Research ArticleServiceability Assessment for Cascading Failures inWater Distribution Network under Seismic Scenario
Qing Shuang,1,2 Yisheng Liu,1,2 Jing Liu,1,2 and Qigang Chen3
1Department of Construction Management, School of Economics and Management, Beijing Jiaotong University,Beijing 100044, China2Beijing Center for Industrial Security and Development Research, Beijing 100044, China3School of Civil Engineering, Beijing Jiaotong University, Beijing 100044, China
Correspondence should be addressed to Qing Shuang; [email protected]
Received 29 November 2015; Revised 23 March 2016; Accepted 31 March 2016
Academic Editor: Stefano de Miranda
Copyright ยฉ 2016 Qing Shuang et al. This is an open access article distributed under the Creative Commons Attribution License,which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
The stability of water service is a hot point in industrial production, public safety, and academic research. The paper establishesa service evaluation model for the water distribution network (WDN). The serviceability is measured in three aspects: (1) thefunctionality of structural components under disaster environment; (2) the recognition of cascading failure process; and (3) thecalculation of system reliability. The node and edge failures in WDN are interrelated under seismic excitations. The cascadingfailure process is provided with the balance of water supply and demand. The matrix-based system reliability (MSR) method isused to represent the system events and calculate the nonfailure probability. An example is used to illustrate the proposed method.The cascading failure processes with different node failures are simulated. The serviceability is analyzed. The critical node canbe identified. The result shows that the aged network has a greater influence on the system service under seismic scenario. Themaintenance could improve the antidisaster ability of WDN. Priority should be given to controlling the time between the initialfailure and the first secondary failure, for taking postdisaster emergency measures within this time period can largely cut down thespread of cascade effect in the whole WDN.
1. Introduction
Water distribution network (WDN) is a basic componentin civil infrastructure systems. Its stability and reliability areimportant to ensure industrial production and public safety.Nowadays, WDN has developed into a large-scale networkwith thousands of pipes and nodes [1]. In such circumstance,how to design, improve, monitor, and repair the componentsefficiently in WDN becomes a critical problem in risk andreliability analysis. However, such analyses are often chal-lenging due to complex network topology [2], componentsinterdependency [3], and hydraulic calculation. Componentfailures may lead to the cascade effects and secondary failuresunder seismic scenario [4, 5]. This cascade reaction willprolong the repair cycle and lead to economic losses [6].Therefore, it is necessary to quantify the effects of suchcascading failures, to develop a system reliability evaluation
method under natural hazards and to further analyze theserviceability of the WDN.
Studies on the WDN reliability under seismic scenariohave been attracting extensive attention. Shinozuka et al. [7]analyzed the WDN of Memphis and evaluated the consumerdemand under seismic scenario. Hwang et al. [8] analyzedthe damage of pipes and the soil liquefaction under seismicscenario by using GIS technology. The serviceability wassimulated by using Monte Carlo method. Regarding thepower system as the backup ofWDN,Adachi and Ellingwood[9] made serviceability analysis on the interaction influenceof the power system andWDNunder seismic scenario. Brinket al. [10] evaluated and compared the WDN emergencymeasures of Los Angeles Hydropower Board under seismicscenario. However, the studies above are concentrated on thereliability and serviceability of pipe network under seismicscenario, instead of considering the effect of cascade.
Hindawi Publishing CorporationMathematical Problems in EngineeringVolume 2016, Article ID 1431457, 10 pageshttp://dx.doi.org/10.1155/2016/1431457
2 Mathematical Problems in Engineering
The identification of critical nodes is an important aspectof the system design and antidisaster ability of urban infras-tructure [11]. Research shows that the failure of the criticalnodes or edges may trigger the disastrous consequences,such as widespread avalanche and complete collapse [12, 13].The research is focused on cascading failure of complexnetwork, which is to find out the critical nodes of the network.Furthermore, an important step in WDN evaluation underseismic scenario is to identify the nodes which influence theserviceability seriously [14, 15].Hence, toWDN, analyzing theserviceability of the system and identifying the critical nodesare crucial in guaranteeing the urban safety.
To analyze the serviceability in WDN, technologies fromcomplex network and system reliability can be used. Theserviceability can be measured in three perspectives: (1)
functionality of structural components under disaster envi-ronment; (2) recognition of cascading failure process; and(3) reliability calculation and further evaluation of the systemserviceability.
This paper studies the serviceability ofWDN in cascadingfailure caused by seismic action. Three factors, that is,the seismic attenuation, cascading failure, and reliability ofwater supply, are taken into consideration. The cascadingfailure process is provided with the balancing of watersupply and demand. The node and edge failures in WDNare interrelated under seismic excitations. The matrix-basedsystem reliability (MSR) method is used to represent thesystem events and calculate the nonfailure probability. Theinfluence of serviceability is evaluated with system reliabilityand cascading failure process. This method is applied toWDN. On the consideration of the antiseismic reliabilityof the single component and the whole system, the paperanalyzes the influence of the network tolerance parameter onsystem serviceability. The method adopted in the paper canhelp the decision-makers to identify the critical nodes, resistthe possible widespread network failure under disasters, andimprove the WDN serviceability.
