Page 252 - Pipeline Risk Management Manual Ideas, Techniques, and Resources

Assigning risk scores 11/229
Corrosion is defined in the broadest sense here-any degra- to any internal corrosion potential promulgated by that product.
dation of a material in its environment. This encompasses many Therefore, ali portions have exposure to internal corrosion. If
possible mechanisms such as temperature degradation, graphi- the pipe is not exposed to the atmosphere, then the model
tization, emhrittlement, chemical deterioration of concrete, and assumes it is exposed to soil and is treated as being in a subsur-
other processes. face corrosive environment. For each exposure type-atmos-
As with other failure modes, evaluating the potential for cor- pheric, internal, subsurface-an assessment is made of the
rosion follows logical steps, replicating the thought process relative corrosivity of the environment. Each pipeline’s imme-
that a corrosion control specialist would employ. This involves diate environment is characterized based on its relative corro-
(1) identifying the types of corrosion possible: atmospheric, sivity to the pipe material-steel, concrete, or plastic, for
internal, subsurface; (2) identifying the vulnerability of the example.
pipe material; and (3) evaluating the corrosion prevention In the scoring system presented here, points are usually
measures used, at all locations. assigned to each condition independently and then summed
Corrosion mechanisms are among the most complex of the together to represent the corrosion threat. This system adds
potential failure mechanisms. As such, many more pieces of points for safer conditions. For example, for the subsurface
information are efficiently utilized in assessing this threat. corrosion variable, three main aspects are examined: environ-
Because corrosion is often a highly localized phenomenon, and ment, coating, and cathodic protection. The best combination
because indirect inspection provides only general information, of environment (very benign), coating (very effective), and
uncertainty is usually high. With this difficulty in mind, the cor- cathodic protection (also very effective) commands the highest
rosion index reflects the potential for corrosion to occur, which points. An alternative approach (also described in Chapter 4)
may or not mean that corrosion is actually taking place. The that may he more intuitive in some ways, is to begin with an
index is therefore not directly measuring the potential for fail- assessment of the threat level and then consider mitigation
ure from corrosion. That would require inclusion of additional measures as adjustment factors. Here, the evaluator might wish
variables such as pipe wall thickness and stress levels. This is to begin with a rating of environment--either atmosphere type,
further discussed later in this chapter (corrosion rate discus- product corrosivity, or subsurface conditions. Then, multipliers
sion) and again in Chapter 5. are applied to account for mitigation effectiveness. For exam-
Three potential types of corrosion are commonly encoun- ple, in a scheme where increasing points represents increasing
tered in a pipeline system: atmospheric, internal, and suhsur- risk, perhaps a subsurface environment of Louisiana swamp-
face (Table 11.2). Atmospheric is considered to he the least land warrants a risk score of 90, very corrosive, while a dry
aggressive form of corrosion under normal conditions. Internal Arizona desert environment has an environmental rating of 20,
corrosion is a significant threat for unprotected water pipe, but very low corrosion. Then, the best coating system decreases or
less of a factor in most gas distribution systems. Subsurface offsets the environment by 50% and the best cathodic protec-
corrosion is seen as the highest corrosion threat for most metal- tion system offsets it by another 50%. So, the Louisiana situa-
lic pipelines. The higher threat is a result of potentially very tion with very robust corrosion prevention would he 90 x 50%
aggressive subsurface corrosion mechanisms, including vari- x 50% = 22.5. This is very close to the Arizona desert situation
ous types of galvanic corrosion cells and interference potential where no corrosion preventions are employed, but the environ-
from other buried structures, as well as the general inability to ment is very benign. This is intuitive because a benign environ-
inspect and gain knowledge of actual corrosion on subsurface ment is really roughly equivalent to a corrosive environment
components. Background issues of all types of corrosion are with mitigation, from a corrosion rate perspective.
discussed in Chapter 4. Further discussion of scoring options can he found in
The first step in assessing the corrosion potential involves Chapter 2. See also discussions regarding information degrada-
evaluating the pipe’s environment. This can be done most effi- tion on pages 25-3 1.
ciently by a risk model that has been populated with pertinent We now discuss the Chapter 4 corrosion variables as applied
information. The following discussion illustrates one approach to distribution systems. See Chapter 4 for background discus-
to characterizing each pipe’s environmental exposures (the sions of all corrosion mechanisms noted here.
threats to the pipe from its immediate environment).
The computerized risk model first searches for indications of Atmospheric corrosion
atmospheric exposure, including casings, tunnels, spans, valve
vaults, manifolds, and meters. These occurrences are noted in Where pipe materials exposed to the atmosphere are not sus-
the database and identify one of the potential threats as atmos- ceptible to any form of degradation, or where there are no
pheric corrosion. The model assumes that all portions of the atmospheric exposures, this variable can be scored as no risk
system are exposed to the product being transported and, hence, from atmospheric corrosion. The evaluator is cautioned about
discounting entirely the possibility of atmospheric corrosion.
For example, while plastics are often viewed as corrosion
Table 11.2 Corrosion index possible variables and weights proof, sunlight and airborne contaminants (perhaps from
nearby industry) are two degradation initiators that can affect
Variable Weight certain plastic materials. Note also that casings, tunnels, valve
vaults, and other underground enclosures allow the possibility
Atmospheric corrosinn 10 of atmospheric corrosion.
Internal corrosion 10
Subsurface corrosion 80 Where there are many atmospheric exposures and the pipe
Corrosion index total I00 material is susceptible to corrosion, the weighting of this
variable may need to he increased.