Seismic monitoring of the Subsurface along Transportation Infrastructure using passive measurement on TeleCOM fiber-Optic – SITcomOptics

Passive seismic:

The proposed geophysical investigation method is based on passive seismic, which is based on the reconstruction of the Earth's impulse response between pairs of seismic sensors by interferometry of ambient seismic vibrations. In this project, sensors deployed in linear fashion are used to carry out multichannel analysis of surface waves reconstructed by interferometry. The origin of ambient seismic vibrations is rail and/or road traffic.

DAS:

Vibration measurements will be carried out on optical fibers deployed horizontally, using DAS (Distributed Acoustic Sensing) technology. In contrast to three-component geophones which carry out point measurements of particle velocity of the seismic signal in the 3 dimensions of space, the optical fiber measures the longitudinal deformation. DAS technology allows to carry out seismic measurements distributed every 2 m at up to 2 kHz over lengths of 50 km. The amount of data can be up to several Tb/day, which represents a major challenge to handle.

Machine Learning (ML):

In this context, ML can provide interesting solutions both for identifying signals of interest in DAS data (and thus reducing data amount), as well as for increasing the quality of the data (data denoising).

Digital simulation:

The use of numerical simulation is necessary to understand the data on longitudinal deformation sensors such as optical fibers. The evolution of deformation measurements as a function of the evolution of geomechanical properties in a numerical model will make it possible to evaluate the sensitivity of the approach.

Measurement campaign in real and controlled environment:

The project lies on the acquisition of new data along a 40 km pilote telecom optical fiber in a railway context. To calibrate these measurements, a 500 m linear platform is built as part of the project near the pilot fiber. Different optical fibers with different types of ground coupling (directly buried and in a buried PVC pipe) and different types of cable coupling (loose and tight-buffered design) are installed on the platform in order to carry out measurements in a controlled environment.

Calibration platform construction:

A calibration platform (500 m) was built along the 40 km pilot line identified before the beginning of the project , which is already equipped with an SNCF telecom cable. Two types of fiber optic cables (an SNCF telecom fiber optic cable, and a fiber cable dedicated to vibration measurement) were installed in different coupling configurations with the ground (directly buried in the ground and in HDPE pipe).

Calibration acquisition phase:

A 5-day calibration acquisition phase made it possible to collect train recording data, ambient noise, and controlled impulse source impact. Different configurations were tested in terms of optical pulse length, gauge length and spatial and temporal sampling steps. Acquisitions were also made synchronously on the existing SNCF telecom cable. For calibration, 4 broadband seismometers were deployed, as well as a three-component geophone line.

Improvement of the DAS measuring bench:

Also, an approach based on polarization diversity was implemented in the laboratory on a prototype of the SNCF DAS measurement bench. The results show a significant improvement in the signal-to-noise ratio at frequencies which are those of the seismic signals generated by the passage of trains (a few tens of Hz).

The processing of data from the calibration phase will make it possible to estimate the optimal acquisition parameters to launch the long-term acquisition phase planned as part of the project.

The new DAS measurement prototype will be installed to carry out additional measurements on the calibration platform. By comparing with previously acquired data, the degree of improvement can be estimated on field data (train recording). If the results are conclusive, the new prototype will be used to carry out the long-term acquisition.

From the first data acquired during the calibration acquisition phase, specific machine learning algorithms will be developed to identify the spatio-temporal windows that can be used to carry out geophysical measurements from DAS data on telecom fiber.

The climate change makes the near-surface geohazards risk mitigation a priority, and there is a need for the monitoring of earth material beneath cities using new sensing strategies. Distributed acoustic sensing (DAS) is a recent breakthrough in opto-electronics, which allows recording seismic vibrations on fiber optic (FO). This, in turn, may be used for near-surface geohazards assessment.
The application of this technology on existing telelecom FO already deployed in cities for monitoring purposes is appealing, but remains challenging. This project aims to evaluate how the existing telecom FO can be used to monitor the near-surface beneath transportation facilities. For that purpose, an integrated study combining telecom FO DAS measurements in a railway context, machine learning, geophysical processing and physical modeling is proposed. The outcome of this research will allow developing safe and sustainable smart cities from the leveraging of existing communication infrastructure.