2. Simulation of the CascadingFailure Process in WDN
2.1. Cascading Failure Model. The load distribution on thenetwork is determined by many factors. The load can bethe material, information, and energy [12]. In many entitynetworks, the load is transmitted along the edges basedon the strategy of shortest path. The nodal capacity is themaximum load that a node can bear. As the goal of systemoptimization is tomaximize the system operation effects withthe minimum cost, it can be assumed that there exists aproportional relation between the bearing capacity and theinitial load:
๐ถ = (1 + ๐ผ) ๐ฟ, (1)
where ๐ถ is the node capacity; ๐ฟ is the node load; ๐ผ is thetolerance parameter, 0 โค ๐ผ โค 1. ๐ผ is a tunable parameter[16]. It gives a way to control the strength of the capacity.
When the load is out of the capacity range, the node losesits functions and triggers the flow redistribution. The flow
redistribution may lead to new failure nodes. This step-by-step process is a cascading failure [17โ19]. Time ๐ก is used todescribe the cascading failure step. In the paper, ๐ก is used foran algorithmic step. ๐ก = 0 indicates no failure in WDN. ๐ก = 1
describes the initial failure. ๐ก = 2, 3, . . . show the cascadingfailure process.
2.2. Cascading Failure Modelling in WDN. Whether a nodein WDN can provide sufficient pressure and flow to the cus-tomers is the basic condition to judge the system operation.The service pressure is defined as the load.The nodal pressureis neither too high nor too low. In this scenario, a water nodeis operational if it can operate effectively and it functionsas intended. It required that the nodal pressure is neitherless than the design lowest nor higher than its capacity. Thehighest capacity is calculated by (1). After the initial node isattacked, if there are other nodesโ loads out of its capacityrange, new failure nodes are generated. The end condition isthat no other node loses its function.
In order to measure the cascading failure in WDN aftera particular node failed, it is assumed that the consequenceof a failure node is completely damaged. Its adjacent edgesquit from the system after the attack. Therefore, the failureis equivalent to deleting the node and its adjacent edgesfrom the water network, whichmeans the networkโs topologychanged. With changes in the WDN topology structure, thewater flow redistributed. Pressure of the operational nodes iscalculated according to the laws of conservation of mass andenergy. The Newton-Raphson method and node equationsare involved to simulate the pressure throughout the network.Figure 1 shows the flowchart of cascading failure simulationin WDN.
3. System Reliability
3.1. Matrix-Based System Reliability Method. In recent year,the MSR method has been developed into an effective toolfor calculating system reliability [20โ22]. The principle ismatrix-based calculation.This method has been proved to becapable of being applied to the components with statisticalcorrelation, series system, and parallel system. Assume that asystem has ๐-independent components. The ๐th componenthas two states, that is, nonfailure or failure. Then, the samplespace can be divided into ๐ = 2
๐ mutually exclusive andcollectively exhaustive (MECE) events which are representedby ๐๐(๐ = 1, 2, . . . , ๐). The system events can be described
by event vector ๐. When ๐๐represents a system event, the
element value of ๐ corresponding to component ๐ is set as 1;otherwise it is set as 0. Set the probability of ๐
๐as ๐๐= ๐(๐
๐).
According to the mutual exclusiveness of ๐๐, ๐(๐ธsys) of ๐ธsys
can be calculated by the sum of ๐๐. Therefore, the probability
of any system event (๐ธsys) can be calculated through the innerproduct of two vectors:
๐ (๐ธsys) = โ๐:๐๐โ๐ธsys
๐๐= ๐๐
๐, (2)
where ๐ contains the probability vector of ๐๐, ๐ = 1, . . . , ๐.
Mathematical Problems in Engineering 3
Compute the nodal load and capacity
Compute the nodal pressure of each operational node
Is there a new failure node?
Select an initial attack node i
Yes
No
Start
t = t + 1
Evaluate all nodes?
End
No
Yes
i = i + 1
Figure 1: The flowchart of cascading failure simulation in WDN.
Let ๐๐ธ denote the event vector of event ๐ธ; the vector of itscomplementary event ๐๐ธ is
๐๐ธ
= 1 โ ๐๐ธ
. (3a)
Then the intersection event and union event of events๐ธ1, ๐ธ2, . . . , ๐ธ
๐are, respectively,
๐๐ธ1โฉโ โ โ โฉ๐ธ๐ = ๐
๐ธ1 . โ ๐๐ธ2 . โ โ โ โ . โ ๐
๐ธ๐ (3b)
๐๐ธ1โชโ โ โ โช๐ธ๐ = 1 โ (1 โ ๐๐ธ1) . โ (1 โ ๐๐ธ2) . โ โ โ โ . โ (1 โ ๐๐ธ๐) , (3c)where โ.โโ represents the multiplication between the ele-ments of the matrix, which can be calculated by MATLAB.Equation (2) can be used to calculate the vector of any systemevent ๐ธsys.
