QUANTITATIVE WATER MANAGEMENT IN RABAT, SALE AND 1 [617803]

1
QUANTITATIVE WATER MANAGEMENT IN RABAT, SALE AND 1
TIMISOARA DRINKING WATER SYSTEM, USING GIS 2
3
Ziane Imane1*, Mohammed Karim Ben Hachmi1, Rares Halbac -Cοtοara -Zamfir2 4
5
1Department of Process Engineering and Environment, University Hassan II of Mohammedia, 6
Science and Technology Faculty of Mohammedia, 20650 Mοrοccο 7
2 Department of Hydrotechnics, Faculty of Civil Engineering, Polytechnic University of 8
Timisoara , George Enescu Street, No. 1/A 300022 – Timisοara, Romania 9
10
Abstract 11
12
For efficient management of the losses of drinking water distribution networks, it is essential to 13
have a precise knowledge of the system and its infrastructures. In this context, our study consists 14
in describing two drinking water distribution networks, from two different countries: Ra bat-Salé 15
from Maroc and Timisoara from Romania, given water losses over a period of four / five years. 16
The work consists of calculating some performance indicators to establish thematic maps using 17
Geographical Information System (GIS) in order to identify and prioritize the critical sectors by 18
the method AHP (Analytical Hierarchy Process) based on technical criteria (night flow, linear 19
loss index, linear leakage index, pipe age) . The city of Timisoara (Romania) has a huge volume 20
of water losses, inefficient use of resources, scarce data and poor control. The Rabat and Salé 21
(Morocco) network also has significant water losses, but thanks to the sectorization of the 22
network and flow measurements, data are available for local managers. The technical 23
performance indicators and the results obtained gave us an idea of the reliability of the leak 24
management methods adopted by each manager . 25
26
Keywords: drinking water system, geographic information system (GIS), leaks 27
28
1. Introduction 29
30
Technical management of drinking wa ter supply systems is a national and international 31
issue. Each year, more than 32 billion m3 of treated water does not reach subscribers due to 32
numerous failures such as leaks, faulty connections or piping. In some low -income countries, 33
losses account for 50-60% of the water supplied when the global average is 35% (Farley et 34

*Author to whom all correspondence should be addressed: E -mail: [anonimizat] ; Phone: +212615372167.

2
al.2008). Water loss occurs in all distribution systems, only the volume of loss varies. This 35
depends on the characteristics of the pipe network and other local factors, the water utili ty's 36
operational practices and the level of technology and expertise applied to controlling loss. Irene 37
(Karathanasia, 2016) . The basic function of water distribution system is to satisfy water users’ 38
needs. This water must meet standards of good quality, potability, pressure and enough quantity 39
(Haidar, 2006). This type of management requires precise knowledge about the system and both 40
its infrastructures and hydraulic operation, while maintaining a regular system maintenance. For 41
better management, the us e and integration of IT tools will facilitate and help managers make 42
decisions and develop action plans and interventions. Among these tools, Geographic 43
Information Systems (GIS), they are presented as efficient technologies. They play a vital role in 44
wate r resources management because this technology is considered as one of the most important 45
technologies for integration, analysis and manipulation of data. Moreover, this technology is well 46
known and implemented in many countries of the world and has shown a great efficiency in the 47
field of resource management through their performance in georeferencing. This technology 48
offers appropriate tools for spatial data combination as well as models on the same graphic 49
support. This will allow data communication betw een stakeholders to ensure good coordination 50
of activities, Unfortunately, In Morocco and Romania, this technology is little used in 51
infrastructure at national or regional level. 52
In Morocco, the Department of Water and Environment stated that the state of degradation of the 53
networks in the cities had resulted in a loss of 35% of the distributed water "Rabouli 2014". This 54
explains the development of several methods to estimate current and future performance. In this 55
article, the goal is to provide an interac tive decision support tool that will help managers 56
optimize management and intervention. This decision -support tool is proposed to provide 57
general information (collection, analysis, and prioritization) and to sort the priority intervention 58
sections taking into account several performance indicators. 59
Do drinking water system operators in Romania and Morocco have the same network problems? 60
Is their leak management strategy reliable? 61
62
2. Case studies presentation and Methods 63
In order to help the drinking wat er supply network managers successfully accomplish the 64
mission entrusted to it, we propose a structured methodology in several successive and distinct 65
stages to improve the strategy of the detection of the leaks and which will lead to the 66
optimization of t he service. 67
The methodology is based on collection, processing and analysis of different data. It aims 68
to develop support for study and management. We chose two cases, belonging to two different 69

