Environmental Engineering Reference
In-Depth Information
Table 5. Criticality matrix for system S
1
These methods are used to predict the probability
of accident of any system S
ij
of S
i
.
This criticality assessment is used to support
the subjective expert judgment on the initial power
grid system state. The more system criticality
calculated on (12) the more confident expert's
opinion on the criticality of each node of PG.
System S
1
Severity of failure mode
H
M
L
Fail-
ure
rate
H
S
12
M
S
13
L
S
11
Bayesian Belief Network as a Model
for Power Grid's Safety Assessment
level hierarchy. For example, considering the
criticalities of S
11
, S
12
, S
13
as subsystems of S
1
its
criticality could be calculated as:
The state of each PG system is determined by
types of influence mentioned above. The Figure 11
represents the different types of networks, which
characterize the same PG. Hence, all networks
have the same nodes as PG systems, but different
types of influence, which stipulate the different
causal links between nodes. The different colors
are used to show different types of influence (green
- physical influence, blue - geographical, brown
- organizational, red - logical, yellow - informa-
tional and black - societal). The different types
of influence are characterized by its own weight.
The more weight of the given type of influence
the more sensitive PG's safety to this type of influ-
ence. Apparently the physical influence is more
important, when PG safety is considered. But all
types of influences should be taken to provide a
more accurate PG safety evaluation. For each type
of influence might be introduced its own type of
PG system particular criticality. It means that PG
could be more vulnerable to the change of one type
of influence and at the same time be insensitive
to other type influence change.
Considering the types of influence mentioned,
it is assumed that the total PG system criticality
is a function of power grid system's particular
criticalities stipulated by the different types of
influence, i.e.
Crt S
(
)
=
P S
(
)
×
Sev S
(
)
+
P S
(
)
×
Sev S
(
)
+
i
1
1
2
2
I
∑
P S
(
)
×
Sev S
(
)
=
P S
(
)
×
Se
v S
i
(
).
3
3
i
i
(12)
It is suggested to treat criticality as power grid
system's safety inverse index. The more system
criticality the less its safety and vice versa.
It is worth to note that a probability of the sys-
tem accident and its severity could be handled as
a linguistic or numerical value. Hence, criticality
is also treated correspondently either linguistic or
numerical value.
The set of states Ω
Si
of any PG system S
i
is
determined as:
Ω
Si
= {Crt (S
i
)=High, Crt (S
i
)=Medium, Crt
(S
i
)=Low}.
(13)
Any accident or failure of the power grid system
leads to the change of criticality of all connected
systems. When a failure of one system occurs,
our technique recalculates the criticalities of all
dependent systems.
The prognosis and assessment of PG system
service life based on real time measurements will
help to identify grid systems most likely to fail.
The potential estimation methods and equipment
service life prediction for complicated systems
consist of deterministic, statistical, physical-
statistical and methods based on expert knowledge.
Crt S
f Crt
(
)
=
i
org
(
(
S Crt
),
fhys
(
S Crt
),
geo
(
S Crt
),
log
(
S Crt
),
soc
(
S Crt
),
inf
(
S
))
i
i
i
i
i
i
(14)
Search WWH ::
Custom Search