When the components are statistically correlated, (2)is not able to calculate the system probability. Instead, theminimum (maximum) probability of a system event can beestablished by linear programming (LP), so as to get theoptimization boundary of the system event. Its mathematicalexpression is
min (max) ๐๐
๐
s.t. ๐1๐ = ๐1
๐2๐ โฅ ๐2,
(4)
where ๐1and ๐2are probability vectors; ๐
1and ๐2are coeffi-
cient matrixes, whose rows correspond to the event vectorof ๐1and ๐2. The system probability interval obtained by LP
method is the narrowest boundary which can be obtainedfrom the given information [23]. Therefore, the systemprobability obtained byMSRmethod is themost accurate andeffective estimation.
3.2. System Reliability under Seismic and Cascaded Scenario.Assume that the earthquake has caused the failure of a nodeon WDN. The nodal failure can probably trigger cascadingfailure. At this time, the new failures caused by cascadingfailure are independent of the earthquake. The subsystemreliability is measured by the operational nodes.
Use LP of MSR method to calculate the boundary of thesystem reliability because of the following: (1) the failuremode of WDN is internally correlated; a nodal failure caninfluence others; (2) different failure nodes lead to differ-ent failure paths due to the topological relations betweennodes and pressure capacity; and (3) analyzing the reliabilityboundary of the disconnected nodes, Lim and Song [24]calculated the average of the upper and lower boundary.Results showed that the disconnection probability is over-valued when ignoring the spatial correlation. Therefore, LPmethod is adopted to calculate the reliability boundary. Theaverage of the boundary is computed to avoid deviation.Let ๐๐ธ๐ denote the probability of the ๐th node antiseismic
reliability. The objective function of the system reliabilityevent can be calculated as
๐ธsys =
๐พ
โ๐=1
โ๐โ๐ถ๐
๐ธ๐. (5)
The system reliability event in (5) is defined in termsof components ๐ถ
๐, ๐ = 1, . . . , ๐พ, whose joint realizations
constitute realizations of the system state ๐ธsys. ๐ถ๐ denotes theset of components that constitute the ๐th subsystem.The sub-system is a connected network constituted by the remainingoperational nodes after cascading failures. Different nodeslead to different failure process. The subsystem reliabilities
4 Mathematical Problems in Engineering
1 20
(a) Series system
0
1
2
(b) Parallel system
Figure 2: Hypothetical system examples of series and parallel system.
constituted by the operational nodes at each time are alsodifferent. The reason is due to the hydraulic properties of theWDN and the topological structure of the pipe network.
Firstly, from the aspect of hydraulic properties, theWDNis a directed network. The upstream failure leads to thedownstream failure which is directly connected with theupstream one. On the contrary, the downstream nodal failuredoes not result in the failure of the upstream nodes. Theupstream nodes can supply water continually.
Secondly, from the aspect of network topology, whencascading failure occurred, (1) if the failure nodes correspondto the series position, the subsystem reliability increases. Forexample, in Figure 2(a),Node 0 is the supply node of the seriessystem. The reliabilities of the three subnodes are 0.5, 0.3,and 0.3, respectively. Assume that the nodes are independent;the system reliability is ๐ธ
0โฉ ๐ธ1โฉ ๐ธ2
= 0.5 ร 0.3 ร 0.3 =
0.045. If Node 2 at the tail end fails while Node 0 andNode 1 can still supply water (i.e., the pressures within thecapacity), the system reliability increases to ๐ธ
0โฉ ๐ธ1= 0.5 ร
0.3 = 0.15. (2) If the failure nodes correspond to the parallelposition, the subsystem reliability decreases. For example, inFigure 2(b), Node 0 is the supply node of the parallel system.The reliabilities of the three subnodes are 0.5, 0.3, and 0.3,respectively. The system reliability is ๐ธ
0โฉ ๐ธ1โช ๐ธ2
= 0.5 ร
(0.3 + 0.3) = 0.3. If Node 2 fails and it does not affect thesupply of Node 0 and Node 1, the system reliability decreasesto ๐ธ0โฉ ๐ธ1= 0.5 ร 0.3 = 0.15.
3.3. Serviceability. The serviceability of WDN after earth-quake contains two aspects: (1) the simulation of the failurestate and (2) the analysis on the serviceability of the damagednetwork. After analyzing the influence of cascading failureand seismic action on WDN, it is necessary to measure theserviceability of the subsystem. Adachi and Ellingwood [9]developed the serviceability ratio. The serviceability ratioof WDN is related to the connectivity number of waterdistribution nodes. The serviceability ratio is given as
๐ =โ๐
๐=1(๐๐๐๐)
โ๐
๐=1๐๐
, (6)
where ๐ represents the serviceability ratio; ๐ is the weight; ๐is the functionality of the infrastructure;๐ = 1 indicates thatthe functionality of the infrastructure is effective; otherwise๐ = 0; ๐ is the number of nodes in WDN.
The serviceability of WDN after earthquake should con-sider (1) the influence of seismic action on work nodes;(2) failure state caused by cascading failures; and (3) the
serviceability of subsystem, comprehensively. Therefore, theaseismic reliability factor of the subsystem is taken intoconsideration based on (6). Equation (6) can be improvedinto (7). Equation (7) describes the serviceability of WDNafter earthquake:
๐ =โ๐
๐=1(๐๐๐๐)
โ๐
๐=1๐๐
๐ (๐ธ๐ก
sys) , (7)
where ๐(๐ธ๐กsys) is the system reliability event constituted bythe work nodes at time ๐ก. The smaller the computed result ofโ๐
๐=1๐๐๐๐is, the fewer the function-effective infrastructures
will be. The smaller the value of ๐(๐ธ๐กsys) is, the lower thereliability probability of the system will be. As a result, ๐
is ranging between 0 and 1; the smaller ๐ is, the greaterthe serviceability loss caused by the cascading failure afterearthquake will be.
