Strategic planning optimization of “Napoli Est” water distribution system

(in lingua inglese)

Coupling the DMA design with the existing water flow remote measurement system will provide the following benefits:
- remarkably reducing water losses;
- reducing the number of maintenance operations, after reducing the operating pressure of the system;
- easy finding the most vulnerable areas of the network;
- making quick maintenance and repair in the network.
Obviously, results of numerical simulations are only preliminary, and should be confirmed in the future by implementing and monitoring one or more DMA and comparing numerical simulations with in situ measurements.

Scarica il PDF Scarica il PDF
Aggiungi ai preferiti Aggiungi ai preferiti

Articoli tecnico scientifici o articoli contenenti case history

da Alessia De Giosa


Parole chiave: 

Estratto del testo
STRATEGIC PLANNING OPTIMIZATION OF ''NAPOLI EST' WATER DISTRIBUTION SYSTEM Maurizio Giugni(1), Nicola Fontana(2), Davide Romanelli(3), Davide Portolano(4) (1)Department of Hydraulic and Environmental Engineering G. Ippolito, University of Naples Federico II, Via Claudio 21, 80125 Naples, Italy, phone: +39 081 7683443; fax: +39 081 5938936, e-mail: (2)Department of Engineering, University of Sannio, Piazza Roma 21, 82100 Benevento, Italy, phone: +39 0824 305564, fax: +39 0824 325246, e-mail: (3)ARIN S.p.A., Via Argine 929, 80147 Naples, Italy, phone: +39 081 7818111, fax: +39 081 7818190, e-mail: (4)Centre of Environmental Research C.I.R.AM., University of Naples Federico II, Via Mezzocannone 16, 80134 Naples, Italy, phone: +39 081 401023, fax: +39 081 409782, e-mail:
ABSTRACT The District Meter Areas (DMA) design is an innovative methodology of water networks management, based on the pressure patterns control and on the water flows monitoring, in
order to reduce water losses and to optimize the water systems management. In the present
paper the DMA design of the ''Napoli Est' water distribution system (approximately
65.000÷70.000 customers), performed with the support of the Water Agency ARIN S.p.A., is
discussed. After analysis of authorized consumption, by means of a monitoring campaign of water fluxes over the area, the system water balance was performed, showing significant water
losses, as a consequence of high pressure patterns. District Meter Areas, therefore, were
designed, and the corresponding hydraulic and water quality investigations and simulations
were carried out. Six District Meter Areas were planned, assembling 14 intercepting valves
and 9 pressure reducing valves to prevent the downstream pressure head from exceeding the
set value, achieving a remarkable water saving, approximately equal to 34% of the physical
losses. Keywords: water networks, water systems management, District Meter Areas design, pressure
patterns, physical leaks, water losses reduction.

1 INTRODUCTION From a recent report of the Committee for the Vigilance on the Use of Water Resources (2004), with reference to approximately 61% of Optimal Territorial Ambits (ATO), in Italy
the total water losses (physical and administrative) range between 20÷65%, with an average
value around 40%. As well known, the greater share of physical leaks is localized in distribution networks: an active leakage control in order to reduce water losses would concur, therefore, a significant
water saving with economic and environmental benefits. Beside the advances in leak detection practices and techniques (leak localising, leak location, repair of leaks), in the 90ies the District Meter Areas (DMA) design was introduced
in the management of the water distribution systems (Cheong, 1993). A District Meter Area is
an area supplied from few water inputs, into which discharges can be easily measured to
determine leaks. It is well known the pressure-leakage relationship, so in the DMA an
effective control of the head patterns is performed, by means of pressure reducing valves to
prevent the downstream hydraulic grade from exceeding a set value. Such technique is,
therefore, an effective instrument to optimize the management of a water distribution system, with beneficial effects on the water losses reduction and probably also on the pipe failure
rates and rupture frequencies. The DMA design can be chosen in alternative to the traditional approach of heavy looped distribution networks. These systems on the one hand offer a remarkable level of reliability
(ready emergency response with little impact on the customers, for example in the case of a
main break; simple territorial expansion of the distribution system; etc.) but on the other hand
they involve a less effective control of the water network and greater difficulties in the
leakage management activities (Artina et al., 2005; Giugni et al., 2005). D.M. 99/1997 introduced in Italy the DMA design, but until now only few examples of application are known (Modena, Reggio Emilia, Bologna). This in consequence of economic
factors and technical difficulties, due to the complexity of the DMA design operations and
sometimes to the lack of an accurate acquaintance of the water networks topology.