3

countries: Rabat and Sale in Morocco located in North Africa and Timisoara in Romania located 70
in Eastern Europe. We will study each independent case with the data available and provided by 71
the managers. The methodology adapted to do the work is presented as follows in Fig.1: 72
73
Fig. 1. General organization chart of t he intervention plan. 74
75
After calculating and analyzing the performance indicators for each drinking water supply 76
system, we will have used the AHP (Analytical Hierarchy Process) for the Rabat , Sale city , this 77
AHP method is broken down into four stages (S aaty TL, 1996): ranking of indicators by 78
importance from most important to least important, constructing a matrix based on the 79
comparison of indicators two by two, determination of the weights associated with each indicator 80
using a vector calculation metho d clean and finally check the consistency of the result. 81
82
The first phase consists in structuring the decision problem in a hierarchical structure by 83
identifying the elements (indicators), the second phase, a pairwise comparison matrix is 84
established to ev aluate the importance of each of them. Then, the overall weights are calculated. 85
For comparison, the experts determined the relative importance using pairwise comparisons, the 86
values being suggested by Saaty , scale from 1 to 9 (Harker, P.T., 1989), allowin g the decision 87

4
maker's judgments to be brought closer to reality. , the meanings of which are given in Table the 88
comparison leads to obtaini ng the decision matrix (Eq.1) 89
A= [
] =
(1) 90
91
92
A: is the decision matrix, aij (individual priority) the comparisons between the elements i and j 93
for all i, j ∈ {1, 2 … n}. 94
Then we have to check the consistency of the result. At this point, we have the "weight" of 95
each of the elements. The AHP method then proposes to validate the reliability of the result in 96
Calculating the coherence ratio (CR) the consistency ratio is is calculated as follows Eq. 2 . 97
Saaty (2012) has shown that a consistency ratio (CR) of 0.10 or less is acceptable to continue 98
the AHP analysis. If the consistency ratio is greater than 0.10, it is necessary to revise the 99
judgments to locate the c ause of the inconsistency and correct it. 100
101
CR= CI / RI (2) 102
103
With CI: index of coherence and RI: index of a random -like matrix 104
after the prioritization of the indicators the criteria are used to build new G IS thematic maps, 105
with high, medium or low risks, which are used to identify critical sectors. 106
107
2.1. First study area: Rabat and Sale cities 108
109
Rabat -Sale-Kenitra region covers an area of 18,194 km2 and has 4,581 thousand 110
inhabitants, a density of 251.8 i nhabitants per km2 and an area of 2.56% of national territory. It is 111
limited North by Tangier -Tetouan -Al Hoceima region, East by Fez -Meknes region, and South by 112
Beni Mellal -Khénifra region and Casablanca -Settat region, and West by Atlantic Ocean. In our 113
study we wi ll focus on both cities Rabat and Sale. 114
Data is provided by REDAL (Autonomous Water and Electricity Distribution Authority), 115
which is responsible for delegated management of liquid sanitation and drinking water and 116
electricity distribution service s in Rabat – Sale – Zemmour – Zaers Wilaya, and which serves 117
nearly 2 million people inhabitants. Network heritage represents a total linear of 3980 km in 118
drinking water. It is broken down into an infrastructure network and the non -structuring network 119
service that are distinguished by pipes diameter a nd materials that compose them. 120

5

121
Fig. 2. Hydraulic sectors of the cities Rabat and salé 122
123
2.2. Second study area: Timisoara city 124
125
Timisοara is the capital of Timiș Country; situated in the far western part of t he country, 126
with a population of 319.272. Within a radius of 600 Kilometres’, there are seven major capital 127
cities of C entral and south Eastern Europe. (Neumann, 2015) . 128
Aquatim S.A. Timisoara, operates the water supply and sewerage systems in the area of 129
operation starting with 2010, the area of operation at the end of 2015 includes 102 water supply 130
systems in 102 localities including a municipality Timisoara (Aquatim SA database). In 131
Timisoara city, the sectorization method is still under study, so we will not have data by sector 132
but of the whole city. 133
134
3. Results and d iscussion 135
136
3.1. Analysis of drinking water system performance parameters Of Rabat and Sale 137
138
Night flow: Regular measurement of night flow sectors makes it possible to locate the 139
important le aks which exist, but also to be informed quickly about the emergence new leaks in 140
areas where important ones are likely to occur. Flow measurements are carried out continuously 141
concerning the four last year’s evolutions 2012, 2013, 2014 and 2015 for Rabat and Sale cities in 142
(Fig.2 and Fig.3) 143
144

