Medical Gas Professional Healthcare Organization

Leading through education, we save lives


Medical Vacuum Pump Failure

  • Wednesday, January 13, 2021 7:25 PM
    Message # 9857883
    Al Moon (Administrator)

    The following is a list of what, I believe to be true.

    ( see picture )


    Starting off like a ( ESCAPE ROOM ) statement right?


    1) very large rotary oil vane pump over 30 hp

    2) very large healthcare facility

    3) unit was serviced and maintained per the factory o & m

    4) standard oil was used per the o & m

    5) unit used for a combo medical surgical and wagd vacuum source 

    6) several random oxygen analyzer samples showed levels of over 23.6% O2 at the roof top exhaust line

    7) the healthcare facility has 30+ surgery suites

    8) the healthcare facility has a high volume of covid 19 patients


    So other then the probability of the fire triangle being completed.

    What answers or questions can be brought to lights?



    1 file
  • Thursday, January 14, 2021 8:07 AM
    Reply # 9858639 on 9857883

    I have seen this same scenario 3x before, although the pump did not melt the housing through. The oxygen content was not measured at the time, I believe that it is not the root cause. The facilities are smaller and did not have high surgery cases or COVID patients at the time of failure. 

    This is what I believe happened to the pumps:

    1) 1 or multiple exhaust filters fell out of its slot and was sitting in the pump housing. (see pictures, one filter fell out and was touching the pump housing).

    2) This allowed all the oil to run out of the exhaust through oil vapor (some oil residue could be present on the roof). 

    3) Pump ran almost completely out of oil. Normally this would trip high temp and shut the pump off, in all 3 of my experiences the pumps either did not have a high temp cut off or it had failed to initiate. The facility staff measured the housing at over 600 degrees in one case after the incident.

    4) The oily filters touching the housing became the fuel, the heat created the ignition with the oil, and the exhaust being open to atmosphere and pushing through at minimum atmospheric air was enough oxygen to create a fire out of the exhaust filters. 


    Even if the pump is properly maintained, I have seen exhaust filters fall out due to worn out clips or improperly installed.

    The oxygen enriched exhaust could have made the fire burn hotter than normal but is likely not the root cause. The extra quantity of exhaust filters in your scenario would have burned longer which is why the housing melted through.


    Curious to know, did the pump have a high temp shut down installed? Have the filters clips been replaced before or how long ago was that?



    2 files
    Last modified: Thursday, January 14, 2021 8:11 AM | Charlie Campbell
  • Tuesday, June 29, 2021 8:00 PM
    Reply # 10712511 on 9857883

    Pump in Al's comments were serviced with synthetic oil the Thursday prior to the fire in the exhaust filter chamber. The exhaust filters was the only part of the exhaust box which caught fire. The exhaust filters were seated and new (less than 3 days old).

    Forensic consultant blames the exhaust filter for the ignition. I am not convinced.

    The monitors have been installed on all 4 vacuum pumps and are recording tons of data being collated by a third party.

    The high temperature sensor is located in another chamber in the exhaust box monitoring the oil temperature which is around 190 degrees at the time of the fire.


    Fire started 8 minutes after all OR cases started at 7:30 am at this facility.


    Once a statistically viable amount of data is collected a white paper is in process.


    Again the source of ignition is not conclusive it is the exhaust filters.


    The pump manufacturer states they have an exhaust filter rated up to 230 degrees C in 100% oxygen.


    The challenge with the monitor is the level of oxygen at -23" Hg and the code statement of 23.5% oxygen for the monitor to mitigate the high oxygen atmosphere is definitely inconsistent with the reality of the monitor placement in the vacuum stream and the evidence of where the fires occur in these systems. 

    The problem is the oil mist incompatibility with the oxygen analyzer technology deviates from it's analytical accuracy when coated in a thin layer of oil.


    Your thoughts and experiences you may have to share on this subject is appreciate. 


    I am currently active on the team monitoring the pictured vacuum system for the past number of months and we currently meet at least one a week to followup on this system and monitoring being placed.

16339 Kranker Drive, Stilwell, KS 66085

mgpho@me.com