Hi All,
I need some advise on the following problem, and would like to know if the design is safe and sound.
Typically PSV is not required for centrifugal pump if the downstream system is designed for pump shutoff. In our case, the plate exchangers are unfortunately not designed for the pump shutoff pressure.
Refer to the attached for the simple process flow diagram for the seawater system.
We have a seawater lift pump supplying seawater to the diesel engine coolers.
- Seawater lift pump is a submersible centrifugal pump with shutoff head of 12 barg.
- Two engine coolers are plate exchangers. The plate and frame exchangers are designed for 6.9 barg. The flow through the coolers are maintained by the individual restriction orifice. Combined flow to the coolers (60 m3/hr) are slightly lower than the required pump minimum flow (80 m3/hr).
- A pressure control valve is provided for pump minimum flow, which is set at 7.5barg.
Questions:
1. With all the manual valves locked open at the inlet/outlet of plate exchangers, we eliminate the blocked outlet scenario which may cause the plate exchangers to be overpressure. Similarly thermal blocked in can be avoided too. Therefore PSV is not required. Correct?
2. If the plate exchangers are choked/blocked, then there is a potential that the exchanger is subjected to upstream pump pressure. As per API 521, we can take credit of the normal outgoing path with only the subjected equipment/valve shut. Given that the pump minimum flow valve is open during normal operation since the combined flow to engine coolers are lower than the pump min flow, then the exchangers will see 7.5 barg upstream. Overpressure of the exchanger is possible. However if we reduce the pressure setting for the control valve to 6.5 barg, then overpressure is not possible. Is this design considered safe and acceptable?
3. Typically the external gasket for the plate and frame exchangers will give way/leak at high pressure, is PSV really required to protect the exchangers?
4. Alternatively we can replace the PCV with another RO. In this scenario there will always be flow through the system. Double jeopardy to consider RO and plate exchangers to block at the same time.
Thanks all.
Attached Files
Edited by mykid, 06 October 2016 - 04:10 AM.