6
145
Fig. 3. Evolution of the minimum night flow of Sale 146
147
2012 and 2013 recorded high night -flow values, especially in sectors B, H, I1, I2 and J. 148
For 2014 -year, leakage losses are not negligible in almost all hydraulic s ectors except for Areas 149
D, E, Sala Al Jadida (S -B, S-H) and E-B (Bouknadel ) saw no increase in night flow. For year 150
2015, Compared with previous years, it is clear that sectors B, H, I1, I2 and J still require an 151
intervention at detection level and leaks r epair. It is recommended to check network and water 152
tightness in order to anticipate loss reduction measures such as sectorization and pressures 153
modulation. 154
155
Fig. 4. Evolution of the minimum night flow of Rabat 156
Fig. 4. Evolution of the minimum night flow of Rabat 157

7
From 2012 to 2015, we note that D2, E, F, G, J, K and L sectors have very high night – 158
time values. It should also be noted that 2013 does not show an effective reduction in leakage 159
losses especia lly in sectors F, E, D2 and G. 2014 b ased on previous graph of average 2013 night 160
flow evolution. We note that night flow values of sectors D2, E, F, and G are always in increase 161
knowing that these sectors are modulated. Sectors J, K, L, in 2015 experien ced a sharp increase; 162
it is recommended to check network of these sectors in order to anticipate loss reduction 163
measures and react with a maintenance plan or pipe renewal for their optimization. 164
165
Linear loss index: is used to monitor network evolution an d to evaluate leakage losses 166
on distribution network. This index also makes it possible to compare physical state of two or 167
more networks. Linear water loss index (LLI) use, expressed in cubic meters per day and per 168
kilometre of pipeline. LLI is calculated as follows Eq. 2 : 169
170
LLI = Q/L (3) 171
172
With Q: Night flow, L : Network length (m3 / km / d). 173
174
Processing results linear loss index data from 2012 to 2015 by ARCGIS software are 175
presented in LLI form evaluation cards by sector of each city below (Fig.5 ) 176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192

8
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
Fig. 5. Linear loss index (LLI) m3 / J / Km evaluation in Rabat and Sale cities sectors in 217
2012,2013,2014,2015 : (a) linear loss index m3 / d / Km of the Rabat and Sale cities sectors in 218
2012, (b) linear loss index m3 / d / Km of the Rabat and Sale cities sectors in 2013, (c) linear 219
loss index m3 / d / Km of the Rabat and Sale cities sectors in 2014, (d) linear loss index m3 / d / 220
Km of the Rabat and Sale cities sectors in 2015 221
222
According to maps analysis of losses linea r index presented previously, one notes that 223
they present different network categories according to LLI namely: 224
– Sectors presented in green indicates that drinking water system is in good condition, 225
– Sectors shown in yellow indicates that drinking water system is u nacceptable condition, 226
– Sectors presented in red indicates that drinking water system is in poor condition. 227
(b)
(c)
(d) (a)

9
We note that it has a status degradation in most systems of Rabat city and Sale based on 228
losses linear index, for example in 2012, Raba t sectors H / I / J / K / L and Sale’s Bouknadel 229
were in good condition, then in 2013 in acceptable condition, then in a bad state. This gives us an 230
indication that strategy for managing and reducing leaks in Rabat and Sale drinking water system 231
is unrelia ble. It is recommended to check these sectors ‘networks and revise strategy to solve 232
problems in drinking water system. 233
To confirm the analysis of drinking water system category based on linear loss index of 234
Rabat and Sale cities, the following table is u sed which presents of linear loss indices 235
classification according to drinking water system type (Table 1). 236
237
Table 1. Linear loss indices classification according to drinking water system type 238
239
Drinking water system
category Rural Semi -rural Urban
Good LLI < 1 LLI < 3 LLI < 7
Acceptable 1 ≤ LLI ≤ 3 3≤ LLI ≤7 7 ≤ LLI ≤ 12
Bad LLI > 3 LLI > 7 LLI > 12
240
ILI leakage linear index : is expressed from minimum flow of a network that is 241
generally observed. It makes it possible to estimate share of leakage -related water losses 242
considering that nig ht-time consumptio n is negligible. His formula is (Eq.4 ). Judgment of 243
network state in leaks terms, based on the ILI is made on data derived basis from work 244
experience of detection agents / search for leaks at REDAL (Table 2 ). 245
246
ILI = Total Leaks Number / Cumulative Swept Lines (Leaks / km) (4) 247
248
Table 2. Linear leakage indices classification by network type 249
250
Drinking water system category
Good ILI< 0,3
Acceptable 0,3 < ILI < 0,6
Bad ILI > 0,6
251
Pipes Age: Historical installation data are taken from GI S records. It can be seen from 252
figure below that some periods were more conducive to certain materials use . 253