2 THE CASE STUDY: THE ''NAPOLI EST' WATER DISTRIBUTION SYSTEM In the present work are reported the noteworthy points of a research aimed to the DMA design of ''Napoli Est' network, pertaining to the system of Water Agency ARIN S.p.A.
(Neapolitan Water Resources Agency). In greater detail, the East Naples area has a surface approximately equal to 920 hectares (nearly 8% of the Municipality of Naples) and covers most of the eastern zone of the city
(East Naples; Fig. 1). Resident number is approximately 65.000 ÷70.000 and the area elevation varies in the range 11 ÷78 m above sea-level. The network is supplied by ''San Sebastiano' reservoir, with 6 tanks (of which only 5 are currently working) for a total storage
volume of 30.000 m3. The reservoir maximum water depth is 5.40 m and the overflow
elevation is 112.50 m above sea-level. Figure 1. Location of ''East Naples' area in the Naples Municipality. The distribution network is heavy looped (Fig. 2) and pipes are in reinforced concrete (going back to 60ies, with diameter DN 1000), grey cast iron, steel and ductile cast iron, with
diameters ranging from DN 40 to DN 1000. By means of a preliminary field survey, the water system operational conditions were characterized and a database for the calibration of the hydraulic simulation model was
implemented. Along the reservoir exit pipe an electromagnetic flow meter was installed (Fig.
2): in Figure 3 the network daily discharge (24 ÷27/05/2006) was showed as an example. Collected data showed that: ' the network daily discharge has low week and seasonal variation. The peak hour demand is around 440 ÷450 l/s, the minimum around 220÷230 l/s. The average-day demand is practically constant and equal to approximately 340 l/s; ' the minimum night flow practically is never smaller than 220 l/s and the peaking factor is approximately equal to 1.3. It follows that ''Napoli Est' network has a high
percentage of physical losses, since the area is characterized by small night activities
(do not exist noteworthy industrial systems or business activities carried out in night
hours). On the basis of collected data, the mass balance (IWA, 2000) of the network (showed in Table 1) was performed, comparing the average daily water volume supplied by ''San
Sebastiano' reservoir (System Input Volume) with the average daily volume billed to the
customers (Billed Authorized Metered Consumption). Daily average volume [m3/d] Daily average flow [l/s] SYSTEM INPUT VOLUME 29.422 340 BILLED AUTHORIZED METERED CONSUMPTION 9.790 113 Water losses 19.632 227 Table 1. Water balance of ''Napoli Est' distribution system.
The billing data were collected by ARIN S.p.A. by means of a management software integrated with a Geographic Information System (GIS). The billed consumption is equal to
9.790 m3/d, corresponding to an average-day demand of 113 l/s. 10 00 CA P (Ex Ca sm ez) 10 00 C A P ( Ex C a sm ez ) 10 00 CA P ( Ex Ca sm ez) 1 00 0 C AP (E x C as m ez) 1 0 0 0 C A P ( Ex C as m ez ) 1 0 00 C AP (E x C asm ez) 100 0 C AP (Ex Ca sm ez) 10 00 C A P ( Ex C as m ez ) 10 00 CA P (Ex Ca sm ez) 10 00 CA P ( Ex Ca sm ez) 10 00 CA P ( Ex Cas me z) C C Measure point n. 2 Junction J230 Via Leto Pomponio ang. Via Madonnelle Measure point n. 1 Junction J140 Viale Europa - Camera partenza DN 1000 ex CasMez Measure point n. 4 Junction J002 Via Nuova ENEL ang. Via Ravioncello Measure point n. 3 Junction J243 Via Porchiano ang. Via Mario Palermo Measure point n. 5 Junction J150 Via De Meis ang. Via del Purgatorio Measure point n. 6 Junction J248 Via Mastellone 100 "San Sebastiano" reservoir flow meter Electromagnetic Figure 2. Layout of ''Napoli Est' water distribution system (with pressure transducers). 