Development of an asset lifetime model for 2 distribution network management 3

Aging infrastructures 10 provide and maintain a specific level of service to consumers. In this regard, efficient replacement 11 polices are needed. This paper proposes a method for improving renewal efficiency in water 12 supply systems through a reliable asset lifetime model that will lead investments to those elements 13 with greater impact in service provision to the end user. As uncertainties are minimized, the 14 likelihood of failure will be more accurate and renewal investm 15 Therefore, the failure predictor model has been built in a reliable manner through collected data 16 from Madrid distribution network which comprises more than 17.000 km with over 400.000 water 17 pipes. It is based on the statis 18 gathered during four complete years. Additionally, detailed information from more than 4.400 19 disturbance events was recorded through field visits and laboratory essays of soil and pipe 20 materials when failures were repaired. Examination of such large series of data recorded allows a 21 better understanding of explanatory factors of failures. It is an essential step for building a 22 consistent asset lifetime model. According to this model, a renewa 23 on the risk of service disturbance and involved costs. It supports planning and operation decisions 24 reaching failures reduction and system resiliency improvement. 25


29
Providing an appropriate level of service is the primary goal for water suppliers in urban water 30 distribution systems. Even though a certain level of service is defined by regulatory standards, 31 service responsible agents use to set their own commitments of service with thresholds that go 32 beyond such specifications for improving customer satisfaction level. In this regard, the level of 33 service, considered as the guarantee of supply at every property, ca 34 continuity of service, pressure and quality provided to the end user.

35
On the other hand, service suppliers make remarkable efforts to establish suitable asset 36 management policies. Aging infrastructures require increasing asset inv 37 level of service. So that, developing a cost effective asset renewal and replacement strategy is  Aging infrastructures maintenance results in increasing asset investment in order to provide and maintain a specific level of service to consumers. In this regard, efficient replacement polices are needed. This paper proposes a method for improving renewal efficiency in water upply systems through a reliable asset lifetime model that will lead investments to those elements with greater impact in service provision to the end user. As uncertainties are minimized, the likelihood of failure will be more accurate and renewal investments will become more efficient. Therefore, the failure predictor model has been built in a reliable manner through collected data from Madrid distribution network which comprises more than 17.000 km with over 400.000 water pipes. It is based on the statistical analysis of historical data from over 55.000 system failures gathered during four complete years. Additionally, detailed information from more than 4.400 disturbance events was recorded through field visits and laboratory essays of soil and pipe rials when failures were repaired. Examination of such large series of data recorded allows a better understanding of explanatory factors of failures. It is an essential step for building a consistent asset lifetime model. According to this model, a renewal strategy is proposed. It based on the risk of service disturbance and involved costs. It supports planning and operation decisions reaching failures reduction and system resiliency improvement.
Asset lifetime model, system Failure, service disturbance, asset management.
Providing an appropriate level of service is the primary goal for water suppliers in urban water distribution systems. Even though a certain level of service is defined by regulatory standards, e responsible agents use to set their own commitments of service with thresholds that go beyond such specifications for improving customer satisfaction level. In this regard, the level of service, considered as the guarantee of supply at every property, can be measured in terms of continuity of service, pressure and quality provided to the end user.
On the other hand, service suppliers make remarkable efforts to establish suitable asset management policies. Aging infrastructures require increasing asset investment for maintaining the level of service. So that, developing a cost effective asset renewal and replacement strategy is essential. Two different strategies for replacement optimization have been analyzed in the literature he optimal replacement time for a pipe while the other is based on prioritizing the pipes in the network for replacement under a certain budget.
Actually, water suppliers are focused on improving efficiency in CAPEX & OPEX for some but all issues involved in asset management are linked to uncertainty. The key for a better asset management is to set criteria, methods and systems to facilitate that www.mdpi.com/journal/xxxx Development of an asset lifetime model for maintenance results in increasing asset investment in order to provide and maintain a specific level of service to consumers. In this regard, efficient replacement polices are needed. This paper proposes a method for improving renewal efficiency in water upply systems through a reliable asset lifetime model that will lead investments to those elements with greater impact in service provision to the end user. As uncertainties are minimized, the ents will become more efficient. Therefore, the failure predictor model has been built in a reliable manner through collected data from Madrid distribution network which comprises more than 17.000 km with over 400.000 water tical analysis of historical data from over 55.000 system failures gathered during four complete years. Additionally, detailed information from more than 4.400 disturbance events was recorded through field visits and laboratory essays of soil and pipe rials when failures were repaired. Examination of such large series of data recorded allows a better understanding of explanatory factors of failures. It is an essential step for building a l strategy is proposed. It based on the risk of service disturbance and involved costs. It supports planning and operation decisions urbance, asset management.
Providing an appropriate level of service is the primary goal for water suppliers in urban water distribution systems. Even though a certain level of service is defined by regulatory standards, e responsible agents use to set their own commitments of service with thresholds that go beyond such specifications for improving customer satisfaction level. In this regard, the level of n be measured in terms of On the other hand, service suppliers make remarkable efforts to establish suitable asset estment for maintaining the level of service. So that, developing a cost effective asset renewal and replacement strategy is optimization have been analyzed in the literature for a pipe while the other is based on Actually, water suppliers are focused on improving efficiency in CAPEX & OPEX for some but all issues involved in asset management are linked to uncertainty. The key for a better asset management is to set criteria, methods and systems to facilitate that