10
254
255
Fig. 6. Drinking water system pipes age by type 256
257
Emergency level : It is at this level that we can identify prioritized sectors that are most 258
deficient with respect to drinking water supply system and optimize decision -making. 259
Performance of a criterion is calculated from indicators performance associated with it (Ellis. et 260
al., 2004). 261
We used AHP (Analytical Hierarchy Pr ocess) method, this AHP method is broken down 262
into four stages (Saaty, 1996): indicators prioritization by importance from most important to 263
least important, a matrix construction based on comparison of two by two indicators, weights 264
determination associat ed with each indicator, thanks to calculating method eigenvectors finally 265
checking result consistency. 266
According to collected data and criteria of the studied area, and according to sort 267
performed during data management, the most important parameters that have weight when 268
studying are: night flow, linear loss index, and linear leakage index and installation date. 269
Night flow: weight 5, 270
Linear loss index (LLI): weight 4, 271
Linear leakage index (ILI): weight 3, 272
Pipes Age: weight 2. 273
The number of class is set at three for better readability and a good interpretation of 274
resulting maps. This approach will lead to thematic maps production and water leakage risks for 275
years 2012 to 2015, this approach will be programmed in ArcGIS in order to automatically 276
determine emergency levels for each sector in map form that will be easily exploited by 277
managers. Criteria classification will lead to thematic maps of three classes that are risk, high, 278
average and low classes. 279
This methodology has been applied to identify priorit y sectors on drinking water system 280
of Rabat, and Sale cities. This implementation of this step has been facilitated by ArcGIS software use. 281
Results obtained by software are automatically presented in maps form below (Fig.6) . 282

11
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
Fig. 7. Water leaks risk maps in hydraulic sectors for 2012,2013,2014,2015 : 310
(a) Water leaks risk maps in hydraulic sectors for 2012, (b) Water leaks risk maps in hydraulic sectors for 311
2013 , (c)Water leaks risk maps in hydraulic sectors for 2014, ( d) Water leaks risk maps 312
in hydraulic sectors for 2015 313
By analyzing water leaks risk maps in Rabat city hydraulic sectors, sector O 314
modulation in 2013 reduced water leakage risk but network condition then was degraded after 315
2014 which gives us an indication th at the methodology used by REDAL is reliable over a short 316
period of time. For sectors D1 / D2 / E / F J / K and L, they are still in a critical state, and require 317
an urgent intervention. 318
(a)
(b)

(c)
(d)

12

For Sale city, Sector A, H, I1, are still in critical condition, thu s there must be an 319
intervention to reduce water leakage and to innovate in these sectors. This indication makes it 320
possible to establish an action plan for leaks detection and orientation of research for better 321
management of drinking water system, as well as anticipating emergency program for immediate 322
drinking water system leaks repair. 323
To confirm our analysis, we will evaluate the reliability of REDAL leakage strategy followed by 324
gain percentage compared to each previous. 325
326
327
328
329
330
331
332
333
334
(a) 335
336
337
338
339
340
341
342
343
(b) 344
Fig. 8. The gain percentage in Rabat and Sale for 2013, 2014, 2015 : (a) The gain percentage in Rabat and 345
Sale for 2013, 2014, 2015, (b) The gain percentage in Rabat and Sale for 2013, 2014, 2015 . 346
347
After the treatment of Rabat and Sale’s data, several sectors ha ve priority over other 348
sectors. Hence establishing interest an action plan for leaks detection and team’s orientation for 349
drinking water system better management, as long as regular and anticipated verification of an 350
emergency program for repairing network leaks. Water leakage risk thematic maps of pipeline 351
sector obtained at end of this work for both cities were consistent with distributor's perspective. 352
This proves geo -computer system reliability that was created during this work and offers 353
opportunity t o exploit this system on a larger scale and save time and manage resources, the 354
thing which is altarpiece for this latter, not forgetting its characteristic and its capacity in 355