200 220 240 260 280 300 320 340 360 380 400 420 440 460 0. 00 1. 00 2. 00 3. 00 4. 00 5. 00 6. 00 7. 00 8. 00 9. 00 10. 00 11 .0 0 12. 00 13. 00 14. 00 15. 00 16. 00 17. 00 18 .0 0 19. 00 20. 00 21. 00 22. 00 23. 00 Time [] Fl ow [ l/s ] Instantaneous flow (24/05/06) Average flow (24/05/06) Instantaneous flow (25/05/06) Average flow (25/05/06) Instantaneous flow (26/05/06) Average flow (26/05/06) Instantaneous flow (27/05/06) Average flow (27/05/06) Figure 3. Network daily discharge (24÷27/05/2006). Therefore, from the water balance of ''Napoli Est' distribution system (Table 1) it results -even with inevitable approximations due to measurement errors and uncertainties in the area
delimitation- that the network global losses are about 67%. The heavy losses are confirmed by the high number of maintenance operations performed in the area in 2005 (Table 2). It is probably due to the high number of steel
pipelines (with consequential corrosion phenomena) and above all, as it is shortly going to
illustrate, to high pressure heads, that surely have a strong impact on water leaks and on pipe
failure rates.
Maintenance operations typology n. Pipelines repairing 129 Horizontal repairing 42 Repairing and replacement of valves, junctions, etc. 32 Other 7 TOTAL 210 Table 2. Maintenance operations in ''Napoli Est' network in 2005 (ARIN''s Archives, 2005).
In order to characterize the piezometric head on the network, ARIN S.p.A. supplied to the installation of six pressure transducers in the most vulnerable areas (Fig. 2). The water
level in the reservoir was also measured in order to estimate its influence on the network
pressure head. Measurements allowed to detect small head losses in the network, due to pipes large diameters. Consequently, the network peripheral areas, characterized by low elevations, are
subjected to elevated pressure heads largely superior than the minimum one for efficient
distribution (Fig. 4). The following relationship exists between pressure in network P and physical losses Ql (Khadam et al., 1991; Khaled et al., 1992; Lambert, 2000; Milano, 2006):
l Q C Pα = '' (1) being C and α coefficients varying according to pipe characteristics and type of loss. This
correlation is confirmed and emphasized by the experimental results on the ''Napoli Est'
distribution system, so the DMA design appeared a suitable solution to achieve a water losses
reduction and for future benefits in monitoring the system water balance. 65,00 70,00 75,00 80,00 85,00 90,00 95,00 100,00 105,00 110,00 0.0 0 1.0 0 2. 00 3. 00 4.0 0 5. 00 6. 00 7. 00 8.0 0 9. 00 10. 00 11. 00 12. 00 13. 00 14. 00 15 .0 0 16. 00 17. 00 18. 00 19. 00 20. 00 21. 00 22. 00 23. 00 Time [] P re s s u re [m ] a n d H y d ra u lic h e a d [m a .s .l.] Junction 230 (z=18.30m a.s.l.) - Hydraulic head Junction 230 - Pressure Junction 243 (z=14.40m a.s.l.) - Hydraulic head Junction 243 - Pressure Junction 002 (z=13.10m a.s.l.) - Hydraulic head Junction 002 - Pressure Junction 150 (z=29.50m a.s.l.) - Hydraulic head Junction 150 - Pressure Junction 248 (z=28.20m a.s.l.) - Hydraulic head Junction 248 - Pressure Figure 4. Daily pressures in network. 3 HYDRAULIC SIMULATIONS Hydraulic simulations were carried out with the EPANET software version 2.0 (Rossman, 2000). Hydraulic layout resulted from the network ''skeletonization' (Hamberg e
Shamir, 1988), achieved by eliminating out of order pipes, integrating pipelines of same
