Pipeline System Relief Scenarios

[Berenger]

Hi all,

I am working on a crude pipeline transporting around 3000 barrels/hour of crude.

I have done the system relief analysis, but have a few of questions:

Is blocked outlet a credible scenario? Especially at pump discharges? If not, why?

Finally, if sizing blocked discharge on a pump, do you have to size it for the entire 3000 bph flow?

Thanks for your help and direction.

Berenger.

 

[BigInch]

Scenario 1 might be: SCADA leak detection indicates abnormally low pressure downstream of the pump station and closes the pump station's discharge ESD valve. Pump goes (or perhaps not goes) into shutdown mode.

Scenario 2: The station's discharge ESD valve fails closed. Perhaps it run out of air supply due to instrument air compressor failure, or corrosion, or dirt in the air supply line.

Scenario 3: Any block valve in the downstream pipeline is closed and pressures at the pump station rises to pump shutoff head/pressure.

It would not be a credible scenario, only if there were no valves in the downstream pipeline.

Relief systems are sized for the flow that is possible at the relief valve's set pressure. It is often the flow that is possible at shutoff head and there must be at least enough flow out of the pipeline to ensure that all pressures remain <= MAOP (MAOP +10%, if it is a B31.4 design).

 

[georgeverghese]

Also check that pressure surge generated due to liquid hammer (caused by sudden closure of the ESD valve at the end of this pipeline) is within MAWP(assuming that this crude is all single phase liquid). This should be superimposed on the max pressure that the pipeline will see, which in turn, would be influenced by how the process controls are set up.

 

[LittleInch]

berenger,

barrels an hour is an odd term - normally its barrels/day or m3/ hour or gallons per minute.

Would be interesting to know what your relief scenarios are to see what is or is not included so far.

Blocked outlet is normally a credible scenario as outlined by BI.

As mentioned earlier, you need to provide pressure relief to prevent overpressure of the piping. If the blocked in maximum pressure, using the maximum inlet pressure to the pump is less than the design pressure of the piping, then it becomes an issue of pump protection and how long the pump can run in a blocked flow condition. It is not uncommon to find that piping downstream of a pump up to the station ESD valve or outlet to the pipeline or pump discharge valve is at a higher pressure rating than the pipeline to cover these issues and prevent the use of a pressure relief system and just use instrumented high pressure trips instead.

If you need a certain flow to restrict the maximum pressure then this becomes your maximum flow through the relief valve. this is not necessarily the same as full flow.

It is more common for the pump to be equipped with various high level pressure protection devices to shut down the pump than use full flow relief systems. You tend to find full flow reliefs downstream of a pressure break at the outlet of a pipeline / inlet into a plant or tank farm etc than at the start pumping staion.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[BigInch]

Before going into a HIPPS design, review the full implications, especially legal and appendix A, as well as historic operating, maintenance and testing performance of the operator of the local, not to mention manufacturing and quality control of all the component devices. Latest recommendations (

http://www.iogp.org/pubs/443.pdf)

reinforce the principal that HIPPS should be the [highlight #EF2929]LAST[/highlight] layer of protection, incorporating several other layers, such as a pump shutoff switch that trigger well before in advance of hipps activations and far in advance of exceeding MAOP.

http://www.hse.gov.uk/foi/internalops/hid_circs/technical_osd/spc_tech_osd_31.htm#pipesprotect

Weren't they trying to manually activate one of those HIPPS-type valves (blowout preventer) when Macondo ran amuck. ... The science works, but the implementation can be far behind. And when one isn't enough, after Macando we seam to be in need of DUAL HIPPS ...

http://www.engineerlive.com/content/first-dual-redundancy-subsea-hipps

 

[racookpe1978]

Seems like - given the high cost of relief valve failure, the high cost of containing the effluent from relief valves in any pipeline or pumping station series, and the extreme enviro and political penalties for "any" leaks from "any" pipelines today, that the industry - and I am NOT in the pipeline industry! - would have a expansion chamber or relief chamber pre-fabricated that could absorb low and intermediate order shocks to absorb such pressure surges without lifting the relief valve.

 

[BigInch]

That most likely won't work. If you include a volume for surge within your net system volume, pressure of the system will immediately fill it. You must have a valve seat, or other trigger machine in between to keep the pressure out of it, until you need to fill it. That's what relieves the surge pressure, having enough low pressure extra volume available to get that surge volume out of the overpressurized part of the system. If it already had the system pressure contained within it, pressure across the system would just increase.

A stand pipe could work, but for high pressure systems it would have to be very, very tall.

 

[LittleInch]

In case you missed it, the OP in this thread

http://www.eng-tips.com/viewthread.cfm?qid=394369

has finally advised this is a PD pump he has....

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[racookpe1978]

OK. Thanks for the clarification. 8<)