The existing design of de-methanizer column relief system is re-validated to ensure process safety and integrity. Relief load estimation for de-methanizer blocked outlet scenario requires careful evaluation as it is a reflux failure scenario. The simplified relief load estimation method adopted in design does not represent the relief load and relieving conditions accurately. The de-methanizer blocked outlet relief scenario involves many pre-events before over pressure happen, hence proper system review is to be done in analyzing the relief scenario and estimating the relief load. This paper presents the comparison of results from both rigorous steady state and dynamic simulation are carried out in estimating the relief load and relieving conditions close to reality.

By original design, relief load estimated from the simplified method resulted in under sizing of the de-methanizer PSV and also metallurgy for flare KOD pump and discharge piping is identified not adequate to handle very low temperature liquid anticipated during relief scenario. The design hypothesis of governing relief scenario of de-methanizer column (blocked outlet scenario) is reviewed with rigorous method / best practices to ensure the adequacy of the design of the relief system and flare KOD pumps.

With the simplified method adopted in design, the assumption of steady state initial conditions considered for relief load estimation results in very cold fluid relief and also the steady state composition of the vapor considered to be relieved will be arbitrary and will not represent the composition of the vapor during the real relieving scenario.

With the rigorous steady state simulation and dynamic simulations approach, carefully considering the pre-events in operation of the NGL recovery unit during the de-methanizer vapor blocked outlet scenario resulted in higher relief load and warmer fluid relief temperature and no significant liquid relief is expected.

Hence, It is very important to evaluate the relief scenario with appropriate relief conditions and operational upsets anticipated during the relief scenario in order to design the relief system appropriate to ensure the process safety in design

From the outcome of this study, it has been observed that the PSV is under sized for the required relief load due to warmer fluid to be relieved. It is also concluded that existing metallurgy for the flare KOD pump and discharge piping will be adequate. It is also observed that the existing flare KOD liquid hold up volume considered in original design is very much high, whereas the relief load estimation by rigorous method and dynamic simulation results clearly highlights the existing KOD is very much oversized.

Any assumptions / approximations and inadequate analysis made in relief load estimation may result in either significant overdesign of the system which will lead to more CAPEX, or undersized system which is a safety concern.

Design of relief system shall consider all the key factors affecting the relief load, by rigorous steady state simulation or dynamic simulation, particularly in estimating the relief load of complex process systems like NGL unit involving column relief.

You can access this article if you purchase or spend a download.