diameter and roughness, replacing dead-end branches and small networks supplied by a single
junction with an equivalent flow. To support the skeletonizated layout, the following input data were used: ' Planimetric and topographic layout of the system and hydraulic parameters of pipelines and junctions. The system is constituted by 100 meshes, 259 junctions and
358 pipes. Every pipeline was defined by length, diameter and roughness coefficient
(the Hazen-Williams equation was adopted in simulations, defining each pipe
coefficient according to material and age); each junction was defined by elevation and
by distributed and billed flow. ' Daily pattern of water levels in the tank. The daily water level in the reservoir was assigned from measured levels during the field measurement campaign (Fig. 5). 107,00 108,00 109,00 110,00 111,00 112,00 113,00 0. 00 1. 00 2. 00 3. 00 4. 00 5. 00 6. 00 7. 00 8. 00 9. 00 10 .0 0 11 .0 0 12 .0 0 13 .0 0 14 .0 0 15 .0 0 16 .0 0 17 .0 0 18 .0 0 19 .0 0 20 .0 0 21 .0 0 22 .0 0 23 .0 0 Time [] Wa te r lev el [m a.s .l.] Water level
Overflow level 112.50 m a.s.l. Figure 5. Daily pattern of ''San Sebastiano' reservoir water level. ' Daily consumption. It represents hourly variation of flow supplied from the junctions (Fig. 6). The consumption pattern was deduced by making the following hypothesis: 9 the average input flow was 340 l/s (corresponding to a daily input volume of 29.422 m3); 9 the total water load supplied by the junctions, according to the billed data, was 113 l/s, corresponding to 9.790 m3/d. The network demand was divided among
the supplying junctions according to authorized and billed consumption
collected by ARIN S.p.A.; 9 the total losses were 227 l/s. The 30% (more or less equal to 68 l/s) was considered apparent losses, assuming a specific per-capita consumption of 240
liters/day. This flow was equally divided among the supplying junctions.
Consequently, physical losses were assumed to be 70% (159 l/s). Spatial distribution of the network maintenance operations in 2005 shows a quite homogeneous distribution. Therefore it was assumed to divide physical losses among all the
system junctions (supplying junctions and leaking junctions). Consequently, the following
relationships were adopted to evaluate the supplied flows qtj in the generic junction j, in the instant t: - supplying junctions: α tj jSUP t tj p C q c q '' + '' = (2) - leaking junctions (emitters): α tj tj p C q '' = (3) where: - qSUP: total billed flow and apparent losses; - ct: hourly coefficient; - C, α: ''emitter coefficient' and ''emitter exponent' (see eq. (1)). 0,0000 0,2000 0,4000 0,6000 0,8000 1,0000 1,2000 1,4000 1,6000 0. 00 - 1. 00 1. 00 - 2. 00 2. 00 - 3. 00 3. 00 - 4. 00 4. 00 - 5. 00 5. 00 - 6. 00 6. 00 - 7. 00 7. 00 - 8. 00 8. 00 - 9. 00 9. 00 - 1 0. 00 10 .0 0 - 1 1. 00 11 .0 0 - 1 2.0 0 12 .0 0 - 1 3. 00 13 .0 0 - 1 4. 00 14 .0 0 - 1 5. 00 15 .0 0 - 1 6. 00 16 .0 0 - 1 7.0 0 17 .0 0 - 1 8. 00 18 .0 0 - 1 9. 00 19 .0 0 - 2 0. 00 20 .0 0 - 2 1. 00 21 .00 - 22 .0 0 22 .0 0 - 2 3. 00 23 .0 0 - 2 4. 00 Time (Time Period = 1 hrs) Ho ur ly Coe ff ic ien t average estimated pattern
measured pattern 24/05/2006 Figure 6. Consumption daily pattern. According to recent literature (Khaled et al., 1992; May, 1994), it was estimate α = 0.80. Coefficient C was assumed to be constant for all the system junctions, and when calibration
was completed, a value C = 0.02 was returned. This value allowed a reliable simulation of the average daily pattern (Fig. 6). Calibration returned, moreover, reliable values of the pipe
roughness coefficients.

