Selection and Concentration of Pipeline Mains for Rehabilitation and Expansion of Water Distribution Network

Available online at www.sciencedirect.com

ScienceDirect Procedia Environmental Sciences 28 (2015) 732 – 742

The 5th Sustainable Future for Human Security (SustaiN 2014)

Selection and concentration of pipeline mains for rehabilitation and expansion of water distribution network B. Bakria*, Y. Araib, T. Inakazub, A. Koizumib, H. Yodac and S. Pallua a

Department of Civil Engineering Hasanuddin University, Perintis Kemerdekaan KM VII, Makassar 90245, Indonesia Department of Civil and Environmental Engineering Tokyo Metropolitan University, 1-1 Minami Osawa, Tokyo 192-0397, Japan c Geoplan Co., Ltd., Tokyo-Japan

b

Abstract This paper proposes a strategy for rehabilitation and expansion of water distribution network (WDN) in developing countries. The main framework of the pipe network rehabilitation is formulated based on the “selection and concentration” of trunk/limb mains pipe strategy. The strategy is to select more appropriate pipeline, diameter and material of pipe mains to ensure cost effectiveness and adequate water pressure. Meeting the objective of the strategy, this study developed Hybrid Genetic Algorithm (HGA) model. HGA-I is applied considering rapid growth of the future water demand. HGA-IIa focuses on selection and concentration of trunk/limb mains pipelines, and HGA-IIb is applied to determine appropriate diameters for water demand in interim plan years. Life Cycle Cost Analysis (LCCA) is applied to identify the potential cost of a project alternative. This analysis proposes the most effective combination of pipe material and diameter of each pipe main. The application of this method may be appropriate for water authorities in planning rehabilitation and expansion of a WDN. They can easily clarify the optimal combination of pipeline, diameter, and material of trunk/limb mains pipes in the network that will result in the minimum life cycle cost over the entire upgrading project period. Published by Elsevier B.V This is an open access article under the CC BY-NC-ND license

© 2015 The Authors. Published by Elsevier B.V. (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of Sustain Society. Peer-review under responsibility of Sustain Society Keywords: Selection; concentration; trunk/limb main pipe

* Corresponding author. Tel.: +6285395956837; fax: -. E-mail address: [email protected]

1878-0296 Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of Sustain Society doi:10.1016/j.proenv.2015.07.086

B. Bakri et al. / Procedia Environmental Sciences 28 (2015) 732 – 742

733

1. Introduction Since most of the population growths are in developing countries, the number of people who lack access to safe water is likely to increase. Even today, the provisioning of water supply facilities is not sufficient to meet the increasing water demand with rapid population growth and the advance of commerce and industry. Developing countries are confronted by two main problems relating to the water supply system, namely, how to optimize the distribution network for achieving cost-effectiveness and at the same time how to meet increasing water demand (Vairavamoorthy, 2008). In other hand, proper selection of material pipe is also one of the factors for succeeding the rehabilitation and expansion both in terms of the cost and technical. In many cases, the pipelines are being installed to meet the water demand in conformity with a standard target year of projects (normally, 10-15 year target). It is often the case that the diameter of the pipes installed is no longer sufficient after the project and needs to be replaced. A method is required to identify the potential cost of pipe design alternatives, including those for initial pipe installation, replacement, and leak repair. To deal with these issues, this research develops a new approach for rehabilitation and expansion of the WDN. This study intends to show that rehabilitation and expansion of water distribution network process is more effective, in term of hydraulic and cost, when making “selection and concentration” of trunks/limb main pipe. Achieving the objective of the concept, Genetic Algorithm (GA) is developed to determine not only the most effective pipe diameter but also proper selection of pipeline mains. In other words, GA in this paper is applied to select the most effective pipeline mains to be rehabilitated and to discard others while at the same time searching the optimal diameter solution for the pipeline to ensure the costeffectiveness and adequate water pressure at each node (Bakri et al. 2015). The first, Hybrid Genetic Algorithms I (HGA-I) is applied, considering rapid growth of the future water demand. The second, HGA-IIa, focuses on selection of trunk/limb mains pipelines, and HGA-IIb is to determine appropriate diameters for water demand in interim plan years. Life Cycle Cost Analysis (LCCA) is applied to select the most effective combination of pipe material and diameter of each pipe main. 2. Current State of and Issues for Pipe Network in The Target Area To clarify the effectiveness of the concept, this study is conducted in Makassar, capital of South Sulawesi Province in the eastern part of the Republic of Indonesia. The area is developing rapidly as a center of administration, industry, commerce, and education in East Indonesia. Its population in 2010 was 1,339,374 people, with an annual growth rate of 2.2% (2004-2010), much higher than Indonesia’s 1.6% national average growth rate (BPS-Statistics Indonesia, 2011). The expansion of water supply facilities has not been able to keep pace with the advance of urbanization. Many people, particularly in the north and west part of the city, suffer from chronic water shortage and low water pressure (0-0.5 kg/cm2). It may be caused by the inadequate alignment of distribution pipelines due to the rapid increase in customers and high rate of non-revenue water (NRW), 45% in 2010. Many old pipes in the central area installed in the 1920s are still in use, while smaller mains of the PVC pipes are installed in the newly developing areas, which at 100-150 mm, are not sufficient to achieve stable and continuous water supply. The water supply system in this area is serviced by PDAM Makassar, it is city city-owned waterworks and one of twelve drinking water companies classified into a large group in Indonesia, which has over 100,000 customers. The PDAM, although recognizing the urgent need for rehabilitation and reinforcement of the existing distribution pipe network, is not capable of allocating sufficient funds. Thus in undertaking the rehabilitation of water pipelines in the study area, replacement or reinforcement of the existing distribution pipe network at minimum cost and meet future water demand increase is considered essential for overcoming all of the problems above. This study intends to focus on the analysis of the Somba Opu Distribution (SOD) system, one of two major distribution systems in Makassar. Its service area includes the city center (commerce and administration) and newly developing residential areas. Facilitating this study, the network target is adjusted as shown in Figure 1. It consists of one storage reservoir, 43 nodes, and 77 pipelines. The nodes in the target area are on mostly flat terrain. Water is distributed by gravity flow downward from the distributing reservoir 49 (m) above mean sea level. In this study, we consider a pipeline upgrading plan for the target area over a project period of 80 years (2090) starting in 2010 as the base year, divided into four periods of 20 years. Figure 2 summarizes the current water demand in 2010 and its forecasting in 2030, 2050, 2070 and 2090.