4. Data and Application Results
Theproposedmethod is illustrated by an example.TheWDN[25] is shown in Figure 3.The pipe information includes pipenumber, flow direction, pipe diameter, and pipe length. Thenode information includes node number, nodal demand, andground elevation. Assume the minimum pressure is 6m toensure the basic operation.
The overall procedure is as follows:
Step 1 (network information)
(1) Provide WDN topology structure.(2) Provide WDN hydraulic features.(3) Calculate nodal load and capacity.
Step 2 (hazard information)
(1) Assess the ground motion parameters of WDN com-ponents under seismic scenario.
(2) Calculate the failure probability of each component.
Step 3 (cascading failure)
(1) Take one node in WDN as the initial failure node.Simulate the network topology after flow redistribu-tion.
(2) If no new failure nodes are found, the cascading fail-ure process stops.
Mathematical Problems in Engineering 5
2(300)
1700
3(300)
1000 4(250)
900
8(200)
14007(250)
1100
6(300)
900
9(200)
900
1(350)
1400
5(200)
1350
11(150)
1200
10(200)
1000
12(300)
1100
13(200)
800
17(250)
1200
18(200)
800
19(150)
900 20(200)
1400
21(150)
1200
16(200)
110015(250)
800
14(200)
1400
1โ165.51
180
26.94178
311.57179
410.42180
79.26182
89.26181
513.89181
1013.89182
610.42183
129.26181
159.26179
166.94180
1313.89183
913.89180
1710.42181
149.2618411
6.94181
abc
Junction numberJunction demand (L/s)Ground elevation (m)
Pipeline number(dia in mm) Length in m
30M
Figure 3: The layout of the WDN.
(3) Simulate the cascading failure process of each nodesuccessively. Conduct the system event vector.
Step 4 (serviceability)
(1) Calculate the system reliability according to the sys-tem event vector and failure probability.
(2) Evaluate the serviceability ratio using system reliabil-ity.
4.1. Hydraulic Analysis. Each operational node in WDNutilizes the laws of conservation of mass and energy todetermine pressure and flow distribution throughout thenetwork [26]. Formulations for water system are describedas follows:
(1) For a junction node,
โ๐in โ โ๐out = ๐ext, (8)
where ๐in and ๐out are the inflow and outflow of the node;๐ext is the external demand or supply.
(2) For any path,
โ๐โ๐ผ๐
โ๐ฟ,๐
+ โ๐โ๐ฝ๐
โ๐,๐
= ฮ๐ธ, (9)
where โ๐ฟ,๐
is the head loss across component ๐ along the path,โ๐,๐
is the head added by pump ๐, and ฮ๐ธ is the difference inenergy between the end points of the path.
The friction (โ๐ฟ) in each pipe is calculated using the
Hazen-Williams equations, given as
โ๐ฟ= 10.654 (
๐
๐ถ)1.852 1
๐ท4.87๐ฟ, (10)
where ๐ is the flow rate; ๐ถ is the Hazen-Williams coefficient;๐ท is the pipe diameter; and ๐ฟ is the pipe length.
(3) For closed loops that begin and end at the same point(ฮ๐ธ = 0),
โ๐โ๐ผ๐ฟ
โ๐ฟ,๐
= 0. (11)
4.2. Seismic Hazard. In order to measure the vulnerabilityof WDN components under seismic action, the paper takesa single seismic origin as the case and the earthquakemagnitude is set as 7.0.The earthquake magnitude representsthe strength of the earthquake. It is related to the energyreleased by the earthquake. The greater the release of energyis, the greater the magnitude of the earthquake will be.The peak ground acceleration (PGA) is used to propagatethe earthquake intensity. PGA is equal to the maximumground acceleration that occurred at a site during a particularearthquake. As the selection of attenuation function is notin the scope of this paper, the seismic intensity attenuationrelation proposed by Toro et al. [27] is applied to describe thevariation of PGA:
ln (PGA) = 2.20 + 0.81 โ (๐๐ค
โ 6.0) โ 1.27
โ ln (โ๐ 2 + 9.32) โ (1.16 โ 1.27)
โ max[ln(โ๐ 2 + 9.32
100) , 0.0]
โ 0.0021โ๐ 2 + 9.32 + ln (1.52) ,
(12)
where ๐๐ค
is the earthquake magnitude. The earthquakemagnitude is assumed to 7.0. ๐ is the distance between thesite of interest (water node) and the epicenter. The distancesof Nodes 1โ17 in the network are 37.189, 37.804, 38.079, 38.075,37.487, 37.693, 37.350, 37.323, 36.864, 36.666, 36.592, 36.754,36.560, 35.567, 35.583, 35.881, and 35.531 km, respectively.
6 Mathematical Problems in Engineering
Table 1: The cascading failure processes (๐ผ = 0.5 and ๐ผ = 1.0).
