring loaded and hydraulically balanced.




  • On February 14, 2018