734

B. Bakri et al. / Procedia Environmental Sciences 28 (2015) 732 – 742

3. Methodology 3.1. Proposing Strategy In the ideal water distribution system, pipes are laid out suitably with trunk mains having the largest diameter, followed by limb mains, and then service mains as shown in Figure 3. In the case of the distribution network in the study area, however, such a clear distinction is not built into the network design concept. The problem is that trunk and limb mains do not have a large enough diameter. The hydraulic problem arising from this situation is the large friction head loss in pipes at the upstream portion, so that sufficient water pressure cannot be obtained in downstream pipes. In devising a pipeline upgrading plan for the target area, it will be necessary to change the existing inappropriate pipe diameters. Of particular importance from the standpoint of meeting the higher demand volumes expected in the future and solving the inadequate pressure problem will be to incorporate in the plan the priority provisioning of trunk and limb mains. Replacement of all pipe mains for rehabilitation and expansion is maybe ineffective way and requires sufficient budget for rehabilitation. In undertaking the appropriate reinforcement of water pipelines, “selection and concentration” strategy is needed from the following two standpoints. One is to satisfy the hydraulic constraints of the WDN dealing with future water demand, and the other is to ensure cost effectiveness. This strategy may be more effective than, and hydraulically preferable to, replacement of all mains for rehabilitation. Judging from this strategy, it is necessary first to identify trunk/limb mains on the network and then to seek a solution for proper diameter of mains to improve the overall network (Bakri et al. 2013). To find the optimal solution from the many possible alternatives for pipeline mains in a network, we applied Genetic Algorithm (GA). The next step, rehabilitation or reinforcement of the pipelines is then carried out selectively based on result of the GA. In this study, the pipes that are considered as trunk and limb mains with their major role in the distribution network are therefore defined as a main pipeline network. The pipelines that have hydraulic and economic impacts shall be prioritized in the reinforcement process. 3.2. Genetic Algorithm (GA) GA has been used successfully in optimal design of pipe distribution network in across the globe. GA was developed by John Holland (1975) at the University of Michigan. The main approach of GA is to minimize cost, subject to hydraulic constraints. In contrast with traditional method where the design of WDN was based on experience of planner or engineer, a GA searches the optimal solution for the network based on natural selection and the mechanism of biological background (Goldberg, 1953). Simpson et al. (1994) presents a methodology for optimizing pipe networks using GA and investigates a three-operator GA comprising reproduction, crossover, and mutation. Frey et al. (1996) applied GA to minimize capital and/or life cycle costs for design and operation of WDN. Furthermore, Savic and Walters (1997) describe the development of a computer GA model to the problem of leastcost design of WDN. Their studies show that the GA is effective in finding global optimal or near-optimal solution with required only a relatively small number of evaluations. The optimization model used in this methodology is a Hybrid Genetic Algorithm (HGA) model, which builds a pipeline network analysis program into the GA (Arai et al., 2009). In order to identify main pipeline networks in the target water distribution system, an optimization problem is formulated in which all pipe diameters are decision variables. The objective function (TC) is the total of pipe material costs for rebuilding pipeline network (assuming use of ductile cast iron pipe (DCIP) as the pipe material). The material costs of each pipeline are expressed as the product of pipeline total length Li (m) and unit laying cost c (IDR: Indonesian Rupiah) based on diameter Di (mm). Then, the planning problem for determining the optimal diameter Di of each pipe i is formulated as follows, where the constraints are flow velocity Vi (m/s) in the pipes and effective water head Hj (m) at each node j.