Node Failure process (๐ผ = 0.5) Failure process (๐ผ = 1.0)1 All nodes failed All nodes failed2 2 โ (3โ10, 13, 14, 17) โ (12, 16) 2 โ (3โ10, 13, 14, 17) โ (16)3 3 โ (4, 5, 6, 9, 10, 13, 14, 17) โ (8, 12, 16) 3 โ (4, 5, 6, 9, 10, 13, 14, 17) โ (16)4 4 โ (3, 5, 6, 9, 10, 13, 14, 17) โ (8, 12, 16) 4 โ (3, 5, 6, 9, 10, 13, 14, 17) โ (16)5 5 โ (4, 6, 8, 9, 10, 13, 14, 17) โ (12, 16) 5 โ (4, 6, 9, 10, 13, 14, 17) โ (3, 16)6 6 โ (5, 9, 10, 14, 17) โ (4, 8, 13, 16) โ (12) 6 โ no more failure nodes7 7 โ (8, 14) โ (5, 6, 9, 10, 13, 16, 17) โ (4, 12) 7 โ (8, 14) โ (6, 10, 17)8 8 โ (7, 14) โ (5, 6, 9, 10, 13, 16, 17) โ (4, 12) 8 โ (7, 14) โ (6, 10, 17)9 9 โ (5, 6, 10, 13, 14, 16, 17) โ (4, 8, 12) 9 โ (6, 14) โ (5, 10, 17) โ (4) โ (3)10 10 โ (5, 6, 9, 14) โ (4, 8, 13, 16, 17) โ (12) 10 โ (6, 14) โ (17)11 11 โ (6, 10, 12โ17) โ (4, 5, 9) โ (3, 8) 11 โ (6, 10, 12โ17)12 12 โ (13, 14) โ (5, 6, 9, 10, 16, 17) โ (4, 8) 12 โ (13, 14) โ (6, 10, 17) โ (16)13 13 โ (12, 14, 16, 17) โ (5, 6, 9, 10) โ (4, 8) 13 โ (12)14 14 โ (5, 6, 9, 10, 13, 16, 17) โ (4, 8, 12) 14 โ (6, 10, 17)15 15 โ (14, 16, 17) โ (5, 6, 9, 10, 13) โ (4, 8, 12) 15 โ (14, 16, 17) โ (6, 10)16 16 โ (14, 15, 17) โ (5, 6, 9, 10, 13) โ (4, 8, 12) 16 โ (14, 15, 17) โ (6, 10)17 17 โ (13, 16) โ (6, 10, 12, 14) โ (4, 5, 9) โ (3, 8) 17 โ no more failure nodes
1
2
11
3
7
12
15 16
(a) ๐ก = 2
1
2
11
3
7
15
(b) ๐ก = 3
Figure 4: The state of WDN after Node 5 failed.
Structural fragility curves are essential for evaluationof the functionality of water distribution component underearthquake impact. The structural fragility in Adachi andEllingwood [28] is used. The fragility of each componentin the WDN is described by the lognormal cumulativedistribution function (CDF):
๐๐๐ฅ(๐ฅ) = โซ
๐ฅ
0
1
โ2๐๐๐ exp[โ
1
2{ln (๐ ) โ ๐
๐}
2
]๐๐ , (13)
where ๐ is the ground motion intensity and ๐ and ๐ are themedian and standard deviation of the lognormal distribution,respectively.The fragilities are defined bymedian of 0.4 g andlogarithmic standard deviation of 0.4 [29].
4.3. Cascading Failure Process. The cascading failure pro-cesses ofWDNhave been simulatedwith๐ผ = 0.5 and๐ผ = 1.0.The failure processes are shown in Table 1. It can be foundthat the number of secondary failure nodes has a significantreduction when the constraint condition of node bearingcapacity is relaxed. As ๐ผ = 0.5 and 1.0, the numbers of failurenodes among Nodes 2, 3, 4, 7, 8, 11, 12, 15, and 16 at ๐ก = 2 are
the same; as ๐ก โฅ 3, the failure scale of ๐ผ = 0.5 is larger thanthat of ๐ผ = 1.0. For other nodes, for example, Nodes 5, 6, 7,10, 13, 14, and 17, the failure scale of ๐ผ = 0.5 is larger than thatof ๐ผ = 1.0 at ๐ก = 2. This is obvious for Node 6 and Node17: as ๐ผ = 0.5, the nodal failure of Node 6 and Node 17 leadsto the extra 10 and 11 failure nodes, respectively; as ๐ผ = 1.0,the nodal failure of Node 6 and Node 17 cannot trigger thecascading failure. The cascading failure simulation is used togenerate the connected link, which can be used to conductthe system event vector.
4.4. System Reliability. Take Node 5 (๐ผ = 0.5) as an example.Figure 4 shows the connected state at ๐ก = 2 and ๐ก = 3. Node5 is damaged under seismic action and its failure process canbe divided into three stages. At ๐ก = 1, Node 5 fails under theearthquake and its failure leads to the failure of its connectedpipes which further results in the redistribution of WDN; at๐ก = 2, Nodes 4, 6, 8, 9, 10, 13, 14, and 17 fail successively; at ๐ก =
3, Nodes 12 and 16 fail; then the system reaches a new stablestate; that is, no new failure node occurs. The effective eventvectors of supply node ๐
๐ธ๐ (๐ = 1, 2, . . . , 17) are constructedby (3a), (3b), (3c), and (5).
