QC Inspection of a Relined Freshwater Culvert – Utilising Simulated Data Outputs, UK

Project summary

The client had relined a water tunnel using a 150mm thick reinforced concrete liner. There were concerns that there were voids and other defects within the new lining, not visible on the surface, that would need to be rectified.

The brief:

To support the data interpretation and prove assurance of the GPR results to the client, a simulation of the inspection was run. This data was not complex, but the client had doubts over the GPR’s ability to detect defects.

Our work:

A model was created from the client’s design information of the new tunnel lining. This consisted of a 150mm thick reinforced concrete slab (not curved for simplicity) over a second concrete layer and the ground below. Air voids of various sizes were created within this slab at several depths, above and below the bars. Additionally, areas of ‘wet’ concrete were created by including areas of concrete with higher conductivity and dielectric.

Figure 1: Example model: 150mm of concrete (orange) with bars (blue dots) at 100mm centres. This sits on the original pipe (light blue) and some earth (red). We have added voids (dark blue) and ‘wet’ patches (orange) to show areas of higher conductivity and amplitude.

The simulated data was outputted, processed, and analysed. The simulated data allowed the team to know what to expect from the data prior to interpreting the collected data.

The data was collected through several longitudinal lines over 500m of tunnel at 30-degree intervals.

*Figure 2: Simulated data above and real data below showing an air void above the bars with an area of ‘wet’ concrete.*

The data from the simulation was directly comparable to the data collected onsite . Although the simulation was perfect with straight edges and distinct features, the collected data showed less distinction between features.

For this project, the data was interpreted, and the locations of detected defects were presented in a tabular format within a written report. Using experience and simulated data, it was possible to provide context and definition to the features and anomalies extracted from the data. This allowed the client to drill on indicated locations and repair voids found.