In 2016, Barcelona opened its metro line connecting the city to the airport. This was a much-requested infrastructure for the city expansion and involved a complex construction project. Maria Navarro led the team handling the planning, supervision, quality control and auxiliary works for the monitoring of the Barcelona L9 Metro underground excavations.
Meanwhile, Ángela Lluch was in Singapore working for the Land Transport Authority. She was supervising the construction and monitoring team for the Thomson East Coast metro line works, following her previous experience as a consultant of geotechnical instrumentation and monitoring for the High-Speed Rail Line in Barcelona.
In the same period, Juan Pérez was managing geotechnics, data management and acquisition systems for several projects in his role as Instrumentation and Monitoring Director, among them the High-Speed Rail Line in Barcelona, the Valencia metro, as well as various tunneling works in the main highways in Northern Spain.
Their paths crossed at Worldsensing, where they currently manage key monitoring projects all over the world to digitize operations with Internet of Things (IoT) technology. TBM: Tunnel Business Magazine joined them for a talk on how IoT unleashes an irreversible change that will improve the risk management of tunneling activities.
All of you have over 15 years’ experience in the monitoring industry. Where does monitoring stand in terms of risk management? How have methods and prevention evolved?
Maria: Tunneling is an activity involving major works that inevitably pose risks, most of them related to the unknown geotechnical conditions and others due to human factors. Tunneling methods have evolved and technology has helped to develop more secure excavation methods and machinery, such as tunnel boring machines. Nevertheless, there is always an implicit risk in an excavation itself, as ground conditions change during and after the excavation process, leading to movements induced on nearby structures.
For that reason, monitoring plays an important role in risk detection and prevention as it provides information to find out if real movements happen as expected according to the numerical simulation models, and also gives an idea of the risk management capacity (soil failure). All this information allows for corrective measures to be applied, if needed.
Automating the monitoring process has become more frequent and data availability in near real time has led to an improvement in the understanding of soil/rock behavior and has also increased the risk management capacity.
Although automation in tunnel monitoring projects began in the last decades, there is still a long way to go. Geospatial monitoring is usually the first part of the monitoring project that tends to be automated and it is usually prescribed in the project specifications unlike geotechnical monitoring, as this implies more effort in terms of cost and deployment.
From the design of a tunnel, through construction to operation, what is the involvement of monitoring technologies? How is this included or modified during the process? What areas of improvement are needed?
Ángela: Monitoring is an essential part of any construction project; it is the way to ensure compliance with local regulations and standard codes in terms of safety. During the design phase of a tunnel, certain quantities of instruments are allocated to monitor key parameters. This amount varies according to each design particularity and construction phase. Prior to construction, monitoring points provide a baseline of the existing conditions (e.g. groundwater level, thermic response of buildings) whereas in the post-construction phase, low frequency data is needed until the ground is deemed stabilized.
During the execution phase, however, real-time monitoring is relevant to cope with the pace of the construction and be able to make decisions on time. Several factors such as site constraints, complexity of the design and unexpected unfavorable conditions play their part in the necessity of deploying new technologies.
Traditional methodology, like manual monitoring, might become at some point insufficient and impractical in a tunneling project in terms of human resources, tight schedules, sampling frequency and ease of access and deployment, the latter is also a drawback of cabled instrumentation. All this leads the market to strive for new solutions to meet the demands. A good example of this is the adoption of systems running on IoT low-power wide-area networks (LPWANs), such as Sigfox and LoRa. These technologies not only solve these constraints but also offer increased data accuracy and reliability.
Although new technologies are being adopted slowly, much emphasis should be placed to correct the misconception that monitoring is a complementary or minor activity, especially during the execution phase of the project.
Would you say that the majority of projects worldwide use automation in their monitoring programs? Are IoT and wireless technologies commonly used?
Juan: It depends on the country, but in general, we can say that the majority of the tunnel construction projects have incorporated automation in some parts of their monitoring programs, particularly using Automated Total Stations for monitoring. On the contrary, in-ground sensors (piezometers, borehole inclinometers, extensometers) are still read manually using portable read-out units.