13
coupling with a large and medium software range and stimulation systems, manage ment and 356
programming. 357
Monitoring, under good conditions, should lead to reduce water losses, especially follow – 358
up and results to identify performance indicators at a good level. The objective is to preserve 359
these assets and monitor network evolution so as to ensure as quickly as possible any slippage in 360
water consumption to ensure its reliability and performance. 361
In this work and according to the obtained results, the objective was to detect and identify 362
areas risk zones to reduce and control leaks in Rab at and Sale drinking water system. I t is 363
mandatory to move towards a pressure modeling for high linear index sectors ILI and LLI since 364
excess reduction pressure makes it possible to reduce continuation flow rate of one part, and 365
continuation frequency of a nother part. Monitoring network is an essential step in ensuring 366
sustainable management of distribution from which monitoring of night consumption (night 367
flow). 368
It should also be noted that repair details or leakage repairs, for example, can be found in 369
diagnostics sheets of drinking water connections and pipes, which contain all information needed 370
for future monitoring and diagnosis (failure nature, installation conditions, equipment condition, 371
environment condition, etc.). To ensure this task, tools and means are varied and will depend, in 372
particular, on methods implemented to carry out diagnosis . 373
374
3.2. Analysis of drinking water system performance parameters of Timisoara city 375
376
The most common method of determining water loss is water balance calculation by 377
IWA methodology (International Water Association).The table 4 below shows the volume of 378
water entering the system and the loss of water for each system within the operating area over 379
the last five years in Timisoara: 380
381
Table 4. Water volu me entering the system and the loss of water for Timi soara city . 382
Water
supply
system Year System input
volume Water losses
Apparent losses Real losses
m3 m3 % m3 %
Timisoara
water
supply
system 2011 36.558.840 1.853.574 5,1 12.946.587 35,4
2012 36.396.596 1.837 .377 5,0 13.099.069 36
2013 33.671.343 1.726.054 5,1 11.178.339 33,2
2014 31.239.729 1.627.274 5,2 9.583.773 30,7

14
2015 31.274.780 1.637.259 5,2 9.361.064 29,9
383
Performance indicators : (based on IWA specifications) facilitate continuous analysis and 384
benchmarking. The objective is to describe the resources spent (e.g. repair cost/km) or conditions 385
(e.g. number of failures/km), to compare sub areas or entire networks and to estimate the benefits 386
of a rehabilitation or pressure management program. The IW A water performance indicators for 387
water loss are: 388
– Specific water losses (QSL) 389
– Non-revenue water 390
– Current Annual Real Losses (CARL) 391
– Unavoidable Annual Real Losses (UARL) 392
– Infrastructure Leakage Index (ILI) 393
– Customer leakage index (CLI) 394
In the Ta ble 5 are presented the main performance indicators for several water supply systems. 395
Table 5. Volume of water entering the system and the loss of water for each system within the 396
operating area over the last five years 397
398
Water
supply
system Year Performa nce indicators
QSL CARL UARL
ILI Non-revenue water
m3/km/day m3/day m3/day m3 %
Timisoara
Water
supply
system 2011 54.65 35.470 907 39 14.891.558 40.7
2012 55.30 35.790 921 39 15.027.458 41.3
2013 47.19 30.626 993 31 12.988.571 38.6
2014 40.46 26.257 1.001 26 11.289.146 36.1
2015 39.51 25.647 1.100 23 11.076.510 35.4
After calculating the ILI for water supply system, all values are greater than 8; so he 399
hastenance and a very inefficient use of resources, poor maintenance indicative and sy stem 400
condition in general. An initial assessment of its condition and an immediate description of 401
water loss management performance can be obtained by using World Bank Institute's banding 402
system shown in Table 6. 403
404
405
406
407

