Page 262 - Pipeline Risk Management Manual Ideas, Techniques, and Resources

Incorrect operations index 11/239
quences of the leak. This is the logic behind the discussion of mental actors such as wind, temperature, and the presence
leak detection in the discussions in Chapter7. However, the sit- of other interfering fumes in the area. Therefore, specific
uation for distribution systems is thought to be different. survey conditions and the technology used will make
Leakage is more routine (and even expected, for reasons previ- many evaluations situation specific. An estimate of survey
ously noted) and leak detection and repair is a normal aspect of effectiveness (&loo%) can be made part of the risk assess-
operations. ment. A default for test effectiveness can be used when no
Distribution systems tend to have a higher incidence leaks further information is available-a value such as 70% might be
compared to transmission systems. This is due to differences in an appropriate default. This can be combined with two more
the age, materials, construction techniques, and operating envi- factors to score this variable: Amount of system surveyed and
ronment between the two types of pipelines. With the increased Zme since lasf survey (see discussion of information decay,
opportunity for leaked products to accumulate beneath pave- Chapter 2)
ment, in buildings, and in other dangerous locations and with A possible scoring algorithm could therefore be:
the higher population densities seen in distribution systems,
this higher leak propensity becomes more important, especially 100 (1 0 x years since last test) =time "YO
~
for gas distribution. Furthermore, leak rates often provide early Leak survey score = (maximum points) x (test effectiveness) x (amount
warning of deteriorating system integrity. Therefore, attention of system tested) x (time since last test)
to leaks should be a strong consideration in assessing the risks
of distribution systems. For example, a test method deemed to be 80% effective and
Regular leakage surveys are routinely performed on gas dis- performed annually over 50% of the system would score 9 x
tribution systems in many countries. Hand-carried or vehicle- (0.8) x (0.05) x (0.09) = 3.2 ifthe variable weighting is 9 points.
mounted sensing equipment is available to detect trace amounts The operator's use of established procedures to positively
of leaking gas in the atmosphere near the ground level. Flame locate a leak can be included in this assessment. Follow-up
ionization detectors (FID), thermal conductivity, and infrared actions including the use of leak rates to assess system integrity
detection are some of the technologies commonly used in leak and the criteria and procedures for leak repair should also be
detection equipment. The use of trained animals-usually considered. This variable can logically be weighted higher than
dogs-to detect small leaks is a ground-level technique that has suggested in Chapter 6 due to leak surveys' increased role in
also been successful. distribution systems. The risk model designer should determine
One of the primary means of leak detection for gas distribu- the weighting based in consideration of all other failure
tion is the use of an odorant in the gas to allow people to smell variables.
the presence of the gas before flammable concentrations are
reached. As a special type of leak detection, the use and poten- Training
tial failure of the odorization system can be covered in the leak
impact factor. Score this item as described in Chapter 6.
Other types of leak detection techniques include [6]
Mechanical error preventers
0 Subsurface detector survty-in which atmospheric sam-
pling points are found (or created) near the pipe. Such sam- The role of error prevention devices can be evaluated as
pling points include manways, sewers, vaults, other conduits, discussed in Chapter 6. As noted there, error prevention devices
and holes excavated over the pipeline. This technique may be might include
required when conditions do not allow an adequate surface
survey (perhaps high wind or surface coverage by pavement 0 Three-way valves with dual instrumentation
or ice). A sampling pattern is usually designed to optimize Lock-out devices
this technique. Key-lock-sequence program
0 Vegetation survey-which is also done in transmission lines 0 Computer permissives
as a part of routine air patrol. The observer seeks visual indi- 0 Highlighting of critical instruments
cations of a leak such as dying vegetation, bubbles in water,
or sheens on the water or ground surface. where points are added for each application up to a maxi-
Pressure loss test-in which an isolated section ofpipeline is mum ofpoints.
closely monitored for loss ofpressure, indicating a leak. If a section that does not have any applications (and hence
no opportunity for this type of error) is being evaluated, the
Ultrasonic leak detectors-in which instrumentation is used maximum points are awarded. Note that in scoring a section
to detect the sonic energy from an escapingproduct. for this item, upstream sections may need to be considered
Bubble leakage-used on exposed piping, the bubble leak- because the error can occur there and affect all downstream
age test in one in which a bubble-forming solution can be sections.
applied and observed for evidence of gas leakage.
Maintenance
Other leak detection techniques more commonly seen in
transmission systems are discussed beginning on page 160. A low score in maintenance should cause doubts regarding the
It is beyond the scope of this text to offer specific guidance adequacy of any safety system that relies on equipment opera-
on the effectiveness of various leak surveying methods. The tion. Because overpressure protection is identified as a critical
effectiveness of many leak surveys often depends on environ- aspect in a distribution system, maintenance of regulators and