Sewer Life Span Prediction: Comparison of Methods and Assessment of the Sample Impact on the Results

Abstract

Survival models can support the estimation of the resources needed for future renovations of sewer systems. They are particularly useful, when a large share of network will need renovation. This paper studies modelling sewer deterioration in a context, where data are available for pipes selected for inspections due to suspected or experienced poor condition. We compare the random survival forest and the Weibull regression for modelling survival and find that both methods yield similar results, but the random survival forest performs slightly better. We propose a method for estimating the range in which the actual network survival curve lies. We conclude that in order to reach reliable results, a life span model needs to be constructed based on a random sample of pipes, which are then consecutively inspected and in addition, censoring and left truncation need to be accounted for. The inspection data applied in this paper had been collected with the aim of finding pipes in poor condition in the network. As a result, the data were biased towards poor condition and unrepresentative in terms of pipe ages.

Challenges in Sewer System Maintenance

“…Sewer pipes are a very important and critical part of the sewer system-not only because they play an important role but also because their purpose is to transport waste and rain water to wastewater treatment plants or recipients and, therefore, their construction and repair are expensive but necessary [58]. Limitations in data availability and quality are common for sewer system status data [59].…”

“…Therefore, sewer systems that use the GIS support system are more favourable in terms of management and maintenance than those that do not use GIS, especially if they are longer. Sewer pipes are one of the rarely examined infrastructures [59,89]. Due to their good functioning, it is important that they are regularly examined and recorded, cleaned and repaired.…”

Construction and maintenance of urban underground infrastructure with digital technologies

“…The model was demonstrated robust to smaller calibration sample sizes, indicating its potential for predicting sewer deterioration when only a small portion of pipes was inspected. Similarly, random survival forest and the Weibull regression were evaluated for modeling sewer deterioration, in which the former is a machine learning method for survival analysis while the latter is an established parametric method where the survival times follow the Weibull distribution [97]. The two models generated similar results, but the former performed slightly better.…”

Probabilistic Assessment of Biodeterioration Effects on Reinforced Concrete Sewers

“…The probability distribution, the mean value and coefficient of variation of the variables used in this work were obtained from the literature and are shown in Table 2. The selection of lognormal and uniform distributions was produced based on a literature review [100,[104][105][106][107][108][109][110][111]. The lognormal distributions were appropriate for parameters that cannot be physically negative, while uniform distributions were selected when the range of variation of the parameter was known.…”