1. Integrated Reliability Assessment of Water
Distribution Networks
Presented by:
Azhar Uddin Mohammed
Supervisors:
Dr. Tarek Zayed, Dr. Osama Moselhi
January 26, 2016
8. Introduction Literature
Review
Research
Methodology
Data
Collection
Model
Implementation
Conclusion
8
Objectives
Develop a comprehensive reliability assessment model that
can identify the crucial water mains and prioritize their renewal
Identify and study the factors impacting mechanical and hydraulic
reliability at various hierarchical levels
Develop mechanical and hydraulic models to assess network reliability
Develop an integrated reliability assessment model considering
mechanical and hydraulic aspects
Semi-Automate the developed model using coded scripts
13. Introduction
Literature
Review
Research
Methodology
Data
Collection
Model
Implementation
Conclusion
13
Network Reliability Analysis
Hydraulic Network Analysis
Demand Driven Analysis (DDA)
• Surpass demand driven analysis, particularly for networks under abnormal operating conditions
• Considers the pressure dependency of nodal outflows, and in consequence, the results are more
realistic
• Objective of PDA is to establish the actual supply quantity and pressure at each node in a WDN
• Assumes that consumer demands are always satisfied regardless of the pressures
• Acceptable results when WDNs are subject to normal operating conditions
• When the pressure drops below the required level, no information on how much water would be
delivered by the system
• Results indicate the system is still supplying the full demand at lower, and sometimes, negative
pressures
Pressure Driven Analysis (PDA)
16. Introduction
Literature
Review
Research
Methodology
Data
Collection
Model
Implementation
Conclusion
16
Water utilities usually ignore and do not record the causes of failure
Reliability techniques were either applied to small scale networks or assessed
mechanical and hydraulic reliability independently
Hydraulic models were found to be demand driven
Pressure driven models were analyzed manually rather than analyzing in a
simulating environment
Limitations of Previous Studies
17. Introduction
Literature
Review
Research
Methodology
Data
Collection
Model
Implementation
Conclusion
17
Research Methodology
Failure Rate
Pipes Accessories
Component
Reliability
Segment
Reliability
Network
Minimum Cut
Set Analysis for
allNodes
Record Minimum
Cut Sets
Integrated Reliability
Assessment of WDN
Literature
Review
Failure Probability
Models
Reliability Assessment
Methods
Reliability
Classification
Failure Causes and
Consequences
Data Collection
London Doha
LiteratureReview Best Practices
Run Hydraulic
Simulation
Normal Condition
Failure Condition
(For all cut sets)
Available Demand
Bentley WaterCAD Library
Required Demand
Service
Pressure Head
Available
Pressure Head
19. Minimum
Cut Set
Analysis
Introduction
Literature
Review
Research
Methodology
Data
Collection
Model
Implementation
Conclusion
19
Failure Rate
Pipes Accessories
Component
Reliability
Segment
Reliability
Network
Mechanical
Reliability
Minimum Cut
Set Analysis for
allNodes
Record Minimum
Cut Sets
Path
Matrix
• Find all possible paths from source node to destination node
• Create path matrix
First Order
Cut Sets
• From the path matrix, check if any column is non zero
• Any non zero column is a first order cut set
Second
Order Cut
Sets
• Combine any two columns representing segments in a path matrix
and check if their addition creates a non zero column
• Resultant non zero is a second order cut set
26. Introduction