4 THE DISTRICT METERING PROCEDURE After the network was calibrated, several hypothesis of designing and implementing DMA to reduce physical losses were performed, according the following procedure: 1. District Meter Areas and the transmission mains connecting them to the network were identified. Each District Meter Area had similar elevation, few connections with
the remaining part of the network, and served a population of 10.000÷15.000. 2. Implementing each zone with intercepting valves and/or pressure reducing valves. 3. Preliminary analysis of the district metered network was performed to provide adequate operating pressure of the system. A minimum head pressure of 25 m was
guaranteed with peak demands, as 5-6 stories buildings were present. The network
served as well 8 insulated tower buildings, which were supposed to be served with
their own on site pumps. 4. Many numerical simulations were performed to guarantee adequate head pressure for: - daily demand, especially for peak hours demand;
- break of transmission mains;
- fire hydrant service. Peak demands due to fires were estimated by assuming a concentrated flow of 30 l/s, requiring a minimum head pressure of 5 m. - chlorine residuals analysis, by simulating the transport and decay of chlorine through the network. Initial chlorine concentration was assumed 1 mg/l and
checking for a minimum chlorine residual of 0.2 mg/l, as provided for Italian laws.
To simulate chlorine decay through the network, the reactions occurring in the
bulk flow were modelled with first order kinetic. The instantaneous rate of reaction
was assumed to be concentration-dependent according to the following equation: ) ( ) ( t c K dt t dc b '' = (4) where c is the reactant (chlorine) concentration and Kb a bulk reaction rate coefficient (in h-1 or days-1). Kb is inversely proportional to initial chlorine concentration c0 and it can be estimated according to the empirical relation provided by Fang et al. (1999): b c a K b + '' 0 (5) Coefficients a and b in equation (5) depend on the water quality and can be
estimated by performing a bottle test in the laboratory. Numerical simulations
were performed assuming a value of 0.55 days-1 for Kb, as suggested by Rossman et al. (1999) after laboratory experiments.
Table 3 summarizes minimum head pressures and water losses reduction after the numerical simulations for each district metering implementation were performed. Six
different scenarios were analyzed, in which the number of District Meter Areas was
progressively increased. For each scenario Table 3 shows the number of intercepting and
pressure reducing valves and the physical losses reduction. All the demand simulations previously described (peak hour demand, break of a transmission main, fire hydrant service) provide a minimum pressure head of 25 m and a
minimum chlorine residual of 0.2 mg/l. For example, Figure 7 shows pressure head at the
junctions during peak hour demand, both in the current configuration and after six DMA were implemented, with a significant reduction of pressure head in the lowest part of the network.
Numerical simulation showed as well head pressure fluctuation due to daily demand
variation, so providing a preliminary estimate of physical losses reduction after DMA design.
Simulations analysis showed that the best performance can be achieved by scenario 6, which
provides for six DMA by placing 14 intercepting valves and 9 pressure reducing valves. (a) (b) Figure 7. Pressure head at the junctions during peak hour demand in the current configuration (a) and after DMA design (b).
This configuration should provide a physical losses reduction of 34%, corresponding to 16% of system input volume (i.e. 4665 m3/d and 1.70 million cubic meters per year). For sake of clarity, Figure 8 shows the network map with displayed the DMA, the minimum head pressure junction and the reduction of the system input volume.
