GUIDE  (Ground and Underground Infrastructure Damage Evaluation)

Motivation

It is estimated that in the UK alone we spend over £7 billion annually on streetworks . A large share of this cost is induced by uncertainty or lack of knowledge of location and condition of our underground assets. There are numerous cases where exact position and/or condition of buried assets and their supporting ground are extremely challenging, if not impossible, to determine using current remote sensing technologies. This can unnecessarily prolong the streetworks, lead to damages to the assets or even fatal injuries.

Performance of the available techniques is susceptible to the burial condition of the pipe, e.g., type and physical properties of the soil, burial depth, ground composition, and material of the pipe. Above all, existing techniques for condition assessment of underground utilities fail to capture a complete picture of the ground-utility symbiotic interaction and cannot provide information on the structural integrity of buried infrastructure and their supporting ground.

At the same time, routine and effective condition assessment of buried utilities as well as roads and railways can significantly reduce the probability of failures caused by deterioration and therefore is of significant financial and strategic importance and allows better resources planning.

A novel approach with clearly defined advantages in certain burial conditions, alongside available techniques, provides a more versatile, inclusive, and reliable toolbox for the detection and condition assessment of buried pipes.

Approach and aim

The focus of the GUIDE project was to develop a novel, non-expensive, yet very accurate method to both locate buried utilities and assess the condition of the utilities and their supporting ground.

Methodology

The proposed method uses ground deflections caused by a drop load to back-calculate the stiffness of the system. Here the word system refers to the integrated ground and buried utilities. The back-calculation procedure of the GUIDE is an optimisation procedure that uses i) finite element models of the ground and buried infrastructure to accurately capture the condition of the system and ii) a machine learning algorithm to estimate the location and condition of the buried asset based on the surface deflections. Figure 2 illustrates the two phases of the GUIDE back-calculation, i.e., finite element modelling and genetic algorithm-based machine learning. The proposed back-calculation approach can address the complex environment of underground surface, and soil-asset interaction. Through a set of laboratory experiments and numerical modelling, the GUIDE successfully demonstrated the benefits of the proposed approach in both locating and condition assessment of buried pipes. Large-scale experiments in the National Buried Infrastructure Facility (NBIF) and field trials are planned to be conducted in near future to establish the extent in which the proposed method can be used.

Image: an overview of the GUIDE procedure

GUIDE inversion algorithm is proposed for the following applications:

• Detecting a buried stiffness anomaly, e.g., pipes, in the shallow subsurface
• Condition assessment of buried asset-ground systems by identifying voids and estimating the stiffness of the ground and the asset
• Identifying structurally weakened zones in the shallow subsurface

GUIDE technique is based on inferring underground condition from ground surface deflections under a drop load (deflection bowl). The presence of anomalies or variations in the stiffness of the shallow subsurface affects the surface deflections when a load is applied to the ground. In GUIDE technique a drop load is applied to the ground surface and surface deflection are recorded using an array of geophones, this can be facilitated using a falling weight deflectometer device. This information are then fed to the GUIDE inference algorithm.

The animations below show two sample cases where the presence of a pipe and a leakage void affect the stress distribution in the soil and, consequently, the surface deflections under a drop load.

 

Finite element modelling

In GUIDE project, detailed FE models of buried pipe-ground were developed to simulate mechanical behaviour of the ground under loading. The validated FE models were then incorporated in the inversion algorithm for locating buried pipes. Below, is an example of the developed FE model based on a real scale experiment  where mechanical behaviour of a buried concrete pipe alongside two different voids was tested under loading.

The FE forward model to identify voids adjacent to a buried pipe. The experimental test reported by Peter et al. (2018), (a) the three-dimensional model of the buried pipe-intact soil, (b) to (d) two-dimensional model of the intact soil, small void, and large void, consequently.

Diagrams: above: three-dimensional model of the buried pipe-intact soil. Below (left to right) two-dimensional models of the intact soil, small void, and large void, consequently.