Literature
Review
Research
Methodology
Data
Collection
Model
Implementation
Conclusion
26
Recorded
Minimum Cut
Set(s)
Run Hydraulic
Simulation
Normal Condition
Failure Condition
(For all cut sets)
Available Demand
Network Hydraulic
Reliability
Nodal Demand Allocation
Import Shape Files of
Selected Sub-Network
Required Demand
Service
Pressure Head
Available
Pressure Head
Normal Condition
Failure Condition
(For all cut sets)
Hydraulic Simulation
AllSegments
are Functional
Run Hydraulic
Simulation
Required Pressure Head
Hreq
Recorded Cut Set(s) from
Mechanical Reliability Model
Recorded Cut Set(s) from
Hydraulic Reliability Model
Segment(s) are
Non-Functional
Run Hydraulic
Simulation
Available Pressure Head
Havl
Nodal Demand Allocation
Lreq
27. Introduction
Literature
Review
Research
Methodology
Data
Collection
Model
Implementation
Conclusion
27
Recorded
Minimum Cut
Set(s)
Run Hydraulic
Simulation
Normal Condition
Failure Condition
(For all cut sets)
Available Demand
Network Hydraulic
Reliability
Nodal Demand Allocation
Import Shape Files of
Selected Sub-Network
Required Demand
Service
Pressure Head
Available
Pressure Head
𝐋 𝐚𝐯𝐥 = 𝟎
if 𝐇 𝐚𝐯𝐥 < 𝐇 𝐦𝐢𝐧
Hreq = Required pressure head at a node (m)
Havl = Available pressure head at a node (m)
Hmin & Hmax = Minimum and maximum
pressure head at a node respectively (m)
Lreq = Required demand at a node (m3/day)
Lavl = Available demand at a node (m3/day)
𝐋 𝐚𝐯𝐥 = 𝐋 𝐫𝐞𝐪
𝐇 𝐚𝐯𝐥−𝐇 𝐦𝐢𝐧
𝐇 𝐫𝐞𝐪−𝐇 𝐦𝐢𝐧
if 𝐇 𝐦𝐢𝐧 ≤ 𝐇 𝐚𝐯𝐥 ≤ 𝐇 𝐫𝐞𝐪
𝐋 𝐚𝐯𝐥 = 𝐋 𝐫𝐞𝐪
if 𝐇 𝐫𝐞𝐪 < 𝐇 𝐚𝐯𝐥 ≤ 𝐇 𝐦𝐚𝐱
𝐋 𝐚𝐯𝐥 = 𝟎
if 𝐇 𝐚𝐯𝐥 > 𝐇 𝐦𝐚𝐱
41. Introduction
Literature
Review
Research
Methodology
Data
Collection
Model
Implementation
Conclusion
41
City of London
Order of Cut Sets List of Cut Sets
1 {a}
2 {b,c},{d,g},{s,v}
3
{b,d,f}, {b,e,i}, {b,f,g}, {c,d,f}, {c,e,i}, {c,f,g},
{d,h,k}, {g,h,k}, {i,j,k}, {i,l,m}, {k,n,p}, {m,q,u},
{o,q,t}, {p,t,u}, {r,s,u}, {r,u,v}
Case Studies
London Doha
South Phase
North Phase
Mechanical Reliability Model
Hydraulic Reliability Model
Integrated Network
Reliability
Applying the principle of minimum cut set method, the
mechanical reliability of the north phase sub-network is
calculated to be 0.94
43. Introduction
Literature
Review
Research
Methodology
Data
Collection
Model
Implementation
Conclusion
43
City of London
Case Studies
London Doha
South Phase
North Phase
Mechanical Reliability Model
Hydraulic Reliability Model
Integrated Network
Reliability
Demand
Node
GIS ID Demand (m³/day) Pressure Head (m)
8 N4210 114 34.86
1 N5775 77.5 38.1
4 N5794 93 35.71
3 N5795 93 36
2 N5796 114 38.5
6 N5992 70 35.65
5 N6088 93 32.38
7 N6091 93 35.88
12 N6094 152 34.94
9 N6096 93 35.85
10 N6098 84 34.65
13 N6100 77.5 34.45
11 N6108 95 36.5
14 N6130 46.5 36.7
15 N6131 77.5 36.39
Service Pressure Heads
for Demand Nodes of
North Phase Sub-Network
44. Introduction
Literature
Review
Research
Methodology
Data
Collection
Model
Implementation
Conclusion
44
City of London
Demand Node a closed b + c closed d + g closed s + v closed
8 0 32.73 34.59 34.86
1 38.1 38.1 38.1 38.1
4 0 32.69 35.94 35.71
3 0 0 36.16 36
2 0 38.5 38.5 38.5
6 0 33.31 35.7 35.65
5 0 29.71 0 32.38
7 0 33.56 35.56 35.88
12 0 32.7 34.66 34.94
9 0 33.61 35.57 35.85
10 0 32.35 34.34 34.65
13 0 32.21 34.16 34.45
11 0 34.27 36.21 36.5
14 0 34.47 36.41 36.7
15 0 34.17 36.11 0
Available Pressure Heads
at Demand Nodes of
North Phase Sub-Network
45. Introduction
Literature
Review
Research
Methodology
Data
Collection
Model
Implementation
Conclusion
45
City of London
Available Nodal Demands
for North Phase Sub-
Network