15
Table 6. World Bank Institute’s bandi ng system for developed and developing countries 408
409
WBI
band ILI range
Guideline description of real loss management
performance categories Developed
countries Developing
countries
D
8.0 or
more
16.0 or
more Very inefficient use of resources, indicativ e of poor
maintenance and system condition in general, leakage
reduction programs imperative and high priority
410
Customer leakage index : A reference system of the linear leakage index adjusted 411
proportionally to the customers’ density has ben adopted by SM ERGREG: the customer leakage 412
index (CLI) (Eq. 5 ) (Renaud, 2010). 413
414
CLI = CAWL / (365 ×N) (5) 415
416
CLI: Customer leakage index in m3/customer/day, CAWL: Current Annual Water Losses in m3 417
N: Number of customers. 418
[EEMJ11]
419
Fig. 9. Customer leakage index in m3/customer/day in 2011, 2012, 2013, 420
2014, 2015 for Timisoara city 421
Material used for distribution networks execution : During the historical development 422
of the water distribution networks, the materials used were: gray c ast iron, asbestos, and steel, 423
precast concrete, PVC, HDPE , ductile iron and composite materials (HOBAS -fiberglass).Over 424
the years, it has been found that old water networks made of gray cast iron or asbestos pipes are 425

16
sensitive to external factors (traffi c, laying bed and improper filing) factors that lead to the 426
occurrence of transverse cracks. (Aquatim SA database ). HDPE is the most used. 427
According to some performance parameters calculation of Timisoara city drinking water supply 428
network . The price of a cubic meter of water is (Eq. 6 ): 429
430
Water price/cubic meter: 3,11 lei = 0.66 EUR (6) 431
432
Table 7 . Non-revenue water prize of Timisoara city during 2011 until 2015 in lei and euro 433
434
Unexamined water must be equal to or greater than 40% according to standards . 435
Timisoara city results according to Aquatim since 2011 until 2015 are down, which explains the 436
efforts made by the company for drinking water network management. But these efforts the 437
performance indicators calculated previously indicate that the network’s state is still in a bad 438
state and that it costs enormously expensive, in 2015, these costs of water in 7310496,6 euros, 439
which is a very important am ount. So u rgent intervention is mandatory . It is proposed to move 440
towards sectorization and integration and use of ArcGIS geographic information system, 441
considered as an effective technology. This computer tool is one of the best known technologies 442
in data integration field, analysis and processing. Thanks to their georeferencing performance. 443
Thus, network monitoring to ensure sustainable management of distribution network from where 444
the regular monitoring of interest night consumption (night flow). It shou ld also be noted that 445
details of repair or leaks repair, for example, can be found in diagnostics sheets for drinking 446
water connections and pipelines containing all information needed for future monitoring and 447
diagnosis. To ensure this task, tools and reso urces are varied and will depend in particular on 448
methods used to carry out the diagnosis. 449
450
3.3 Comparison between the two study areas 451 Water
supply
system Year Performance indicators Cost of loss
Non-revenue w ater LEI Euro
m3 %
Timisoara
Water
supply
system 2011 14.891.558 40.7 46312745.38 9828428.28
2012 15.027.458 41.3 46735394.38 9918122.28
2013 12.988.571 38.6 40394455.81 8572456.86
2014 11.289.146 36.1 35109244.06 7450836.36
2015 11.076.510 35.4 34447946.1 7310496.6

17
452
We cannot really speak of a comparison since the work data of each manager are 453
different, and they do not use the same performance indicators, but, our work consisted 454
essentially in making a diagnosis to know the state of the networks. Drinking water in Rabat, 455
Sale and Timisoara cities, and analyze the data. This diagnosis allowed us to identify certain 456
malfunctions, to se ek and propose improvements to a management system that takes into account 457
at the same time realities. Each manager and compliance with the rules for the management and 458
operation of an urban drinking water network. According to the analysis of the data tha t we have 459
been able to collect from the managers, we can see that there is a great deal of effort in each 460
country by the managers, but there are still weak points in each system that is needed to 461
reinforce. The following table 6, summarizes the difference between the two systems: 462
463
Table 8. The difference between the two systems Rabat, Sale and Timisoara . 464
465
466
4. Conclusion s 467
468
Given the difficulties of short -term management of such a network as well as distribution 469
and maintenance constraints from a sustainable development perspective, it was necessary to 470
adopt a structured and adaptive methodology, which would lay the foundations for a 471
management tool. Responding in particular to the following three points: 472
– Assistance in diagnosing the state of net work malfunctions by an approach adapted to 473
the context of an aging and highly disturbed network; 474 Rabat and Sale Timisoara
Sectorization X –
Using the geographical information
system (GIS) X –
Emergency Intervention Method X X
Calculation and mo nitoring of indicators X X
Renewal and rehabilitation of facilities – –
Updated information and data base X –