Number of intercepting valves Number of pressure reducing valves 1 2 2 A+B J230: 25,12 m 6,60% 14,11% 2 6 3 A, B J230: 25,15 m 8,33% 17,81% 3 10 5 A, B, C A5: 25,75 m 11,94% 25,54% 4 10 6 A, B, C, D A6: 25,00 m 13,37% 28,59% 5 11 7 A, B, C, D, E A7: 25,00 m 13,96% 29,85% 6 14 7 A, B , C, D, E, F J119: 25,17 m 16,05% 34,31% SC E N ARI O DMA DESIGN System input volume reduction Physical loss reduction DM A Minimum head pressure WATER SAVING Table 3. Summary of ''Napoli Est' DMA design. 100 0 C AP (Ex Cas mez ) 10 0 0 C AP (E x C as m ez) 100 0 C AP (Ex Ca sme z) 10 00 C AP (E x C asm ez) 10 00 C AP (E x C asm ez ) 10 00 CAP (Ex Ca sme z) 10 0 0 C AP (Ex Cas mez) 10 0 0 C AP (E x C asm ez) 100 0 C AP (Ex Cas me z) 100 0 C AP (Ex Ca sme z) 100 0 C AP ( Ex C asm ez) C C DMA A DMA B DMA C DMA F DMA E DMA D SCENARIO 6 (6 DMA) System input volume reduction: 16,05% Physical loss reduction: 34,31% Minimum pressure head junction (p = 25,17 m) Figure 8. ''Napoli Est' DMA design.
Coupling the DMA design with the existing water flow remote measurement system will provide the following benefits: ' remarkably reducing water losses; ' reducing the number of maintenance operations, after reducing the operating pressure of the system; ' easy finding the most vulnerable areas of the network; ' making quick maintenance and repair in the network. Obviously, results of numerical simulations are only preliminary, and should be confirmed in the future by implementing and monitoring one or more DMA and comparing
numerical simulations with in situ measurements. REFERENCES
Artina S., Bragalli C., Giunchi D., Liserra T., ''Stima delle perdite idriche proposta dal DM 99/97 e da IWA con monitoraggio in telelettura ', Atti del I Convegno Nazionale di Idraulica Urbana-Acqua e Città, Sorrento, 2005. Cheong L.C., International report on unaccounted for water and economics of leak detection, IWA World Conference Budapest, October 1993. Comitato per la Vigilanza sull''uso delle risorse idriche, Lo stato dei servizi idrici. Anno 2004.
D.M. 99/97, Regolamento sui criteri e sul metodo in base ai quali valutare le perdite degli acquedotti e delle fognature , G.U. 18.04.1997, n. 90. Fang H., West J.R., Barker R.A., Forster C.F., Modelling chlorine decay in municipal water supplies , Water Research, 33, 12, 1999. Giugni M., Romanelli D., Fontana N., Corrado V., Primo approccio alla distrettualizzazione nel sistema idrico di Napoli , Atti del I Convegno Nazionale di Idraulica Urbana-Acqua e Città, Sorrento, 2005. Hamberg D., Shamir U., Schematic Models For Distribution Systems Design, Journal of Water Resources Planning and Management ASCE 114(2) 129-141, 1988. IWA Conference, Workshop on Global Perspectives on Managing Water Security, Melbourne, 2002. Khadam M. A., Shammas N. K., Al Feraiheedi Y., Water losses from Municipal Utilities and their Impacts , Water International, n° 16, 1991. Khaled H., Sendil U., Relationship between pressure and leakage in a water distribution network , Proceeding of the AWWA Conference, 1992. Lambert A., What do we know about Pressure-Leakage Relationship in Distribution Systems, Proceeding of the AWWA Conference on System Approach to Leakage Control and Water
Distribution Systems Management, 2000. May J., Pressure dependent leakage, World Water and Environmental Engineering, 1994.
Milano V., Dipendenza delle perdite di una tubazione dalla pressione di esercizio, L''Acqua n. 4, 2006. Rossman L.A., Epanet 2, Users manual, U.S. Environmental Protection Agency, Cincinnati, OH45268, EPA/600/R-00/057, September 2000. Rossman L.A., Grayman W.M., Scale-model studies of mixing in drinking water storage tanks , Journal of Environmental Engineering, Vol. 125, N. 8, 1999.

© Eiom - All rights Reserved     P.IVA 00850640186