Environmental Engineering Reference
In-Depth Information
Table 15.1
Best practice principles in support of asset management tools
Appropriateness
Appropriate level of data collection and analysis re
ecting the level of risk associated with an asset
and the uncertainty within the decision being made
Understanding
Improving understanding of assets and their likely performance
Transparency
Transparency of analysis enabling audit and justi
cation
Structure
Structured knowledge capture encapsulated through fault tree, breach potential, etc.
Tiered assessment and decision-making
In terms of both data and modelling approaches
Collect once use many times
Reusing data through the hierarchy of decision-making stages and supporting tools
-
from
national policy to local detail
Simple use and practical
There is a signi
cant challenge in converting good science into practical tools. Therefore, even
though the underlying analysis may be complex, the user experience must be well constructed
and intuitive
Common/Central Databases
Asset reliability
The reliability of an individual asset can be ex-
pressed in a number of ways. Typically two are
used, including: (i) the probability of failure during
a given time period (e.g. a year); and (ii) the con-
ditional probability of failure for a given load
(Fig. 15.6), referred to as a fragility curve (Casciati
and Faravelli 1991; Sayers et al. 2002).
Reliabilitymethods can be used to derive either
the annual probability of failure or a fragility curve
(Melchers 1999). This involves the specification
and evaluation of a so-called Limit State Equation,
in general form:
Common databases (Fig. 15.5) provide a means of
accessing data and progressively evolving data
quality (supportinga 'collectonce,usemanytimes'
policy). The importance of such a system, and the
difficulty inachieving it inpracticeacrossmultiple
stakeholders, shouldnot beunderestimated.With-
in England and Wales, for example, the National
Flood and Coastal Defence Database (NFCDD)
provides acommonhome for asset data- regardless
of ownership - but significant difficulties associ-
ated with access and data quality have been en-
countered. Similarly, in theUSA, aNational Levee
Database is currently under development. Al-
though not without technical and organizational
difficulties an NFCDD (or its equivalent) is a fun-
damental component of any asset management
system, without which data collection and analy-
sis activities are easily repeated and effort wasted.
p
ð
f
Þ¼
P
ð
R
S
Þ < 0
ð
15
:
1
Þ
where p(f)
¼
probability of failure (typically de-
fined as breach, blockage or failure of a pump); S
¼
¼
resistance of the asset
loading on the asset; andR
to the loads.
In traditional reliability analysis, the uncondi-
tional probability of failure (e.g. the annual proba-
bility of failure) is determined through integration
of the joint density function (f
RS
) of the loads and
strengths over the region where the limit state is
exceeded (i.e. R
Understanding the Performance of an
Individual Asset
Understanding the performance of an individual
asset under load is the first step towards under-
standing how best tomanage it. The geometry and
structural components of the asset together with
the loading it experiences (e.g. waves, water levels,
etc.) and the associated probability of failure are all
important. Inspection methods (intrusive and
non-intrusive; Long et al. 2008) and reliability
analysis (see below) provide vital aids to the asset
manager in developing this understanding.
S
<
0):
ð
ð
f
RS
ð
p
ð
f
Þ¼
P
ð
R
S
Þ < 0 ¼
r
;
s
Þ
ð
15
:
2
Þ
af
To derive a fragility curve, a similar process is
followed. In this case, however, the loads are
treated as known deterministic variables (hence