Minimize Tc

n

¦ c  D , L .....(1) i

i

i 1

subject to hydraulic constraints,

735

B. Bakri et al. / Procedia Environmental Sciences 28 (2015) 732 – 742 

:DWHU'HPDQG/V 

Hj ≥ HjMin …..(2) Vi ≥ Vi Max …..(3)

41

Ø 400 (30)

2300 (8) Ø 300

(23) 900 Ø 400

71 0 (2 7) Ø3 50

Ø 350

Ø 200

5 ) (6 Ø

20

0

90 29

(12) 10 400 Ø 100 2 0

0

1150 Ø 150 (15) 12

(10) 8 490 Ø 500

1638 Ø 500

950 (1 3) Ø 900

9

6

(9) 85 14 0 0 Ø3

(11)

80 ( 19 Ø3 ) 50

) (7

15

Ø 350

(17)

Ø 350 (20)

(21) 150 Ø 400

1450 Ø 400 (1 8) 1600 (14)

(22)

Ø 400 (34)

Ø3 5 (25 0 )

(26 68 ) Ø4 0 00

Ø 400 (32)

500 Ø 350 (33) 350 1510 (36)

Ø 400 530 Ø 400 (37)

Fig. 2 Node water demand

1665 (4) Ø 200

1

Ø 400 (16)

3 (5) 630 Ø 200

500 (31) Ø 350

Ø 200 (74)

(38) Ø 350

16

00

840 (4 9) Ø 40 0

Node

7

1600 (24) Ø 200

(3 )

31

(53) 800 Ø 300

                          

0

66 Ø 2 0 (5 5 0 4)

11

15



17

Ø1

32

(50) 100 00 )3 50 29 (52 Ø 1 Ø 350 28 0 30 Ø 30 ) (51

(28 )

80

(48) 770 Ø 350

33



14

13

Ø 150 (45)

500 (62) 34 Ø 250

90

50

23

45 24 36 Ø 2 0 00 (43 Ø3 ) (41) 50 (44 26 Ø 35 0 ) (39) 200 25 Ø 350 Ø 350 (4 2) (40) 1460 650 Ø 700 Ø 150 (56) 27 Ø 300 (57) 820 Ø 350 (46)

(67) Ø 200 Ø 250 (60)

500 (61) Ø 150

37

500 (66) Ø 300

50 Ø 2 8) (5

1100 (55) Ø 150

35

66 Ø 2 0 (6 0 0 4)

(59) 380 Ø 150

800 (63) Ø 150

38

Ø2

159 0 Ø 20 (35) 0

 

18 19

22

980 (70) Ø 150

Ø 250 (68)

1230 Ø 300 (69)

39

Ø 350 (73) 1070

Ø 150) (47

Ø 250

40







1010 (29) 19 Ø 200

21

Ø 250 (71)

42

(72) 400 Ø 300

2630 (65) Ø 150

(76) 1070

20



14 8 Ø2 0 00

) 77 80 ( 50

Ø1

17

2133 Ø 300 (75)

43





938 (2) 4 Ø 200 50 00 Ø1 10 0

(1)

WT P

Fig. 1 Somba Opu Distribution System (network study)

