Page 338 - Pipeline Risk Management Manual Ideas, Techniques, and Resources

Case studies 14/315
For ACME Pipeline release modeling, a worst case rupture is Table 14.40 Four potential damage ranges for each of the nine
assumed to be guillotine-type failure, in which the hole size is equal to failure scenarios under discussion
the pipe diameter. at the pipeline’s 15,305-kPa (2220-psig) Maximum
Allowable Operating Pressures (MAOP). This worst case rupture is Thermal radiation level (Btu/hr-ft2i Description
further assumed to include a double-ended gas release that is almost
immediately ignited and becomes a trench fire. 12,000 100% mortality in -30 sec
Note that the majority of the ACME Pipeline will normally operate 5,000 1 % mortality in -30 sec
well below its post-installation, pressure-tested MAOP in Canada. 4,000 Eventual wood ignition
Anticipated normal operating pressures in Canada are in the range of 1,600 Onset injury -30 sec
800 to I 100 psig, even though this range is given only a 40% probabil-
ity and all other scenarios conservatively involve higher pressures.
Therefore the worst case release modeling assumptions are very con-
servative and cover all operational scenarios up to the 15.305-kPa impacted by any assumptions relative to leak detection capabilities.
(2220-psig) MAOP at any point along the pipeline. This is especially true since the damage states use an exposure time
Other parameters used in the failure scenarios cases are ignition of -30 seconds in the analysis.
probability and thermal radiation intensity (Table 14.39). Ignition
probability estimates usually fall in the range of 5 to 12% based on
pipeline industry experience; 65% is conservatively used in this Results
analysis. Results of calculations involving nine failure scenarios and four dam-
The four potential damage ranges that are calculated for each of
the nine failure scenarios are a function of thermal radiation inten- age (consequence) states as measured by potential thermal radiation
intensity are shown in Table 14.41
sity. The thresholds were chosen to represent specific potential dam- The nine cases are shown graphically in Figure 14.3. The right-
ages that are of interest. They are described generally inTable 14.40. most end of each bar represents the total distance of any conse-
These were chosen as being representative of the types of potential
damages of interest. Reference [83] recommends the use of 5000 quence type. The farthest extent of each damage type is shown by
the right-most end point of the consequence type’s color. These nine
Btu/hr-ft* as a heat intensity threshold for defining a “high conse- cases can also be grouped into three categories as shown in Figure
quence area.” It is chosen because it corresponds to a level below 14.4, which illustrates that 11% of all possible failure scenarios
which:
would not have any of the specified damages beyond 29 ft from the
failure point. Of all possible failure scenarios, 55% (44% + 11%)
-Property, as represented by a typical wooden structure would not be would not have any specified damages beyond 457 ft. No failure
expected to burn scenario is envisioned that would produce the assessed damage
-People located indoors at the time of failure would likely be afforded states beyond913 ft.
indefinite protection and In these groupings, the worst case (largest distance) IS displayed.
-People located outdoors at the time of failure would be exposed to a For example, the specific damage types can be interpreted from the
finite but low chance of fatality. chart as follows:
Given a pipeline failure, 100% (-44% + -44% + -1 1%) of the
Note that these thermal radiation intensity levels only imply damage possible damage scenarios have a fatality range of 333 ft or less (the
states. Actual damages are dependent on the quantity and types of longest bar). There is also a 56% chance that, given a pipeline fail-
receptors that are potentially exposed to these levels. A preliminary ure, the fatality range would be 167 ft or less (the second longest
assessment of structures has been performed, identifying the types of bar).
buildings and distances from the pipeline. This information is not yet
included in these calculations but will be used in emergency planning.
Case Study B: natural gas
Role ofleak detection in consequence reduction
Table 14.42 shows results of modeling as described in Ref.
The nine failure scenarios analyzed represent the vast majority of all [67]. The analyses were performed on a 150-mm-diameter nat-
possible failure scenarios. Leak detection plays a relatively minor ural gas pipeline using various pressures and hole sizes with
role in minimizing hazards to the public in most of these possible corresponding release rates and ignition probabilities. Two
scenarios. Therefore, the analysis presented is not significantly damage states, based on thermal radiation levels were of inter-
est to these investigators. Failure probabilities are based on
Table 14.39 Additional parameters for the nine failure scenarios European Gas data with adjustment factors as shown in
under discussion Table 14.7.
Ignition probabiliti:
given failure Case Study C: gasoline
Hole size (in.) has occurred (%) Comments
This case study is extracted from Appendix 9B of Ref. [86],
50% to fnll-bore 40 Larger release rates, as which is an environmental assessment (EA) of a proposed
rupture (8-16) driven by larger hole -700-mile-long gasoline pipeline, called LPP, from Houston
diameters, may find more to El Paso in the state of Texas. Portions of this pipeline
ignition sources due to existed and new portions were to be constructed. Existing
the more violent nature of portions were in crude oil service under the former owner-
rupture and larger ship of a company herein referred to as EPC. MTBE refers
volumes of gas.
0.5-8 20 to a gasoline additive that was being contemplated. This
<0.5 5 additive makes the gasoline more environmentally persistent
and hence, increases the chronic product hazard. This EA