Mathematical Problems in Engineering 7
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
t = 2 t = 3
t = 4 t = 0
Node number
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9Sy
stem
relia
bilit
y
(a) ๐ผ = 0.5
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
t = 2 t = 3
t = 4 t = 0
Node number
Syste
m re
liabi
lity
(b) ๐ผ = 1.0
Figure 5: Computed results of the system reliability at each time.
At ๐ก = 2, Nodes 4, 6, 8, 9, 10, 13, 14, and 17 fail and theremaining connected nodes are Nodes 1, 2, 3, 7, 11, 12, 15, and16. The system event vector can be expressed as
๐ธ5-IIsys = ๐ธ
1โฉ {[๐ธ2โฉ (๐ธ3โช ๐ธ7)]
โช [๐ธ11
โฉ (๐ธ12
โช (๐ธ15
โฉ ๐ธ16))]} .
(14)
At ๐ก = 3, Nodes 12 and 16 fail. The remaining connectednodes are Nodes 1, 2, 3, 7, 11, and 15. The system event vectorcan be expressed as
๐ธ5-IIIsys = ๐ธ
1โฉ {[๐ธ2โฉ (๐ธ3โช ๐ธ7)] โช [๐ธ
11โฉ ๐ธ15
โฉ ๐ธ16]} . (15)
Similarly, by the change of network topology after thecascading failure occurred, we can construct the subsystemevent vector constituted by the remaining work nodes at eachtime. After the supply Node 1 fails, Nodes 2โ17 are bound tofail in the absence of other supplies, so the supply node isthe most critical node. Here, we focus on Nodes 2โ17 in theinfluence of cascading failure and seismic action.
Simulations are made successively with Nodes 2โ17 fail-ure. Construct the connected graph of the remaining worknodes at each time. Calculate the functionality of the systemunder seismic action by (13). The subsystem event is consti-tuted by the remaining work nodes after cascading failureoccurs. Different failure nodes lead to different remainingnodes which constitute different subsystems. Consider thespatial correlation of the nodes; use LP method (4) to cal-culate the boundary of system reliability probability. Definethe average of the supremum and infimum boundary as thesystem reliability.
Figure 5 shows the trendline of the system reliabilityat different time. In Figure 5, the purple line representsthe reliability of the single node under seismic action. Theblue line, red line, and green line represent the subsystemreliability at ๐ก = 2, 3, and 4, respectively.The system reliability
is influenced by three aspects: (1) the distance from networkcomponents to the epicenter; (2) the functionality of nodes;and (3) the change of the network topology.The change of thenetwork topology is obtained from the constraints conditionsof load and capacity. First of all, it can be seen that the systemreliability is lower than that of the components. Secondly, asshown in Figure 5(a), the fluctuation of system reliability isobvious; this is mainly because the small value of ๐ผ and thelow upper-limit of bearing capacity lead to the enlargement ofunconnected scope. Thirdly, when ๐ผ = 0.5, Node 2โs systemreliability reduces at ๐ก = 3 which means Node 2 correspondsto the parallel position. Other nodes at each time are close tothe series position with an increased system reliability. When๐ผ = 1.0, Node 5 at ๐ก = 3 and Node 9 at ๐ก = 4 are close to theparallel position. Other nodes at each time are close to theseries position.
4.5. Serviceability. Based on the efficiency analysis of thesubsystem under seismic action, the serviceability ofWDN isanalyzed also. Consider the weight factor๐ as the pipe length๐ฟ. The computed result is shown in Figures 6 and 7.
Figure 6 illustrates the serviceability ratio at each timestep under different tolerance parameters (๐ผ = 0.5 and ๐ผ =
1.0). In Figure 6, the blue bar, red bar, and green bar representthe serviceability ratio at ๐ก = 2, 3, and 4, respectively. Asthe process of cascading failure, earthquake, and servicearea have been taken into consideration, the result showsthat, with the progress of cascading failure, the remainingserviceability is gradually declining while the service failureis increasing. There is a great difference between ๐ก = 2 and๐ก = 3, indicating that the first secondary failures causedby the failure of initial node lead to a significant declineof serviceability. Therefore, it is necessary to control thecascading failure between ๐ก = 2 and ๐ก = 3.
Specific to ๐ก = 2 and ๐ก = 3, Figure 7 shows the differenceon serviceability ratio. In Figure 7, the blue line and the red
8 Mathematical Problems in Engineering
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17Node number
t = 4 t = 3 t = 2
Serv
icea
bilit
y ra
tio
(a) ๐ผ = 0.5
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17Node number
t = 4 t = 3 t = 2
Serv
icea
bilit
y ra
tio
(b) ๐ผ = 1.0
Figure 6: Result of the serviceability ratio (๐๐ค
= 7; ๐ผ = 0.5, 1.0).