Demand Node a closed b + c closed d + g closed s + v closed
8 0 94.66 111.73 114.00
1 77.5 77.50 77.50 77.50
4 0 72.53 93.00 93.00
3 0 0.00 93.00 93.00
2 0 114.00 114.00 114.00
6 0 58.32 70.00 70.00
5 0 58.11 0.00 93.00
7 0 78.12 91.09 93.00
12 0 125.09 148.90 152.00
9 0 78.62 91.33 93.00
10 0 67.94 82.02 84.00
13 0 62.61 75.74 77.50
11 0 81.59 93.37 95.00
14 0 40.10 45.72 46.50
15 0 66.46 76.20 0.00
46. Introduction
Literature
Review
Research
Methodology
Data
Collection
Model
Implementation
Conclusion
46
City of London
After finding the available
demands at all nodes, the
hydraulic reliability of the selected
north phase sub-network is
calculated is found to be equal to
Demand Node a closed b + c closed d + g closed s + v closed
8 0 94.66 111.73 114.00
1 77.5 77.50 77.50 77.50
4 0 72.53 93.00 93.00
3 0 0.00 93.00 93.00
2 0 114.00 114.00 114.00
6 0 58.32 70.00 70.00
5 0 58.11 0.00 93.00
7 0 78.12 91.09 93.00
12 0 125.09 148.90 152.00
9 0 78.62 91.33 93.00
10 0 67.94 82.02 84.00
13 0 62.61 75.74 77.50
11 0 81.59 93.37 95.00
14 0 40.10 45.72 46.50
15 0 66.46 76.20 0.00
0.7062
47. Introduction
Literature
Review
Research
Methodology
Data
Collection
Model
Implementation
Conclusion
47
City of London
Case Studies
London Doha
South Phase
North Phase
Mechanical Reliability Model
Hydraulic Reliability Model
Integrated Network
Reliability
40
15
13 13
30
21
0
5
10
15
20
25
30
35
40
45
Corrosion Permeation Pipe
Deterioration
Pipe
Deterioration
Hydraulic
Changes during
Maintenance and
Emergencies
Tuberculation
Mechanical Hydraulic
No.ofFailures
Failure Causes
51.51% 48.49%
The integrated reliability of the north phase sub-
network is assessed to be equal to 0.8255
49. Introduction
Literature
Review
Research
Methodology
Data
Collection
Model
Implementation
Conclusion
49
Comparison of Results
The segment connecting the source of water to the network is the most
crucial segment of the network
Although all the minimum cut sets are crucial segments in a network, few
among them can be termed ‘more crucial’ than others
0.965 0.957 0.938
0.706
0.826
0.990 0.988 0.968
0.696
0.845
0.000
0.200
0.400
0.600
0.800
1.000
Average Component
Reliability
Average Segment
Reliability
Mechanical Reliability
of Sub-Network
Hydraulic Reliability of
Sub-Network
Integrated Network
Reliability
Reliability Comparison of Results
North Phase South Phase
50. Introduction
Literature
Review
Research
Methodology
Data
Collection
Model
Implementation
Conclusion
50
Comparison of Results
0.965 0.957 0.938
0.706
0.826
0.990 0.988 0.968
0.696
0.845
0.000
0.200
0.400
0.600
0.800
1.000
Average Component
Reliability
Average Segment
Reliability
Mechanical Reliability
of Sub-Network
Hydraulic Reliability of
Sub-Network
Integrated Network
Reliability
Reliability Comparison of Results
North Phase South Phase
35.65
38.1
32.38
34.86
39.56
43.06
40.86
43.29
0
5
10
15
20
25
30
35
40
45
50
70 77.5 93 114
Pressure
m
Demand
m3/day
Required Pressure
North Phase vs South Phase
North Phase South Phase
South phase sub-network requires higher pressure than the
north phase, to supply equal demands of water
Elevation
51. Introduction
Literature
Review
Research
Methodology
Data
Collection
Model
Implementation
Conclusion
51
Sensitivity Analysis
To study the behavior of pipes with different diameters in north and
south phases throughout their useful life
Simulated in normal condition by keeping all the properties of network as
constant except C factor and diameter
As the pipe ages, it gets tuberculated reducing the value of C factor
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 10 20 30 40 50 60 70 80 90 100
