Q & A > Question Details
Currently my plant is experiencing overhead vacuum fluctuation from 20 mmhg to 40 mmhg.
The design overhead vacuum is 20mmhg and maximum throughput is 20MB.
The ejector system consist of 3 stage ejector.
The first stage ejector consist of 2/3 ejector and 1/3 ejector load.
The second stage ejector consist of 3 ejector, and normally 2 out of 3 online.
The third stage ejector also consist of 3 ejector, and normally 2 out of 3 online.
We had perform field survey and found that the second stage ejector temperature is relatively low compared to the other ejector (26degC vs 70 degC)
Earlier, we suspect air ingress in to the ejector and we had perform online inspection. and indeed, we found 1 coin size leak at one of the first stage ejector and the leak had been repaired. however, the vacuum fluctuation is still there.
We had also verified all the other ejectors for leaks but unfortunately no leak was found.
We are also having issue with the ejector condenser. the third stage ejector outlet temp is relatively high compared to the other condenser (65degC vs 40 degC). This problem was there since a few years which had eliminate the condenser as the root cause of the fluctuation.
Currently we are trying to search of other weak point which can cause air ingress into the ejector/vacuum system.
Appreciate your feedback on the matter.

02/07/2015 A: Sudhakara Babu Marpudi, Dangote Oil Refinery Company, m_sudhakarababu@yahoo.com
Ejector nozzle erosion also could lead to vacuum fluctuations. However for checking the new nozzle performance one may need a plant shutdown. One refinery experienced the severe vacuum fluctuations due to the suspected liquid entrainment to the ejectors and the overloading of ejectors (plant capacity was increased and ejectors could not handle that load). This was resolved by manipulating the top reflux flow and then by adding additional ejectors.
01/07/2015 A: Abdulkarim Aldabass, Saudi Aramco, dabassaa@yahoo.com
You need to check the Barometric legs as they might be plugged. Check the temp over the line by skin temp or by splashing water on the pipe, the temp shall be much lower than normal, an X-ray can be used to check for deposit. Also, you need to conduct pressure survey for the vacuum condenser and ensure no fouling. You may need to increase the cooling water rate. Also check the motive steam pressure across each ejector.
01/07/2015 A: S Sivasankaran SABANAYAGAM, CPCL, cpclsiva@yahoo.com
There could be air ingress from the flange gasket joints apart from corrosion of pipeline @downstream of ejector.
Helium leak detection is one the method suitable to detect leak in any vacuum system. We have seen the effectiveness of the helium leak detection system in one of the surface condensers where air ingress even in the valve glands have been identified and rectified subsequently.
The only constraint in the above technique is that approach is required wherever we need to check the integrity.
30/06/2015 A: keith bowers, B and B Consulting, kebowers47@gmail.com
When did the vacuum fluctuation first appear? Is flooding on a tray in the tower a possibility?
Check steam supply conditions vs design values. It the tower stable or are the temperatures fluctuating.
How long is the pressure cycle---instant, seconds, minutes, hours?
2nd ejector low outlet temp = low compression ratio, loading 3rd stage which may be choking.