5 replies to this topic

[kkala]

An existing exchanger will heat desalter feed (water, tube side, 35ton/h) from 125 to 129 oC and cool gasoil (shell side, 14 ton/h) from 146 to 130 oC. Design and operating pressures are:
Tube side (water) : 31.6 kgf/cm2 g * (operating 21 kgf/cm2 g).
Shell side (gasoil) : 14 kgf/cm2 g * (operating 10 kgf/cm2 g).
* same for upstream / downstream piping.
Note: Certified hydrostatic test pressure of shell side : 21 kgf/cm2 g

According to API 521 (para 5.19: heat transfer equipment failure) “complete tube rupture is a remote but possible contingency” in this case. This could be cured by pressure relief on the shell side or connected vicinity piping. Nevertheless “the tube rupture scenario can be mitigated by assuring that an open flow path can pass the tube rupture flow without exceeding the stipulated pressure”.

Gasoil circuit is shown on the attached diagram. Downstream exchanger, gasoil passes from an air cooler and a LCV (level control valve) interlocked with the crude tower side, then ends to storage. There are also a lot of manual valves along the way, none of them locked open.

Suppose that a tube rupture occurs suddenly under normal operating conditions, and there is no LCV installed. Water flows with gasoline to storage, and resulting ΔP is not high enough to create overpressure, as long as all manual valves remain open. Can we avoid PRV installation, considering the above line as an open flow path? According to a vew closing a manual valve would make a double contingency.
Now suppose the previous event, but with the LCV already installed on the line. Its opening depends on Level in the Crude Distillation Tower, not affected by tube rupture. Can the line be considered as an open flow path?
Advise on the point and the questions would be appreciated.

It is pointed out that tube operating pressure exceeds shell design pressure.

Attached Files

•  ETR.doc   32.5KB   96 downloads

Edited by kkala, 06 May 2010 - 04:05 PM.

[fallah]

In my opinion,only under below conditions mentioned route could be considered as open flow path:

1-No LCV installed:All inline manual valves should be LO and also line Delta P value during tube rupture be such that shell pressure couldn't pass design value.

2-LCV already installed:All conditions in case 1 would be provided and also the LCV furnished by internal mechanical stop to prevent full closing.

Hope above helps.

[Zauberberg]

This is somewhat different configuration than what I used to see in Crude Units, but I can give some thoughts about it. I can only refer to Crude/Diesel, Crude/TPA, Crude/Kerosene, or Residue/Crude exchangers but the philosophy may be reflected to your case as well.

1. As Fallah has pointed out, you need to consider the worst possible scenario that may occur in reality. Normally you will not pump the desalter water into process without having gas oil (or tower bottoms) flowing to the battery limit. This would mean that the "open path" will normally exists throughout the gas oil loop - BUT:

2. During startup or shutdown or process upset, you may encounter the situation where the Tank Farm people close the gas oil inlet valve to any of the storage tanks, for whatever reason, while you are continuing to pump the desalter water into process. I have dozens of Crude Unit startups in my past and I know these situations can be quite common. In such case, you don't have an "open path" anymore.

3. Without this concern, how tolerable is water in the gas oil - at process temperatures, and in the product rundown tank?

4. In all exchangers (Crude/Residue), Crude being at higher pressure, PRV's are provided. Atmospheric Residue flows to the Vacuum Distillation Unit and, if enriched with light components from the Crude, it can seriously affect VDU operation.

I believe that having relief valves is the most common, and the safest configuration. We may opt for alternative configurations and the question we need to answer is - at what cost, and whether is worth to do so?

[kkala]

Sincere thanks to fallah and Zauberberg for their thoughtful answers. Client was unwilling to install a PRV discharging into a far drum, in a unit already crowded. But future operation would feed the gas oil to a coalescer (to retain water), whose water leg HHLA shuts its feed pipe through a control valve. This would occur in case of exchanger tube rupture, passing water into gas oil. So a PRV shall be installed to protect exchanger, leading to a satisfactory end.

Searching the matter meanwhile, I have found a design practice of 1980s saying that: in investigating a potential open flow path (i.e. gas oil route), isolation valves can be assumed fully opened and control valves in a position equivalent to minimum normal flow (of course if they do not close because of the tube rupture).

I wonder whether these statements can be still considered reliable for design, or have been turned more conservative in recent 25 years. Can some member give an opinion? Before looking into the matter, I thought that all valves of the path should be locked open, yet I see practices give some relaxation.

PS (16 May 10): The question also concerns the "desalter feed cooler" mentioned in ETR.doc and installed downstream Desalter after a manual valve. Brine passes through tubes in this exchanger and the Desalter itself is protected by PRV. In the specific case there is almost no doubt that Desalter PRV could protect "desalter feed cooler", but the question is more general. For instance, what if several manual valves and control valves were placed between?

Edited by kkala, 16 May 2010 - 11:16 AM.

If I understand correctly from the case description, water side operating pressure is 21kgf/cm2 and shell side design pressure is 14 kgf/cm2. I think the way tube rupture case should be looked is, if the low pressure side (LPS) blocked in when the tube ruptures, the high pressure side (HPS) fluid will not over pressure the LPS above 1.5 times the MAWP of the LPS. In this case, 14*1.5 = 21. I don't think it needs a RV to protect HX shell side, unless the HPS operating pressure is higher than 21 psig.

[kkala]

If I understand correctly from the case description, water side operating pressure is 21kgf/cm2 and shell side design pressure is 14 kgf/cm2. I think the way tube rupture case should be looked is, if the low pressure side (LPS) blocked in when the tube ruptures, the high pressure side (HPS) fluid will not over pressure the LPS above 1.5 times the MAWP of the LPS. In this case, 14*1.5 = 21. I don't think it needs a RV to protect HX shell side, unless the HPS operating pressure is higher than 21 psig.

API 521 (5th edition, 2007), Heat-transfer equipment failure, para 5.19.2 states: Loss of containment is unlikely to result, where pressure in LP side during the tube rupture does not exceed the corrected hydrostatic test pressure.

Data stated for the exchanger: tube operating pressure=21 kgf/cm2 g (design pressure=31.6 kg/cm2 g) &
shell design pressure=14 kgf/cm2 g (shell hydrostatic test pressure=21 kgf/cm2 g).
Assuming tube operating pressure during tube rupture = 21 kg/cm2 g, requirement of API is fulfilled.
But we are not certain that operating pressure in tubes would not be higher than normal during tube rupture. One step for more safety would be to assume max operating pressure. And a further step is to consider design pressure instead of operating pressure. Under the latter view, requirement is not fulfilled (hydr test pressure of LP side=21 kgf/cm2 g, HP side design pressure=31.6 kg/cm2 g).
The rule of 2/3 (10/13 for modern exchangers) concerns design pressures for both exchanger sides, at least as we apply it at work. I believe this is widely accepted.
Therefore this "conservative" view of comparing design pressure of HP side to hydrostatic test pressure of LP side seems to be widely applied.

Edited by kkala, 03 July 2010 - 01:49 AM.