HydraulicReliability
% Reduction in C Factor
Sensitivity Analysis
North Phase vs South Phase
North Phase 150mm North Phase 200mm North Phase 250mm
South Phase 150mm South Phase 200mm South Phase 250mm
Small diameter pipes are less
reliable than larger diameter pipes
As the diameter of pipe gets
increased, the failure tendency gets
reduced and the pipes behave
likewise
Pipes with same diameter
hydraulically performs better in
south phase sub-network in normal
condition
52. Introduction
Literature
Review
Research
Methodology
Data
Collection
Model
Implementation Conclusion
52
Reliability indices provided at hierarchical levels of component, segment and network
Algorithm of minimum cut set analysis of a network is extended to identify the minimum cut sets
for all demand nodes
Mechanical reliability evaluation is automated to curtail its tediousness
Hydraulic reliability of a water distribution network is evaluated considering the effect of pressure
An integrated reliability assessment model is developed
Contribution
54. Introduction
Literature
Review
Research
Methodology
Data
Collection
Model
Implementation Conclusion
54
Consider more factors to compute the failure rate of pipes
Develop a failure rate prediction model for water distribution network components other than pipes
Consider the effect of rehabilitation on reliability
Investigate the nodal demands for identifying the demand multipliers to run extended period hydraulic
simulation
Supplement with rehabilitation scheduling, budget allocation and life cycle cost models
Integrate with reliability prediction models of other infrastructure by exploring interdependencies among them
Future Work
55.
56. Introduction
Literature
Review
Research
Methodology
Data
Collection
Model
Implementation
Conclusion
56
Best Practice Guides
Types of failures Causes Consequences
Mechanical
Corrosion
Permeation
Failure due to aging and weathering
Contamination of Mains, Fittings, and
Appurtenances
Contamination of Storage Facilities
Contamination Due to the Absence or
Operational Failure of Backflow
Prevention Devices
Hydraulic
Pipe Deterioration
Hydraulic Changes during Maintenance
and Emergencies
Tuberculation and Scale
External Contamination
Sedimentation
Reduction in Hydraulic Capacity and
Associated Increase in Pumping Costs
(NRC, 2006)
57. Introduction
Literature
Review
Research
Methodology
Data
Collection
Model
Implementation
Conclusion
57
City of London
Case Studies
London Doha
South Phase
North Phase
Mechanical Reliability Model
Hydraulic Reliability Model
Integrated Network
Reliability
Demand
Node
GIS ID Demand (m³/day)
Pressure Head
(m)
10 N10333 70 39.56
7 N11569 124 37.86
4 N11570 62 34.55
3 N11594 114 40.66
2 N11596 57 43.59
1 N11638 62 41.9
11 N16836 77.5 43.06
12 N16837 114 40.51
13 N16839 133 39.94
8 N16841 10 43.96
5 N16844 114 43.29
9 N16846 62 41.56
6 N16849 76 41.09
Service Pressure Heads for
Demand Nodes of South Phase
Sub-Network
58. Introduction
Literature
Review
Research
Methodology
Data
Collection
Model
Implementation
Conclusion
58
City of London
Available Pressure Heads at
Demand Nodes of South Phase
Sub-Network
Demand
Node
a closed
b + d
closed
c + g
closed
n + q
closed
10 0.00 0.00 39.48 39.56
7 0.00 0.00 37.76 37.86
4 0.00 0.00 0.00 34.55
3 0.00 0.00 40.67 40.66
2 0.00 43.59 43.59 43.59
1 41.90 41.90 41.90 41.90
11 0.00 0.00 42.97 0.00
12 0.00 0.00 40.43 40.50
13 0.00 0.00 39.86 39.93
8 0.00 0.00 43.88 43.97
5 0.00 0.00 43.22 43.29
9 0.00 0.00 41.48 41.56
6 0.00 0.00 41.01 41.09