Assessment Methods for Structural and Hydraulic Properties of Concrete Sewer Pipes

Abstract

Asset management is a tool for maintaining the required level of serviceability of urban drainage systems, which are costly to construct and in some cases even more costly to replace. The required asset management effort to achieve a certain level of service is unclear due to a limited knowledge on sewer failure mechanisms and due to scarcity of information on the functioning and conditions of urban drainage systems. An important question is: what information of what quality is necessary for cost-effective sewer asset management? In The Netherlands the majority of urban drainage systems are made of concrete elements (about 72%), including nearly all main sewers > 500 mm, thus making the information on the concrete sewer pipes of significant importance. This research aims at (i) identifying the required information on the structural and hydraulic performance of the sewer pipes, (ii) quantifying the uncertainties of information and (iii) improving the quality of this information in order to further understand the changes in processes/status. Sewer failure mechanisms explain the structural and/or operational failures of sewer elements. In order to be able to identify the main processes and defects responsible for the structural and/or operational failures of sewer elements, as well as the possibility of obtaining the information about them, a HAZard and Operability (HAZOP) approach was applied. This technique was demonstrated to be applicable for analysing the information needed for sewer asset management. Structural strength and hydraulic capacity are two essential parameters in the assessment of the need for sewer rehabilitation. In sewer systems where corrosion is the dominant failure mechanism, visual inspection by closed circuit television (CCTV) and core sampling are among the methods mostly applied to assess sewer pipe condition. A study was carried out on visual inspection and drill core analysis in order to enhance a further understanding of the limitations and potentials of both methods and the added value of combining the information from both sources. Both methods have been applied on a selected sewer reach in the city of The Hague, which was reportedly subject to pipe corrosion. The results show that both methods, visual inspection and core sampling, are associated with large uncertainties and that there is no obvious correlation between the results of visual inspection and the results of drill core analysis. The conclusion is that information of a certain quality (depending on the circumstances) on the actual status of the assets is a prerequisite for adequate sewer asset management. For instance, especially concrete pipes suffer from loss of wall thickness due to biochemical corrosion and, consequently, a decreasing structural strength along with an increase in hydraulic roughness. Unfortunately, routinely used visual inspection methods do not allow a quantification of the internal pipe geometry, which would enable not only detection but also the quantification of the progress of biochemical corrosion. Advances in laser technology and digital cameras theoretically allow a cost-effective application of laser profilers to measure the interior geometry of sewer pipes. An analysis of associated uncertainties revealed that the position and alignment of the laser in commonly used laser profiling techniques are the main sources of measurement errors. A full-scale laboratory set-up demonstrated, based on tests on a new and an 89 year old corroded sewer pipe, that laser profiling is indeed capable of measuring the interior geometry accurately enough to determine wall thickness losses for corroded pipes, provided that the position and alignment of the laser and camera are accounted for. Further, drill core samples are taken for an analysis of the material characteristics of concrete pipes in order to improve the quality of decision making on rehabilitation actions. It was shown in this study that core sampling is associated with a significant uncertainty. The results of core samples are compared with the results of full-scale pipe cracking lab experiments. It is shown that the concrete of deteriorated sewer pipes shows a significant variability in material characteristics. Further it is shown that the formation of ettringite due to biochemical sulphuric corrosion is not necessarily limited to the crown of the pipe and that also degradation of pipe material, measured by the carbonation depth, is occurring at the inside and outside of the pipe. It is concluded that tensile splitting strength and carbonation depth (i.e. loss of 'healthy' wall thickness) are material property parameters of core sampling with a sufficiently high correlation (R^2 > 0.90) with the constructive strength of the pipe. The thickness of the remaining healthy concrete material is the optimal parameter in terms of correlation with collapse strength, as this requires the smallest sampling size. Furthermore, in sewer asset management, decision making on rehabilitation or replacement should preferably be based on the actual functionality of a sewer system. In order to judge the ability of a sewer system to transport sewage, hydrodynamic models are used: hydraulic roughness is one of the key parameters. For new pipes, the hydraulic properties are well known, but for aged pipes, with uneven deterioration along the cross section, information on the hydraulic roughness is lacking. The potential of laser profiling methods for accurate, non-invasive and non-intrusive assessment of the hydraulic roughness of concrete sewer pipes is described, demonstrated and discussed. Processing of raw scanned data consists of two steps: (i) spatial interpolation with uncertainty analysis and (ii) statistical analysis for estimating the hydraulic roughness. Moreover, a statistical analysis was carried out to determine the minimal scanning resolution required in order to yield results accurate enough for subsequent modelling uses. The results show a promising potential of the laser scanning approach for a simple and fast quantification of the hydraulic roughness in a sewer system. A Prototype v1.0 (in this study) of an unbiased laser profiler was developed to improve the accuracy of collected information. However, there is a need for more accurate apparatus. The new design of the Prototype v2.0 presented provides accurate measurements (sigma < 2 mm) of the cross section and, from frame to frame, an accurate 3D image of a pipe. The potential applications of the improved laser profiling technique are comprehensive e.g. enhancement of inaccurate visual inspection, deposit measurements, roughness measurements. The combination of the two methods, i.e. to use the laser profiler to determine the pipe interior geometry as well as to identify representative patches where roughness should be measured, is an opportunity to strengthen laser profiling as a method that may partially replace a CCTV inspection as a dominantly applied sewer investigation technique. Additionally to achieve a higher accuracy, there are several improvements that can be applied to a potential third version of the prototype. Currently the amount of raw data that is generated during the experiment over 1 m of pipe length is around 4.86 GB: the data flow is too high for an embedded application but this can be significantly reduced. There are still some improvements to be done with the presented hardware to make the data acquisition faster and easier. Future work will concentrate on the development and improvement of the laser profiling technique accuracy and possibilities for its use - Prototype v3.0. The results of this research will be used for future development of inspection strategies using core sampling. For sewer rehabilitation decisions, it is necessary to be able to calculate the remaining strength of the soil-pipe construction environment for deteriorated pipes. Further research will concentrate on simulations with a Finite Element Method (FEM), with the pipe geometry information provided by laser profiling and material properties by core sampling. The model will be used to determine the remaining load-bearing capacity of a sewer pipe and to determine the type of information needed to further enhance the decision making process.