3 replies to this topic

[nosaj olegna]

Good day,

I would like to ask if i will still need a pressure relief valve on my crude oil pipeline design.

Below was the design parameters:
1. Pipe length was 29 km
2. Pipe diameter was 8 in
3. Velocity was 1.5 ft/s
4. MAOP of pipe was 616 psi.
5. Operating pressure was 155 psi.
6. Flowrate was 241 gal/min

Your response will be highly appreciated.

Thank you in advance.

Jason CHE '05

[GS81Process]

How is the pipeline protected against transient pressure surges?

[asade abiodun]

Determine the transient surge pressure that could occur in the pipeline due to valve closure or any sudden obstruction. If the pressure is below the MAOP of the pipeline, then you can decide not to install PSV. Also, you must ensure that the pipeline is adequately protected.

According to API RP14C, Seventh edition, March 2001, PSV should be installed but its non-installation should satisfy the objections stated in table A-9.2.c.2-5 of the document.

[kkala]

1. Suppose a 29 km horizontal pipeline of Sch80 with 1.5 ft/s liquid velocity (213.5 gpm). A valve at end closes suddenly enough (say in 42 s or shorter) to create full water hammer effect
α. If transferred liquid is water (at ambient temperature), resulting water hammer will create a pressure rise of 6.3 bar = 91 psi (according to Perry, Hydraulic transients, p. 6-44 in 7th edition). In this calculation steel wall elastic modulus = E = 207E9 Pa, water bulk modulus of elasticity = β = 2.2Ε9 Pa, found pressure wave velocity = 1376 m/s, using Perry's data and symbols.
β. Attachment "psi_chart_as_per_pipe_wt.pdf" (from Web) estimates water hammer of case α at 213.5*0.422 = 90 psi, indicating same base.
γ. If transferred liquid is crude (at ambient temperature), density = ρ = 890 kg/m3, bulk modulus of elasticity = b = 280000 psi = 19305 bar = 1.93E9 Pa can be assumed (see attached "crude.jpg"). Thus pressure rise due to water hammer is similarly calculated to 5.6 bar = 81 psi (wave velocity=1378 m/s).
2. Assuming 155 psig pressure at the beginning of pipe, there may be a centrifugal pump upstream or not. In the pump case, pressure on pump closure will be shut off +water hammer pressure, probably not far from 1.25*155+81 psig = 275 psig (*). Chances are that pipe line design pressure is not exceeded, in this probable case PRV or other means for water hammer protection is not needed. But this has to be verified using actual plant data and preferably software to investigate water hammer cases.
Note (*): Factor 1.25 is not related to water hammer, but to the fact that a lot of centrifugal pumps have shut off head ≤ 1.25×rated head, the latter assumed (corresponding to) 155 psi. Water hammer investigation through software has to consider pipe line elevations, maximum crude flow and whole line starting from the pump (if any).
3. TLnet was used here in a similar local case (transferring fuel at ~ 20 km), where total pressure (including water hammer) did not exceed 10% of pipe line design pressure (it can be 0% depending on code), thus no need of surge protection. Like in this thread, liquid velocity was quite low, probably aiming at low surge over pressures. (http://www.cheresources.com/invision/topic/13890-pipeline-surge-relief-system, towerds end of 1st page).
4. In another similar case, surge tanks were considered to alleviate water hammer. I have the impression that PRVs for water hammer protection need very fast response and (at least today) their use is limited. See http://www.cheresources.com/invision/topic/12797-sizing-relief-valve-to-protect-against-surge-pressure. At any case advice on this is welcomed.

Attached Files

•  psi_chart_as_per_pipe_wt.pdf   1.49MB   48 downloads
•  crude.JPG   118.07KB   9 downloads

Edited by kkala, 21 January 2013 - 02:58 AM.