Environmental Engineering Reference
In-Depth Information
number of independent lengths considered to be in
Condition Grade i and Condition Grade j.
By assuming independence between the part of
the asset in Condition Grade i and that in Condi-
tion Grade j, a third estimate of annual probability
of failure can be estimated:
A simplified field-based approach to risk
attribution
As well as the rigorous approach to the attribution
of risk there is often a requirement to provide a
more simplified evaluation based on site inspec-
tions (without recourse to complex computational
modelling). A simplified tool, 'RAFT - Risk As-
sessment Field-based Tool' (Environment Agency
2009), can be used to provide a first estimate of:
.
the annual probability of failure (breach) - taking
account of geometry, structural condition and
loading;
.
the consequential impacts should a given asset
fail - taking into account a range of receptors;
.
the risk (taking account of probability and con-
sequence) attributed to an asset in its current
condition and assuming improvement to its target
condition.
RAFT provides the practitioner with an ability
to assess the contribution of an individual asset to
riskwith aminimumof data andmodelling. RAFT
uses the physical characteristics of an asset - i.e.
crest level and materials of construction - to iden-
tify the most suitable high-level RASP fragility
curve. It then uses the fragility curve, alongside a
user-specified asset length and extreme loading
conditions, to estimate the annual probability of
failure as follows:
P
fc
¼ 1ð1
P
fi
Þ:ð1
P
fj
Þ
ð
15
:
7
Þ
where P
fi
and P
fj
represent the annual probability
of failure for the proportion of the asset in Condi-
tion Grade i and Condition Grade j respectively.
The annual probability of failure assigned to the
asset as a whole is simply given as:
P
f
¼ max
P
fi
;
P
fj
;
P
fc
c
ð
15
:
8
Þ
This process ensures that the strength of the
asset is not greater than its weakest link (regard-
less of length) whilst reconsidering that as the
asset length increases, so will the chance of failure
(assuming all other aspects remain unchanged).
Below a limiting length of 300m the annual prob-
ability of failurewill be represented by theweakest
linkwithin the asset. The influence of asset length
on the annual probability of failure is shown in
Figure 15.10.
A potential inundation extent is estimated
based on the head of water above the floodplain
during the event, enabling the user to enter the
number of properties that may be inundated in the
event of a failure (either estimated or pre-calcu-
lated within a geographical information system)
and the associated risk calculated taking account
of both the probability of failure (i.e. breach) and
the associated consequences.
n
P
fi
¼ 1ð1
P
fCg
ð
i
Þ
Þ
ð
15
:
6
Þ
where P
fCg(i)
¼
the annual probability of a single
independent section of a given asset i failing (cal-
culated by integrating the fragility curve over all
loading conditions); and n
¼
number of indepen-
dent defence lengths within asset i that can be
considered to be Condition Grade j. The number
of independent lengths is simply calculated as the
total length of the asset divided by either 300m(for
hard defences such as walls and embankments) or
600m(for soft defences suchasbeachesanddunes).
In some instances, however, the condition of
a single asset may not be uniform. For example, a
given asset may have localized problems over
a short length with the remainder of the asset in
a better condition. In this case, the annual prob-
ability of failure can be calculated based on the
Developing Adaptive and Optimum
Intervention Strategies
Often, asset management consists of implement-
ing a range of physical interventions and data im-
provements staged in time and space, and the asset
manager is faced with many difficult questions:
.
What is the existingflood risk?Where is it?What
are the drivers?
.
Which assets contribute the most to flood risk?