Fig. 3 Conceptual Fi C t l pipeline i li li system t

As HGA fitness function (FV), the reciprocal of objective function Tc in equation (1) is used. FV is calculated as inverse of the total network cost (Tc). Pipeline network analysis is necessary for guaranteeing that each of the candidate solutions obtained by HGA application satisfies the constraints of equation 2 and 3. A penalty function is set so that solutions that do not satisfy the constraints have their fitness reduced to 1/10 each time they violate a constraint. In the following application, the hydraulic constraints concerning pipes and nodes are set to VMax =3.0 (m/s) and HMin =17 (m). A binary number is generated as the initial population of GA. This study takes a 2-bit binary number such as 0 and 1. Diameter pipe candidates Di for each pipeline i based on the present diameter, with 16 pipe diameter options for the diameter changing stage, are shown in Table 1. 3.3. Selection and concentration process The aim of selection and concentration process is to determine suitable diameters that will meet future water demand while minimizing cost needed for upgrading pipelines and meeting the hydraulic constraints. To this end, the study applied three types of HGA model based on the diameter options as shown in Figure 4 and Table 2. Of these, the objective of applying HGA-I, focusing on the final goal of 2090, was to determine sufficiently large pipe diameters so as to handle the growing water demands of the future. The solutions that could be obtained by HGA-I were accordingly aimed primarily to increase the sizes of all pipes as shown by the diameter options in Table 2. Based on this approach, diameters that are suitable for trunk and limb mains could be obtained. HGA-IIa was used to apply a process of selection and concentration on the proposed plan obtained by means of HGA-I (Bakri et al., 2015), from the standpoint of improving cost performance of the plan. It induced the selection of pipes with one rank lower diameter for pipelines that could be downsized, or of the smallest diameter (150 mm) among existing pipes, for the sake of lowering costs. If 150 mm was selected in the first round of application (FVi-1),

736

B. Bakri et al. / Procedia Environmental Sciences 28 (2015) 732 – 742

a pipeline was determined not to have high importance as a main pipeline. Pipes of the smallest diameter (150 mm) were then excluded from the second round of HGA-IIa application (FVi), and application was repeated until a main pipeline network was obtained with the highest FV. Note that 150 mm pipes included in the results of HGA-I application were likewise excluded from the candidates for main pipelines in subsequent iterations. Since number of pipes would end up being reduced as some pipes were discarded, it was necessary to include a larger diameter than existing as one of the diameter options thus this process was also to select the diameters that were same or one rank larger obtained by HGA-1. The purpose of applying HGA-IIb was to determine appropriate diameters for interim plan years. With the HGAIIa result as starting point, these results would be used in studies feeding back the results in order from the distant future (t period) to the near future (t-1 period), that is, from 2090 to 2070, from 2070 to 2050, and from 2050 to 2030. Since the water demand volume was assumed to rise on a steady curve, such that the t-1 period volume would be less than the t period volume, downsizing diameter candidates would include the option of maintaining the same diameter, as shown in Table 2. The GA parameters set in HGA-I were population 2000, generations 3000, crossover value 0.03, and mutation rate 0.8, while in HGA-IIa and IIb were population 1300, generations 3000, crossover value 0.03, and mutation rate 0.8. Table 1. Diameter and cost stage (Cost in IDR/m x 10^6) No Dia. (mm) Cost No Dia. (mm)

Cost

No

Dia. (mm)

Cost

No

Dia. (mm)

Cost

1

150

0.644

5

350

3.555

9

600

10.535

13

1000

29.500

2

200

1.151

6

400

4.653

10

700

14.374

14

1100

35.748

3

250

1.804

7

450

5.899

11

800

18.814

15

1200

42.601

4

300

2.605

8

500

7.295

12

900

23.856

16

1300

50.060

Table 2. Diameter options and binary numbers

Fig. 4 HGA methodology

Pipe diameter candidates

Current pipe

Option 1

Option 2

Option 3

Option 4

HGA-I

Present diameter

1 up stage

2 up stage

3 up stage

HGA-IIa

Min. diameter (150 mm)

