Another Spin on Pneumatic Testing Energy

[zdas04]

After reviewing the great references in Thread794-26767 I still have a question about the total energy in a pneumatic test.

I'm testing 8 km of 20 cm (8-inch), schedule 40 pipe to 51 barg - when I do the math, I get a stored energy of over 500 kg of TNT. The hoop stress during the test is 27% of SMYS, the line is WAY out in the boonies (closest house is 25 km), and it will be burried under 1.5 m of dirt so I'm not real concerned with injury to the public.

The question is: there is no way that the entire stored energy in a pipe-bomb this long will be spent in one catastrophic explosion (since the relative dP from any control volume to any other is very low only a few dozen meters from a rupture)so how do you approach "de-rating" the explosive force to an effective force?

David

 

[PVRV]

1. 27% of SMYS

1.1 Verify P/T class design & test pressure
1.2 Base the SMYS on the weakest member
1.3 Verify blind/spade pressure rating
as bolted/welded flat heads.

2. Consider,

2.1 dT versus dP
2.2 Consider why SV pop fully open Vs. PRV's

Granted the pipeline is under ground and its not clear how long the line is pnematic testing requires strict safety requirements.

Cheers

 

[zdas04]

PVRV,
This line is adding some functionality to an existing system, the MAWP is set by the existing system and there is no benefit to having a small subset of the system with a MAWP higher than the rest of the system - the 27% of SMYS is the correct test pressure for this pipe.

The line is 8 km long.

I've been considering that the worst-case scenario would be for one joint (about 9 m) to fail in the HAZ of the entire longitudinal weld (the pipe is ERW), so the length of one joint would be totally involved in the explosion. Then I calculated a choked-flow velocity out the end of the two adjoining joints. Finally I converted the velocity to a volume flow rate and used the AGA compressible flow equations to look the upstream and downstream pressures at each joint to find where you no longer have sonic flow in a joint - probably very conservative, but it looked like a good first cut. I'm sure there is a more elegant approach, but this brute-force method looks like it has promise.

David