Radiation Distance and Measurement Length

From AS2885 INFO
Jump to navigation Jump to search

Some general cautions:

  • The radiation contour calculations in AS 2885 (using API 521 equation parameters) are deliberately conservative. Where more appropriate conditions are known designers can apply less conservative parameters provided they can be justified
  • All calculations of radiation distance (and hence measurement length) can only ever provide rough estimates. For a real failure the actual pressure, release location, pipe length, etc may be quite different to those assumed for AS 2885.6 Figures B1 - B4 or other pipeline-specific calculations. For this reason there is no point in obsessing over relatively minor differences in assumptions and calculation results.
  • The objectives of these calculations are firstly to provide a reasonable basis for location classification, and secondly to help with consequence assessment in an SMS. Neither requires high precision.

Some calculations of measurement length include an assumption that the release rate is doubled to account for flow from both ends of a fully severed pipeline. There is no equivalent statement for the graphs in Appendix B of AS 2885.6, and the statement in the last paragraph of B3 states that “As the hole size increases relative to the pipe diameter ... the results will approach the values for full bore rupture ...”. Do the radiation distances in Figures B1 – B4 of AS2885.6 Appendix B account for two open ends at a rupture?

It is true that AS 2885.6 does not explicitly state that Figures B1 - B4 are based on flow from both ends of a ruptured pipe, but that is certainly the intention as implied (perhaps a bit too subtlely) in B2(b) which says the rupture is assumed to be in the midpoint of the 50 km pipeline length.

The statement that release from a large hole approaches that from a rupture is considerable simplification. While the release from a hole equal to pipe diameter can clearly not be equal to the release from two open ends, it may in principle be considerably more than half the two-end release due to the complexities of pressure-flow dynamics inside the pipe.

The graphs in Appendix B are for specific pipelines conditions, including 50 km length. How would the results change as the pipeline becomes longer?

A ruptured pipeline will initially release gas at an extremely high rate which decays as the pressure at the opening drops and flow has to come from further up the pipeline with associated friction losses. The initial drop in flow is extremely rapid, which is why AS 2885 uses the flow 30 s after rupture; the peak flow can be very much higher but it is also very brief. As time progresses the flow decays tails off and is eventually very slow.

The rate of decay in the flow depends on the total volume of linepack. For long pipelines the linepack will sustain the flow for longer so the decay will be slower. Without doing further analysis, it is likely that the flow 30 s after rupture will be not be much affected by pipeline length for pipelines longer than about 20 km (still assuming mid-point rupture). On the other hand as pipeline length becomes shorter the lower total linepack is increasingly likely to reduce the release rate at 30 s.

In summary, the radiation contours in Appendix B are believed to be fairly realistic for long pipelines but conservative for shorter lines. For short lines, but not long ones, doing a transient flow analysis for the specific parameters of that pipeline may justify use of reduced radiation distances and shorter measurement length.

The 2012 edition of AS 2885.1 stated in Clause 4.10 that radiation contours shall be calculated based on flow 30 seconds after the rupture: "Where this Standard requires use of energy release rate or radiation contour it shall be established by calculation of the quasi-steady state volumetric (or energy) flow 30 seconds after the initiating event". AS 2885.6–2018 (4.9.3) has removed the clause to calculate radiation contours (and hence measurement length) at 30 seconds but Appendix B has kept the radiation contour examples as being calculated 30 seconds after rupture. On what basis does AS 2885.6-2018 require the radiation contours to be calculated?

There has been no intentional change to the basis for calculating radiation contours (and measurement length), so they should still be based on the flow 30 s after release. The change in wording between the 2012 and 2018 documents is an unfortunate omission; a clarification will be considered at the next revision.