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
0.5
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
๐ผ = 0.5๐ผ = 1.0
Node number
Serv
icea
bilit
y ra
tio
(a) ๐ก = 2
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
๐ผ = 0.5๐ผ = 1.0
Node number
Serv
icea
bilit
y ra
tio
(b) ๐ก = 3
Figure 7: Serviceability ratio with ๐ก = 2 and ๐ก = 3 (๐ผ = 0.5; ๐ผ = 1.0).
line represent the serviceability ratio under ๐ผ = 0.5 and๐ผ = 1.0, respectively. The serviceability when ๐ผ = 1.0 ishigher than or equal to those when ๐ผ = 0.5. This is mainlybecause the maximum pressure constraint is more relaxed as๐ผ = 1.0. It should be noted that the serviceability ratio ofNode 2 remains unchanged. In Figure 7(a), Nodes 2, 3, 4, 7,8, and 12 have the same serviceability ratio at ๐ก = 2 under๐ผ = 0.5 and ๐ผ = 1.0. In Figure 7(b), Nodes 2 and 5 have thesame serviceability ratio at ๐ก = 3 under ๐ผ = 0.5 and ๐ผ = 1.0.The intersection of Figures 7(a) and 7(b) is Node 2. It is thusclear that the serviceability of Node 2 under any toleranceparameter always is the minimum one, which indicates thatNode 2 is the most critical node. A backup device is neededto Node 2 to enhance the antidisaster ability of the system.
5. Conclusion
To evaluate the serviceability of WDN under seismic sce-nario, the component functionality and the cascading failuresare taken into consideration. The serviceability is measuredin three aspects: (1) functionality of structural componentsunder disaster environment; (2) recognition of cascadingfailure process, that is, the connectivity loss between node-links; and (3) calculation of the subsystem reliability. Resultsshow that the topological position of the failure nodes canbe measure by the subsystem reliability. The critical nodecan be identified. Secondly, under different values of ๐ผ, theserviceability of the subsystem presents a great differencewhich can be attributed to the different failure processes
Mathematical Problems in Engineering 9
under cascade effect. Thirdly, it is necessary to control thecascading failure between ๐ก = 2 and ๐ก = 3. Conductingtimely repair and maintenance to WDN before and duringthis period can enhance the antidisaster ability of WDN.
Further research should be focused on the time. Time๐ก is used as an algorithmic step. Future works should focuson the real meaning of time; that is, the cascading failurestep can be combined with flow velocity to express the hours,minutes, and seconds. Also, a time constraint can be addedin the model to evaluate the dynamic relationship on theserviceability, the failure time, and the repair time. Besides,there is a reduction level of water nodes after the damage.Thereduction level of service should be extended in the model toobtain a more realistic result.
Competing Interests
The authors declare that they have no competing interests.
Acknowledgments
This work was funded by the National Natural Science Foun-dation of China under Grant no. 71501008, the FundamentalFunds for Humanities and Social Sciences of Beijing JiaotongUniversity (no. 2015jbwj014), and the Fundamental Funds ofBeijing Jiaotong University under Grant (no. 2015RC002).
References
[1] A. Yazdani and P. Jeffrey, โComplex network analysis of waterdistribution systems,โ Chaos, vol. 21, Article ID 016111, 2011.
[2] J. Yates and S. Sanjeevi, โAssessing the impact of vulnerabilitymodeling in the protection of critical infrastructure,โ Journal ofGeographical Systems, vol. 14, no. 4, pp. 415โ435, 2012.
[3] T. Adachi, Impact of cascading failures on performance assess-ment of civil infrastructure systems [Ph.D. thesis], GeorgiaInstitute of Technology, Atlanta, Ga, USA, 2007.
[4] L. Duenas-Osorio and S. M. Vemuru, โCascading failures incomplex infrastructure systems,โ Structural Safety, vol. 31, no.2, pp. 157โ167, 2009.
[5] C.-L. Chai, X. Liu, W. J. Zhang, and Z. Baber, โApplication ofsocial network theory to prioritizing Oil & Gas industries pro-tection in a networked critical infrastructure system,โ Journal ofLoss Prevention in the Process Industries, vol. 24, no. 5, pp. 688โ694, 2011.
[6] K. J. Tierney, S. E. Chang, and R. T. Eguchi, Improving Earth-quake Loss Estimation: Review, Assessment and Extension ofLoss EstimationMethodologies, University of Delaware. DisasterResearch Center, 1999.
[7] M. Shinozuka, H. Hwang, and M. Murata, โImpact on watersupply of a seismically damaged water delivery system,โ inProceeding of the Lifeline Earthquake Engineering in the Centraland Eastern US, pp. 43โ57, ASCE, September 1992.
[8] H. H. M. Hwang, H. Lin, and M. Shinozuka, โSeismic per-formance assessment of water delivery systems,โ Journal ofInfrastructure Systems, vol. 4, no. 3, pp. 118โ125, 1998.
[9] T. Adachi and B. R. Ellingwood, โServiceability of earthquake-damaged water systems: effects of electrical power availabilityand power backup systems on system vulnerability,โ ReliabilityEngineering & System Safety, vol. 93, no. 1, pp. 78โ88, 2008.