1 down stage

Present of HGA I or previous of HGA II

1 up stage

HGA-IIb

Previous of HGA IIa or Present of HGA IIb

1 down stage

2 down stage

3 down stage

11

10

10

00

Binary number

B. Bakri et al. / Procedia Environmental Sciences 28 (2015) 732 – 742

737

Fig. 5. Pipe cost per meter based on diameter material pipe

3.4. Life Cycle Cost Analysis (LCCA) Life Cycle Cost Analysis (LCCA) in this paper was performed to assess design alternatives over a particular time frame. It is a tool that can be used to identify the potential cost of project alternatives (Bonton et al., 2012). In the analysis, overall costs of the pipe design alternatives, including those for initial pipe installation, replacement, and leak repair to be required up to the year 2090 were estimated with time escalation namely every 20 years. Minimizing the effects of price escalation, the present study assumed a social discount rate of 0% in the future. The calculations of LCCA in this study covered the following four items: material cost (Cm), civil work cost (Cc), leakage cost (Cl), and demolition cost (Cd) (Bakri et al., 2012). Pipe installation costs included those for materials and civil work. Three pipe materials were considered, DCIP (Ductile Cast Iron Pipe), HDPE (High Density Polyethylene Pipe), and PVC (Polyvinyl Chloride Pipe). Pipe materials of PVC, HDPE, and DCIP were assumed in the current research to have a life span of 20, 40, and 80 years respectively. Material costs depend on the pipe materials and sizes, as shown in Figure 5. while civil work cost per meter was assumed based on local data to be around 30 percent of material cost and was calculated using Equation 4. ‫ܥ‬௖ ൌ ͲǤ͵ ൈ ‫ܥ‬௠ ǥ ሺͶሻ The number of accidents for DCIP pipe was assumed to take place after installation at a frequency of 0.5 points/km/year in the period of 20 years before its life span expires. HDPE and PVC were 2 points/km/year and 5 points/km/year, respectively. Note that leaks were assumed to occur starting in the second year after laying PVC and HDPE, and starting in the 61st year after laying DCIP pipes. Average costs required for a leakage repair per meter of each material pipe was assumed 2 times the total of material costs and civil work costs as shown in Equation 5. ‫ܥ‬௟ ൌ ʹ ൈ ሺ‫ܥ‬௠ ൅ ‫ܥ‬௖ ሻ ǥ ሺͷሻ Demolition cost of each material pipe was assumed based on local data to be around half of civil work cost, and calculated using Equation 6. ‫ܥ‬ௗ ൌ ͲǤͷ ൈ ‫ܥ‬௖ ǥ ሺ͸ሻ

738

B. Bakri et al. / Procedia Environmental Sciences 28 (2015) 732 – 742

4. Result and Discussions 4.1. Application of HGA Model 41

41

41

43

43

43 20

65

31 30 29

75 74

77

19

21

76

42

17

22

69 68

64 63 35

66

37

36

67 61 60 62

44

54

25

42 46 45

56 57

33

55

38 41 37

26

34

59 58

32

43

39

19

18 17

40

11

10

9

5

13

12

10

4

2

15 12

1

36

67 61 60 62

44

4

54

1

16

53

Figure 5a. Existing system Fig. 6a. Existing system

40

41

4

14

3

49

15 12

31 21

76

2

1

32

4

73

40

29

19

17

69

61 60 62 34

58 33

44

42

53

11

10

27

30

8 5

12

50

28

49 12

56

32

4

1

16

53

29

64

7

8 5

10

12

50

49

28

30

2

1 12

5

32

4

1

16

53

31 21

73

40

19

18

35

36 21 44

34

42

58 57 53

31

5 4

1

16

64

30

29

10

37

7

8

10

5

35

28

34

16

36 21

5

32 4

1

16

53

8

14

10

7

8 5

25

40

11

11

13 10

27

29

6

9 9

18

12 2

48 52

54

37

26

42 57

2

7

24

13

24

36

33

1 12

25 15

23

58

3

3 49

66

62

12 2

50

17

22

44

11

13

18

27

34

69

38

6

9

25 11

19

23

14

48 52

54

2

1 12

16

9

27

40

39

8

18 40

33

32

37

41

26

29

21

7

24

13

24

36

31

71

25 15

23

69

62

30

33

76

42

17

22

23

37

3

3 49

28

WTP

75

77

27

34

39

66

2

50

20

29

33

71

64

12

Figure 5f. HGA II (Iteration 4, FV = 1.776) Fig. 6f. HGA-IIa (Iteration 4, FV=1,776)

20

38

10

27

43

75 76

5

11

13

48 29

7

8

41

41

42

56

11

31 WTP

43

77

40

52

54

6

9 10

9

18

25

42

58

57

Figure 5e. HGA II (Iteration 3, FV = 1.773) Fig. 6e. HGA-IIa (Iteration 3, FV=1,773)

WTP

34

37

41

26

8

13

24

44

3

3

16

14

21

33

4

2

36

36

61 62

11

13

66

37

35

31

Figure 5d. 5d HGA II (Iteration 2, 2 FV = 1.756) 1 756) Fig. 6d. HGA-IIa (Iteration 2, FV=1,756)

10

7

24

15

23

69

38

48 52

54

25

34 23

6

9 9

11

18

17

22

71 39

14

27

57

5

2

1

1

27

8

18 40

19

33

73

40

16

25

42

58

3

3

31

34

37

41

26

76

7

24

13

24

44

33

4

2

36

61 62

11

13

48 29

25

29

21

42

15

23

36

31

75

77

17

21

66

37

35

40

52

10

9

18

64

7

25

56 57

54

38 41 37

26

19

4

16

WTP

27

34

69

38 6

9

1 12

5

2

FFigure 5c. HGA II (Iteration I, FV = 1.697) Fig. 6c. HGA-IIa (Iteration 1, FV=1,697)

