Parameters that assure optimum pump mechanical seal reliability

September 16 2013 - William E. Forsthoffer

Experience shows that pump MTBFs (mean time between failures) are directly related to mechanical seal reliability. Pump mechanical seal MTBF is the lowest of all machinery components. Of all machinery components, mechanical seals are the most affected by the process conditions.

Accurate definition of the following parameters on the pump and seal data sheet will go a long way towards assuring optimum pump mechanical seal MTBFs:

Fluid vapor pressure

Fluid specific gravity and viscosity

Fluid pumping temperature (PT)

Suction pressure

Cooling medium temperatures (if a flush cooler will be used)

External flush fluid pressures, temperature, vapor pressure, viscosity and specific gravity

Confirm process conditions are actually as stated in data sheets. If the ‘bad actor’ pump seal reliability, when operating in the EROE, does not significantly improve, a complete check of seal fluid conditions in the seal chamber is required. The process variables that influence seal reliability are:

  • Seal chamber temperature
  • Seal chamber pressure
  • Seal fluid characteristics like cleanliness, vapor pressure, viscosity, specific heat and specific gravity

Temperature monitoring

Since seal chamber or seal flush line pressure gauges are not usually installed on new pumps, the first place to start is with seal chamber temperature. Flush plans with coolers have an option for a thermometer downstream of the cooler which can be used to determine the seal chamber temperature. This value must be compared to the PT (pumping temperature) value listed on the data sheet.

If the measured value does not agree with the data sheet, consult with operations first to see if process changes can be made. If they cannot be made, discuss the measured temperature with the seal vendor representative.

Pressure monitoring

Unless the plant assigned machinery specialist is ‘world class’ there are usually no pressure gauges in the flush line or seal chamber. The measured seal chamber pressure must be compared to the seal vendor’s assumed value, which should be on the seal layout drawing. If this value is not present, the seal vendor should be consulted. Note that the seal vendor assumed seal chamber pressure is the value that was used in the calculation of the seal PV. If the measured value does not agree with the data sheet, consult with operations first to see if process changes can be made. If process changes cannot be made, discuss this fact with the seal vendor representative.

If the above mentioned items are in accordance with the seal design (EROE, seal chamber temperature and seal chamber pressure), a check of the seal fluid characteristics is required. If the measured fluid sample parameters are not as stated on the data sheet, consult with operations first to see if process changes can be made. If process changes cannot be made, discuss this with the seal vendor representative.

Variations in seal chamber pressures and temperatures, if the pump is operating in the EROE, will most likely be caused by malfunction of components in the flush system. Based on the seal flush plan used, the flush system components should be checked in the logical order – starting with the beginning of the flush system and ending with the throat bushing. (The exception is flush plan 13 which begins with the throat bushing and ends at the pump suction pipe.)

FAI ‘seal maintenance best practice’ requires that the flush system be completely checked (including the throat bushing clearance) each time a seal is changed. This recommendation applies to all seal configurations – single, tandem (dual unpressurized) or double (dual pressurized).  

To significantly increase your plant’s mechanical seal and pump MTBFs, confirm the actual seal operating conditions on the data sheet (PT, vapor pressure, SG, P1 and P2) to optimize mechanical seal MTBFs. Failure to properly specify the correct process conditions on the pump and/or mechanical seal data sheet will result in lower than optimum seal MTBFs.

Most ‘bad actor’ seals (seals with more than one failure per year) result from improper speculation of process conditions on the data sheets. Always check all operating seal process conditions against the data sheet values when investigating a ‘bad actor’ seal. If the proper instrumentation (pressure guage, etc.) is not installed, have the appropriate instruments installed at the time of seal replacement.

This best practice has been used since 1990 to troubleshoot mechanical seal problems and to ensure maximum mechanical seal MTBFs (greater than 100 months).