In-ground sensors can help to implement remedial actions by providing the data needed to forecast some of the effects of tunneling on buildings before using other monitoring techniques. With in-ground sensors like extensometers and piezometers, it is possible to monitor not only the consequences (response of buildings to excavation-induced ground movements) but also the causes. Geodetic techniques often just measure the consequences of the movements but rarely the original causes.
Something similar happens with the sensors used to monitor the behavior of structural elements and the soil-structure interaction like vibrating wire (VW) strain gauges, pressure cells and load cells. The data collected from these instruments can be used to verify design assumptions and to check that performance is in accordance with what was predicted by numerical modelling included in the design. In some cases, these data could reveal at the right time the need to introduce changes in the design, reducing uncertainty inherent in geotechnical work.
In conclusion, the reading of the installed in-ground and structural sensors should also be automated as it is currently done with Total Stations.
Why do you think this is not the case so far?
Juan: It is due to several reasons. First, because until recently adequate data acquisition and transmission technologies did not exist for reading geotechnical sensors the way they are deployed at site: installed from the surface in drilled boreholes, attached to the walls of buildings or embedded in structural elements. It was impractical to connect these sensors to centralized wired data loggers due to power requirements and, in some cases, the need for trenches and costly installations.
Second, because tunneling projects are in constantly evolving environments, many times it is difficult to implement a centralized wired data acquisition system from the initial phases just after the excavation, when most of the ground movements and changes in stresses happen. A lot of effort in cable protection would be required and usually these acquisition systems were only installed after construction for monitoring during operation phase.
All in all, if we consider that the density of geotechnical monitoring points in tunneling projects is usually low, a long-range and low-power wireless solution compatible with geotechnical sensors is the most suitable solution. Nowadays it is possible to use the technologies from LPWAN to collect automatic readings of the referred geotechnical instruments. In a tunnel, wireless nodes can be installed just after the excavation at the same time of the sensor installation. In addition, the nodes can be moved, uninstalled and reinstalled according to the different excavation and construction projects.
Finally, sometimes there are additional reasons for not implementing automatic acquisition of the geotechnical instruments related to the contract specifications. In particular when these specifications do not take into account the possibility of implementing wireless monitoring and very low frequencies of measurement compatible with sporadic manually collected readings are requested.
How does wireless monitoring work?
Maria: In a wireless monitoring system, sensors are connected to nodes that are automatically programmed to take readings at the required sampling rate and send data to a dataserver, which can be hosted on a gateway or on a cloud.
Data from the dataserver can be integrated in a visualization platform which can be used to manage data visualization and alarm systems.
Communication from node to the dataserver can be done through different technologies available on the market.
Most wireless technologies depend on local area networks (LAN) which include technologies such as Zigbee, Bluetooth, WiFi or other mesh technologies (2.4 GHz) like wireless Hart. These are mostly used for short distances between 50 and 100 meters but they are not sufficient to cater for most tunnel construction monitoring needs or for the geotechnical market.
Cellular technologies (3G/4G) can reach longer distances and allow for frequent data collection, but their disadvantage is that mobile devices have higher power consumption rates. On top of this, both LAN and cellular technologies are dependent on network signal and line of sight (LOS), which can be difficult to achieve in tunneling projects and, in order to reduce power consumption, it is common for tunneling application data to be sent only once a day or once a week which it is not suitable for real-time or near real-time data acquisition.
LoRa, which stands for Long Range communication, is based on IoT technology and it addresses the issues of LAN and cellular technologies on range and power consumption. With this LPWAN technology, data can be transmitted over large distances while only needing very little power. Batteries can last very long as devices only ‘wake up’ when they have to read and transmit data and go back into sleep mode afterward.
Are wireless-based systems more suitable for a specific kind of tunnel?
Juan: Wireless-based systems are suitable for all kind of tunnels. However, we can distinguish between surface monitoring and underground monitoring.