WTP

23

14

3

3

49

28

31

18

8

13

24

19

22

71 39

21 20 36

29

33

73

40

16

23

36

76

7

24

15

23

4

2

14

20

31 21

42

26 25

34

12

10

51

30

5

6

11

13

27

48 47 52 29 50

7

8

41

75

77

28 27

35

37

53

40

11

10

9

18 17

39

25

42 46 45

6

9

19

43

18

22

71

66

54

1

16

8

14

13

38 41 37

26

56 57

3

5

16

23 22

24

43

20

33

39

64

58

41

75

42

34

7

24

15

36

44

33

12

43

77

32

13

35

6 2

28

51

30

5

36

67 61 60 62

8

11 10

20

35

11

10

26 25

23

21 20

66

37

7

Figure 5b. HGA I (FV = 1.617) Fig. 6b. HGA-I (FV=1,617)

43

38

9

27

48 47 52 29 50

31 WTP

19

18 17

39

25

42 46 45

59 58 56 33 57

38 41 37

26

34

32

24

17

22

69 68

64

6

9

18

28 27

35

8

14

13

19

34

39

38

36 43

5

2

23 22

73

40

16

29

33 32

72 71

7

24

15

21 20

66

37

55

3

3

49

35

6

14

31

26 25

23

69 68

64 63

8

11

27

48 47 52 29 5028 30 51

53

38

7

21

76

42

17

22

35

6

31 30

75 74

77

18

28 27

8

9

19

34

70

39

29

33 32

73

40

72 71

14

21 20 13

24

76

42

16

23 22

36

21

7

24

15

35 38

65

26 25

23

31 30

75 74

77

28 27

34

70

39

18

33 32

73

40

72 71

20

20

50

28

30

31

3

3 49

1 12

16

2

5 4

1 WT P

5g. HGA(Iteration II (Iteration 5, FV = 1.779) Fig.Figure 6g. HGA-IIa 5, FV=1,779)

WT P

Figure 5h. HGA II (Iteration 6, FV = 1.788) Fig. 6h. HGA-IIa (Iteration 6, FV=1,788)

Fig. 6 Application of HGA-I and HGA-IIa (Bakri et al., 2015)

After HGA was applied, HGA-I succeeded to meet increasing water demand, while HGA-IIa was applied six times for obtaining the main pipeline network with the highest fitness value (see Figure 6). The dotted line indicates that pipelines were eliminated. Out of 77 nodes, the solution of HGA-IIa selected 43 trunk/limb main pipes and

739

B. Bakri et al. / Procedia Environmental Sciences 28 (2015) 732 – 742

discarded 34 others (Bakri et al., 2015). Reducing the number of pipes as proposed by HGA-IIa contributed to simplify the network that can achieve a substantial reduction in labor, procurement, repair, and operation and maintenance costs. This result indicates that proper selection of pipeline of trunk/limb mains and focusing most replacement on specific pipelines is effective for ensuring cost-effectiveness and water pressure in the network. Table 3 Diameter proposed by HGA-IIb HGAHGA-IIb IIa Pipe

Pipe

(2090)

(2070)

(2050)

(2030)

1

1300

1300

1300

1300

2

450

400

350

300

3

1200

1200

1200

5

400

300

300

7

350

300

8

250

9

450

10 11

HGAIIa

HGA-IIb

Pipe

(2090)

(2070)

(2050)

(2030)

24

400

300

300

250

25

450

350

350

300

1200

27

300

250

250

250

29

400

350

300

250

250

31

200

150

200

200

200

33

600

400

400

300

34

600

500

500

400

350

36

600

500

450

400

37

12

1000

1000

1000

1000

13

1000

1000

1000

16

700

700

600

18

700

600

21

700

600

23

700

600

45

HGAIIa

HGA-IIb

(2090)

(2070)

(2050)

(2030)

52

350

300

300

250

53

350

300

250

250

250

54

300

200

200

200

300

57

500

350

350

300

150

150

58

400

350

300

300

500

400

300

62

300

300

250

250

500

400

300

64

250

200

200

150

600

500

450

400

66

300

250

200

200

700

500

450

400

69

450

350

350

300

40

800

700

700

600

71

350

300

300

250

1000

42

500

400

400

350

75

250

200

150

150

500

44

500

400

400

350

76

300

250

200

200

500

450

48

500

400

400

400

77

200

200

150

150

500

450

49

700

600

500

400

500

400

50

500

500

500

400

a 2090

2070

2050

2030

Min. Pressure

40 Presure (m)

35 30 25 20 15

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43

Node

740

B. Bakri et al. / Procedia Environmental Sciences 28 (2015) 732 – 742

b 3,5

2090

2070

2050

2030

Maximum velocity

Velocity (m/s)