18
– Consistent and up -to-date management of data infrastructures and networks using a 475
geo-referenced database tool. 476
– Support for the network renewal process in connection with the urgency of the ongoing 477
rehabilitation project, through the development of an evaluation tool for priority areas. 478
The challenge is to create a decision support tool that provides managers with a 479
quantitative means to implement driving r enewal programs in different areas of the city, at 480
different horizons. Taking into account all the parameters characterizing the drinking water 481
network. 482
Although very promising, this tool remains very sensitive to the fact that there is no 483
sufficient basi s for interventions and highlights two things: 484
1) The need for a network -wide operatio nal database to improve results. 485
2) A restructuring of the intervention database is needed to better characterize the different 486
Types of events. The database must provid e: 487
– The introduction of the exact location of the intervention or incident, 488
– The description of the importance of driving, 489
– The presence or absence of internal and external protection on the pipes, 490
– The particular climatic conditions (low temperature l eading to freezing of the pipe), 491
– The exact time of the incident or intervention (close to the minute). 492
These two points show the need to enrich the database in the years to come, which must take into 493
account the first two points, as regards the analysis of dysfunctions. 494
495
References [EEMJ12] 496
497
Blindu I., (2004), Diagnostic aid tool for the drinking water network for the city of Chisinau by 498
spatial and temporal analysis of dysfunctions hydraulic , PhD Thesis, Saint -Étienne School of 499
Mines, France. 500
Ellis J. B., Deuts ch J. -C., Mouchel J. -M., Scholes L., Revitt D, (2004), Multicriteria Decision 501
Approaches to Support Sustainable Drainage Options for the Treatment of Highway and 502
Urban Runoff, Science of the Total Environment , 334–335, 251 -260. 503
Renaud. E., (2010), Tow [EEMJ13] ards a global performance indicator for losses from water supply 504
systems, Water Loss 2010, Sao Paulo, Brazil, Jun 2010, 11. 505
Engelhardt M.O., Skipworth P.J., Savic D.A., Saul A.J., Walters G.A, (2000), Rehabilitation 506
strategies for water distribution networks : a literature review with a UK perspective, Urban 507
Water , 2, 153 -170. 508

19
Dell’orfano F., Esposito V., Gualtieri P., Pulci doria G., (2013), Mean Values of Water Pipe 509
Break Rate in the World and in Geographical Areas, 7th IWA International Conference on 510
Efficient Use and Management of Water (Efficient 2013) Paris, France, 22 -25 October 2013. 511
Haidar H., (2006), Drinking water networks rehabilitation, Methodology for multicriteria 512
heritage analysis and rehabilitation programs , PhD Thesis, National Institute of Applied 513
Sciences of Lyon, France. 514
Nafi A., (2006), Multiannual renewal program for drinking water networks , PhD Thesis, 515
University Louis Pasteur, Strasbourg I, France. 516
Granger D., Caradot N., Cherqui F., Chocat B., (2010), The assessment of the overflow risk of 517
urban water management systems: methodological contributions of two cases of studies 518
(Lyon and Mulhouse), 7th International Conference on Sustainable Techniques and 519
Strategies for Urban Water Management in Rainy Weather, NOVATECH Lyon, France, Jun 520
2010,1 -10. 521
Saaty T., (1996), Decision Making with Dependence and Feedback: The Analytic Network 522
Process, WS publications, University of Pittsburgh Berlin Heidelberg, 4922 Ellsworth 523
Avenue Pittsburg, 15213. 524
Neumann S., (2015), First Monitoring Meeting, Rep ort by the panel of independent experts, 525
Candidate City. On line at: https://ec.europa.eu/programmes/creative -europe/sites/creative – 526
europe/files/files/ecoc2021 -timisoara -monitoring_ en.pdf. 527
528
Farley, M., Wyeth, G., Ghazali, Z.B.M., Istandar, A., Singh, S., 2008. The manager’s non – 529
revenue water handbook: a guide to understanding water losses. Ranhill Utilities Berhad and 530
the United States Agency for International Developement, Bangkok, Thailand. 531