65
Some solutions are focused on the development of water networks models to foresee the system 66 possible behavior. However, when defining assets lifetime models, uncertainty has also to be faced.

89
In this paper, a method for improving renewal efficiency in water supply systems through a

97
This way, the proposed method is based on the minimization of uncertainty linked to the 98 prediction of system disturbances while analysing its consequences and system resilience. In this 99 context, a reliable model of failure prediction for defining assets lifetime is needed to minimize 100 uncertainty. Through this model a more accurate diagnoses of system condition can be obtained in 101 order to lead renewal and replacement policies.

102
Renewal efficiency is considered as the investment option that reduces disturbances impact as 103 well as optimizes investment costs. Therefore, this approach also includes the evaluation of costs by 104 considering the trade-off between replacement cost and the cost related to failures. The cost of 105 failures includes not only service disturbance impact to end users as a cost but also repair costs.

106
From this approach, assets lifetime is governed by an age threshold where renewal cost is below the 107 cost related to failures.

137
In order to analyze relations between elements and failures, some data such as age, material,

189
This way, every element in the system has new attributes related to the impact of its potential 190 failure that is defined as 'properties x hour' with service interruption, its repair cost and its

202
Therefore, benefit of replacement C 1 for a horizon of N years can be calculated by the 203 following formula (4), where A is the old element with an age of k years, and R is the new element. Finally, according to its service perturbation and costs, a priority list of elements that should be 209 replaced can be obtained. In such way, renewal investments will be led in an efficient manner to 210 those elements with greater impact in service provision to the end user while optimizing 211 investments.

213
The water utility Canal de Isabel II Gestión, which supplies water to more than six million

219
The proposed method has been applied in the network in order to assess how the efficiency of 220 investment changes with different hypothesis of the failure predictor model. It has been applied to 221 Madrid Region where the network comprises more than 17.000 km with over 400.000 water pipes.

249
Regarding system's diameter, almost the 90% of distribution water mains present diameters 250 smaller than 250 mm, and the likelihood of failure is greater for pipes with diameters lower than 150 251 mm. In the case of strategic water mains, the 80% of pipes present diameters smaller than 1000 mm 252 and the likelihood of failure is greater for pipes with diameters below 500 mm. Besides, more than 253 90% of service connections present diameters inferior to 50 mm and while the more frequently 254 installed diameter in the recent year is 20 mm, the greater gradient of failures is presented for 255 diameter 40 mm which was installed in the past. The 90% of special elements got diameters below 256 250 mm and failure is more frequent for those where diameter is smaller than 80 mm.

257
Described methodology has been applied to this case study. Firstly assets lifetime has been 258 calculated through the failure prediction model, defining the likelihood of failure for every element.

259
Then, a second module calculates for each of them its potential failure impact in service provision as 260 the risk of service disturbance measured by 'properties x hour' and the repair cost. Then, it is 261 compared with the renewal cost as it is presented in Figure 4 for distribution water mains and in

272
Moreover, elements located in the right side of the threshold proposed include some cases that 273 preventive replacement implies great disturbances. It is the case of critical pipes for providing 274 service where assuming the risk of failure is recommended.

275
As replacement priorities change according to system conditions and previous investments, the 276 proposed method forecasts annually the set of pipes that should be renewed to minimize service 277 disturbance.

279
As conclusion, asset lifetime models reliability is a relevant factor to consider for leading water 280 replacement investments and giving support in water utilities decision making process.

281
In that process the lifetime model provides useful information in terms of planning as well as

292
As replacement priorities change according to system conditions and previous investments, the 293 model of residual useful life forecasts annually the set of pipes that should be renewed to minimize 294 service disturbance. The new renewal strategy based on the proposed useful life model will provide 295 an improvement of investment efficiency for water companies but also it will increase customers' 296 satisfaction from reductions on service disturbance events.