3 2,5 2 1,5 1 0,5 0

1 2 3 5 7 8 9 10 11 12 13 16 18 21 23 24 25 27 29 31 33 34 36 37 40 42 44 48 49 50 52 53 54 57 58 62 64 66 69 71 75 76 77

Pipe Fig 7. Hydraulic analysis of HGA-IIa and IIb (a) pressure at each node; (b) velocity at each pipe

The results of applying HGA-IIb to the demand for the years 2070, 2050, and 2030 are shown in Table 3 along with the suitable diameters for 2090 (HGA-IIa). Note that for HGA-IIb as well, the upstream pipe such as; Pipe 1, 3, 12, and 13 was substituted in case of pipelines that would become larger in the downstream, and this diameter was adopted as the suitable diameter for each year. In this way, a WDN having economical pipeline diameters for meeting the water demands of the final plan year was derived by HGA-I, and HGA-IIa was applied in order to avoid redundancy or simplify main pipelines, thereby designing an economical trunk/limb mains network for the final plan year based on demand conditions. Furthermore, HGA-IIb, aiming at determining economical diameters for each project stage, was applied to obtain most effective and efficient plan. Hydraulic verification of HGA is summarized in Figure 7. The figure shows that solution of HGA IIa and IIb matched the hydraulic requirements both of the pressure at each node and velocity at each pipe. Judging from HGA-I, IIa, and IIb process, obtaining the solution proposed by HGA would be far more difficult and time-consuming using the common method of network design with a number solution space. The application is uncomplicated and does not require a high degree of mathematical sophistication to understand its mechanism (Savic and Walters, 1997). 5. Selection of material pipe Section 4.1 proposed diameter pipe at each project stage with assuming installation by DCIP pipe. Selecting long life cycle pipes at an early stage in the project, on the other hand, requires laying pipes with large enough diameters to meet expected future demand increases and may result in an uneconomical design. In the other hand, when pipes with short life cycle are selected, diameter shall be upgraded in phases to meet water demand. This approach deals with how to select the combination of pipe material and diameter that will result in the minimum life cycle cost (LCC) over the entire upgrading project period using the LCCA equations above. After studying the optimal combination of pipe materials for each pipeline in the main pipeline network (the 43 pipelines determined in 4.1 above), the 42 pipelines other than pipeline 1 can be combined in any of the three ways indicated in Table 4 from Case 1 to 3.

741

B. Bakri et al. / Procedia Environmental Sciences 28 (2015) 732 – 742 Table 4. Result of LCCA analysis

Case

Pipe

Installation time

Material 2010-2030 (20)

1

3,11,12,13,16,18,21 23,36,37,40,49,50

2030-2050 (40)

DCIP

2

2,9,10,25,33,34 44,48,57,69

HDPE+HDPE PVC+PVC+HDPE

DCIP

3

5,7,8,24 27,29,31,52 53,54,58,62 64,66,71,75 76,77

D 1200 = 29417.4

D 1200 = 32363.8

D 350 = 2163.0 D 350 = 1179.0

D 350 = 6970.0

61781.2

5314.6

D 450 = 5314.6

D 300 = 911.6

Total Cost 45147.4

D 1200 = 45147.4

DCIP HDPE+HDPE

2050-2070 (60)

2070-2090 (80)

D 450 = 3302.7

5465.7

D 450 = 3302.7

5393.3

6970.0

HDPE+HDPE

6629.9

HDPE+PVC+PVC

6674.2

PVC+PVC+HDPE

7206.7

PVC+PVC+PVC+PVC

7251.0

This table shows a typical pipeline combination for each of the three cases and the total LCC (though it must be noted that the entire combinations for each pipe with the minimum LCC in each case will not necessarily be those shown in the table). From these typical pipeline combinations, we can see that in Case 1 (Pipe 3) and Case 2 (Pipe 25), upgrading with large-diameter DCIP pipes has the greatest benefit for overall life cycle. In Case 3 (Pipe 7) on the other hand, upgrading to HDPE with a life cycle of 40 years is the most effective choice. From the table we can conclude that consideration of LCC influences the optimal combination of pipe materials for minimizing total LCC in the project period for each pipeline.