[10] S. A. Brink, R. A. Davidson, and T. H. P. Tabucchi, โStrategiesto reduce durations of post-earthquake water service interrup-tions in Los Angeles,โ Structure and Infrastructure Engineering,vol. 8, no. 2, pp. 199โ210, 2012.
[11] T. H. Grubesic and A. T. Murray, โVital nodes, interconnectedinfrastructures, and the geographies of network survivability,โAnnals of the Association of American Geographers, vol. 96, no.1, pp. 64โ83, 2006.
[12] A. E. Motter and Y.-C. Lai, โCascade-based attacks on complexnetworks,โ Physical Review E, vol. 66, Article ID 065102, 2002.
[13] P. Crucitti, V. Latora, and M. Marchiori, โModel for cascadingfailures in complex networks,โ Physical Review E-Statistical,Nonlinear, and Soft Matter Physics, vol. 69, no. 4, Article ID045104, 2004.
[14] Y. Wang and S.-K. Au, โSpatial distribution of water supplyreliability and critical links of water supply to crucial waterconsumers under an earthquake,โ Reliability Engineering &System Safety, vol. 94, no. 2, pp. 534โ541, 2009.
[15] J. Nigg, โPerceptions of earthquake impacts and loss-reductionpolicy preferences among community residents and optionleaders,โ Disaster Research Center, University of Delaware,1998.
[16] J.-W.Wang and L.-L. Rong, โCascade-based attack vulnerabilityon the US power grid,โ Safety Science, vol. 47, no. 10, pp. 1332โ1336, 2009.
[17] J. Wang, C. Jiang, and J. Qian, โRobustness of Internet undertargeted attack: a cascading failure perspective,โ Journal ofNetwork and Computer Applications, vol. 40, no. 1, pp. 97โ104,2014.
[18] J. Wang, โRobustness of complex networks with the localprotection strategy against cascading failures,โ Safety Science,vol. 53, pp. 219โ225, 2013.
[19] J.-W. Wang and L.-L. Rong, โRobustness of the western UnitedStates power grid under edge attack strategies due to cascadingfailures,โ Safety Science, vol. 49, no. 6, pp. 807โ812, 2011.
[20] J. Song and A. D. Kiureghian, โBounds on system reliability bylinear programming,โ Journal of EngineeringMechanics, vol. 129,no. 6, pp. 627โ636, 2003.
[21] W.-H. Kang, J. Song, and P. Gardoni, โMatrix-based system reli-ability method and applications to bridge networks,โ ReliabilityEngineering & System Safety, vol. 93, no. 11, pp. 1584โ1593, 2008.
[22] J. Song and W.-H. Kang, โSystem reliability and sensitivityunder statistical dependence by matrix-based system reliabilitymethod,โ Structural Safety, vol. 31, no. 2, pp. 148โ156, 2009.
[23] D. Bertsimas and J. Tsitsiklis, Introduction to Linear Optimiza-tion, Athena Scientific Press, 1st edition, 1997.
[24] H.-W. Lim and J. Song, โEfficient risk assessment of lifeline net-works under spatially correlated groundmotions using selectiverecursive decomposition algorithm,โ Earthquake Engineering &Structural Dynamics, vol. 41, no. 13, pp. 1861โ1882, 2012.
[25] M. L. Kansal, A. Kumar, and P. B. Sharma, โReliability analysisof water distribution systems under uncertainty,โ ReliabilityEngineering and System Safety, vol. 50, no. 1, pp. 51โ59, 1995.
[26] L.W.Mays,Water Distribution SystemHandbook, McGraw-HillProfessional Publishing, New York, NY, USA, 1999.
[27] G. R. Toro, N. A. Abrahamson, and J. F. Schneider, โModel ofstrong groundmotions from earthquakes in central and easternNorth America: best estimates and uncertainties,โ SeismologicalResearch Letters, vol. 68, no. 1, pp. 41โ57, 1997.
10 Mathematical Problems in Engineering
[28] T. Adachi and B. R. Ellingwood, โComparative assessment ofcivil infrastructure network performance under probabilisticand scenario earthquakes,โ Journal of Infrastructure Systems, vol.16, no. 1, Article ID 010001QIS, pp. 1โ10, 2010.
[29] FEMA, โHAZUS MH MR1,โ Technical Manual, Washington,DC, USA, 2003.
Submit your manuscripts athttp://www.hindawi.com
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
MathematicsJournal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Mathematical Problems in Engineering
Hindawi Publishing Corporationhttp://www.hindawi.com
Differential EquationsInternational Journal of
Volume 2014
Applied MathematicsJournal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Probability and StatisticsHindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Journal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Mathematical PhysicsAdvances in
Complex AnalysisJournal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
OptimizationJournal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
CombinatoricsHindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
International Journal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Operations ResearchAdvances in
Journal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Function Spaces
Abstract and Applied AnalysisHindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
International Journal of Mathematics and Mathematical Sciences
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
The Scientific World JournalHindawi Publishing Corporation http://www.hindawi.com Volume 2014
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Algebra
Discrete Dynamics in Nature and Society
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Decision SciencesAdvances in
Discrete MathematicsJournal of
Hindawi Publishing Corporationhttp://www.hindawi.com
Volume 2014 Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Stochastic AnalysisInternational Journal of