6 replies to this topic

[Martin Khamphasith]

Dear Cheresources community,

I am a young graduate chemical engineer with no experience in pressure safety valve (PSV).
However I managed to find a entrance temporary job where I have to check the adequacy of all existing PSVs in an industrial plant.

The PSVs adequacy check consist of checking that the possible flowrate is sufficiently higher than the required flowrate to be relieved, and if the inlet/outlet pressure losses are respectively less than 3% and 15% of set pressure.

As there isn't any expert or experimented engineer where I am I couldn't find anybody, therefore I would like to ask your advices.

We have many cases where PSVs are installed downstream a pressure reducing valve (PCV) to protect the downstream system in case of the PCV failure (stuck open). Thses PCV are used to reduce the distribution pressure of utilities (steam, N2 and instrument air) to the pressure of the consumers.

In all these cases, the calculated inlet pressure drop is >3% (from 5% to more than 10% sometime). The situation is that all the PSVs are located directly downstream the PCV, in others words there is no possible modification of the piping layout or configuration, excepting to enlarge the inlet diameter. However, even by enlarging the diameter reasonably I do not manage to have an inlet pressure loss <3%.

My question is:
I used, for the inlet loss calculation, the piping elements from the outlet of the PCV until the inlet nozzle of the PSVs and the highest loss is attributed to a tee (with turning flow).
But if I calculate the pressure only from the branch where the PSV is connected (and not from the outlet of the PCV) I find a pressure loss less <3% of set pressure. So is it correct if I do it like this? or I must consider all the piping from the PCV?


Please advice
Thank you!!

Best regards

Attached Files

•  Picture.png   219.95KB   11 downloads

Edited by Martin Khamphasith, 17 March 2016 - 04:26 AM.

[Zauberberg]

Inlet losses = pressure loss between the system and the PSV. This does not include the system itself. In this case the system would be the process line upstream and downstream of the PCV, because it looks like the PSV is there to protect from piping overpressure in case of failure of the PCV (and possibly other reasons as well).

Inlet losses refer to pressure loss between the system discharge flange and the PSV inlet flange. The process line is a part of the system and no friction losses should be accounted for within the system, when calculating PSV inlet piping losses.

[Martin Khamphasith]

Thank sir for answering,

 

So if I understand you correctly sir, in the case attached to this message, the inlet pressure drop shall be calculated only from the branch (red line) and never from the valve (green line) and not from the upstream vessel as well as the massflowrate is fixed by the 2 valves in case of failures?

Please confirm

 

Sincerely yours

 

 

 

Attached Files

•  Picture 1.png   102.42KB   15 downloads

[Zauberberg]

The image is different now and it shows the surrounding equipment but with unknown design parameters.

In order to answer your question, please provide the design pressures of K52000, D59000, piping design pressure, and the set pressure for the PSV.

[Martin Khamphasith]

Hi Zauberberg,

 

The design pressures are 12 barg and 3 barg respectively for K-52000 and D-59000. The piping pressure grade is PN10 (max pressure 10bar).

The set pressure of the PSV is 2.7 barg. Based on old documentation when the plant where designed in 1998, the PSV is there to protect D-59000 from overpressure in case both control valves CV-52012 and CV-52013 fail open.

 

K-52000 is a reactor. At the end of the reaction, the gas phase (containing CF3Br and mainly CO2) is sent to the washing column D-59000 (operating at atm pressure) to scrub the CO2 before sending the washed gas into a cryogenic heat exchanger to liquefy the CF3Br and store it in a vessel.

 

At the end of the reaction, when starting degassing to D-59000 the pressure and temperature are 6 bar and 45°C. The degassing is done via a ramp of 0.04bar/min with the control valves CV-52012 and CV-52013, until the pressure in the reactor attain 0.04 barg.

 

Please advice for the pressure drop calculation.

 

Thank you very much

 

Regards

[Zauberberg]

Thanks for the additional data.

 

With this information we can confirm that the PSV is obviously there to protect from overpressure of D59000. You say the governing case is when both control valves fail open, but what happens with the flow through D59000?

 

1) Do you account for certain flow going through D59000 when control valves fail open, or

2) Do you assume blocked outlet and the entire flow going through PSV? At a first glance that would seem to be a double jeopardy but I don't know how the controls are configured.

 

The reason I am asking this is because my reasoning is as follows. If there is still forward flow through D59000 during relief (Case 1 above), I wouldn't consider the process line from the PSV to D59000 as a part of the PSV inlet line, due to direction of flow (from K52000 towards PSV). On the other hand, if the D59000 outlet is blocked (Case 2 above), I would include the entire line from the PSV inlet flange to the D59000 inlet flange, because in this case the relief stream flows from both sides and if the two lines (process line and PSV inlet line) are undersized for the relief flow, you may end up overpressuring D59000 above the design pressure.

[Martin Khamphasith]

Okay, thank you very much Devan IVANOVIC !

Regards