6. Conclusions This paper deals with how to rehabilitate and expand the WDN in developing countries. The main framework of the pipe network rehabilitation was formulated based on the “Selection and Concentration” trunk/limb mains pipe strategy. The strategy reveals how to select more appropriate pipeline mains and diameter of pipe mains to ensure cost effectiveness and adequate water pressure at each node. This study considered a pipeline upgrading plan for the target area over a project period of 80 years (2090) starting in 2010 as the base year, and divided into four periods of 20 years. Firstly, its focus on the final goal of 2090 is to determine pipeline and diameter of trunk/limb mains pipes. The solution of 2090 will be used in studies feeding back the results in order to determine pipe diameter in 2070, 2050, and 2030. Meeting the objective of the strategy, this study developed HGA model. HGA-I was applied, considering rapid growth of the future water demand. HGA-IIa focused on selection and concentration of trunk/limb mains pipelines, and HGA-IIb was applied to determine appropriate diameters for water demand in interim plan years. LCCA was applied to identify the potential cost of a project alternative, covering the following four items; 1) material cost, 2) civil work cost, 3) leakage cost, and 4) demolition cost. This analysis shows how to select the combination of pipe materials and diameter pipe that will result in the minimum life cycle cost over the entire

742

B. Bakri et al. / Procedia Environmental Sciences 28 (2015) 732 – 742

upgrading project period. Selection and concentration trunk/limb mains pipe (HGA-IIa) selected 43 trunk/limb main pipes and discarded 34 others. The water pressure at each node and velocity at each pipe of HGA solution matched the requirement standard. Reducing the number of pipes as proposed by HGA model also contributed to simplify the network that could achieve a substantial reduction in labour, procurement, repair, and operation and maintenance costs. This result indicates that proper selection of pipeline of trunk/limb mains and focusing most replacement on specific pipelines is effective for ensuring cost-effectiveness and water pressure in the network. In addition, the application of this method may be appropriate for a water authority in planning rehabilitation and expansion of a WDN. They can easily clarify the optimal pipeline, diameter, and the combination of material and diameter pipe in the network for minimizing total LCC. References 1. Arai Y. Koizumi A. Inakazu T. Koo JY. Optimization model for water distribution network considering minimization of total replacement cost and stabilization of flow velocity in pipelines. Desalination and water treatment 2009; (2), 45-51. 2. Bakri B. Arai Y. Kawamura A. Pallu S. Yoda H. Life cycle cost analysis of pipe network rehabilitation and expansion in Makassar-Indonesia. The Proceeding of the IWA International Symposium on Water Supply Technology Yokohama-Japan, 2012. 3. Bakri B, Arai Y, Inakazu T, Koizumi T, Pallu S, Yoda H. Optimal design of a trunk/limb mains reinforced (TMR) pipe network using a genetic algorithm. The Proceeding of the 5th IWA-ASPIRE Conference Daejeon-Korea 2013. 4. Bakri B, Arai Y, Inakazu T, Koizumi T, Pallu S, Yoda H. A multi-step genetic algorithms model for ensuring cost-effectiveness and adequate water pressure in a trunk/limb mains pipe system. Journal of Water Supply; Research and Technology – AQUA, 2015; 64.2 5. BPS-Statistic Indonesia. Makassar dalam angka 2011 (Makassar in 2011). Report of BPS-Statistic Indonesia 2011. Available from: http://www.bps.go.id/eng/ 6. Bonton A., Bouchard C., Barbeau B. and Jedrzejak S. (2012) Comparative live cycle assessment of water treatment plants. Desalination. 42-54. 7. Frey JP. Simpson RA. Dandy CG. Murphy JL. Genetic algorithm optimization; Its application to design and operation of water distribution systems. Computer Conference American Water Works Association, 1996. 8. Goldberg DE. Genetic algorithms in search, optimization and machine learning. Addision-Wesley Professional (Ed.), Boston, USA. 1953. pp. 1-25. 9. Holland J. Adaptation in natural and artificial systems. The University of Michigan 1975. 10. Research team at global water market intelligence (GWI). Global Water Market Report 2011. Media Analytics Ltd, Oxford, United Kingdom; Volume I-III, 2011. 11. Savic DA. Walters GA. Evolving sustainable water networks. Hydrological Sciences Journal, 1997; 42(4), 549-563 12. Simpson RA. Dandy GC. Murphy JL. Genetic algorithm compared to other technique for pipe optimization. Journal of Water Resources Planning and Management, 1994; 423-443. 13. World Bank Group. The Millennium Development Goals Report. 2012. 14. Vairavamoorthy K., Gorantiwar S.D. and Pathirana A. (2008) Managing urban water supplies in developing countries – climate change and water scarcity scenarios. Physics and Chemistry of the Earth, 33 (2008) 330-339.