In surface monitoring we would include all the sensors installed in boreholes drilled from the surface (borehole extensometers, piezometers, in-place inclinometers), sensors installed on and into the buildings in the influence zone of the excavation (tiltmeters, tilt beams, liquid level settlement cells, crackmeters) and sensors installed in the cut and cover tunneling projects (VW strain gauges, load cells, laser distance meters). Monitoring points are spread over a large area and the distance between monitoring points can be significant so an LPWAN type of telecommunication is the most suitable solution to achieve real-time data from geotechnical instrumentation. Many times, wireless-based systems are the only viable solution to collect automatic readings from this kind of instrumentation.
In underground monitoring, wired centralized data loggers can also be used because wiring is feasible and power supply can be available. However, wireless monitoring has some significant advantages.
In TBM projects, the sensors (strain gauges and pressure cells) are installed in the precast concrete plant. Then, after being placed by the TBM and when they become accessible, it is required to access the connection boxes and wire the cables to the datalogger. We can expect in the midterm that the wireless nodes will also be installed in the precast concrete plant, making the segments “smart” and provide automatic readings from the beginning, just after being placed into the lining and without requiring any physical access.
In tunnels constructed using the sequential excavation method (SEM), it is difficult to install a wired data acquisition system to read the geotechnical sensors (borehole extensometers, pressure cells, strain gauges) in the initial phases because the construction site is constantly changing and power and communication cables will be damaged. Considering that the most important information is collected right after the excavation of a section, it is crucial to have a flexible system ready for all these changes: excavation of top heading, initial support, excavation of bench, permanent support etc.
Given all these sequential activities, the best, and sometimes the only feasible way of collecting automatic readings just after installing the sensors is through the use of a wireless system. Long-range systems are useful because the distance between monitoring sections can be significant.
Tunneling infrastructure continues to age as the population continues to grow. How does this impact the future of IoT? What trends are you seeing?
Ángela: IoT technology will play a key role in risk management and ensuring minimal nuisance to the population. As more tunnels are built to meet demand, monitoring will be even more important due to a higher risk of influence over other surrounding structures. All this will definitely enforce the usage of monitoring not only for safety purposes but also to optimize the resources and reduce the duration of the construction phases.
The trend is to be able to gather data from different sources to a centralized domain that enables wise and fast decision-making – a methodology known as Operational Intelligence. For instance, a tunneling contractor would benefit from a tool that would allow him to use data from other contracts within the same project, such as the soil parameters assumed in the design interface, structural and geotechnical monitoring observations made during the different execution phases, and corrective measures adopted during the course of the construction for his own benefit.
All this centralized data would facilitate contractors encountering similar conditions make all the necessary adjustments against their own predictions if deemed convenient. What is more, they could go through their model and check the outcome before any implementation is done to minimize risks.
What are some recent tunneling projects where you have deployed IoT-based systems to monitor risks?
Maria: We are working on some big projects, like for example, the Grand Paris Metro where several contractors are involved in the monitoring execution. Automation of several sensors, such as tiltmeters, has been carried out to work as a complement of Total Stations, or H-Level sensors (installed on basements) to gather information where geodesic methods are not valid due to the impossibility to have external references. In-ground instrumentation like vibrating wire piezometers or chains of in-place inclinometers and structural elements such as struts, by using vibrating wire gauges, have also been automated.
Wireless monitoring is applied on several projects around the world, like metros in LA, Melbourne and Auckland.
What advice would give to operators who are considering using wireless systems to monitor their projects?
Ángela: Several elements should be considered. First is radio range, as the dimension of a tunneling project covers more than 2-3 km and the distance between monitoring points can be significant (>100 m). So the monitoring system should be able to be deployed anywhere and not reliant on signal coverage.
A wireless network has to be plug-and-play. Monitoring plans are usually set at the beginning of the project, but then as the project progresses, and depending on the measurements, corrective measures might be necessary. In this scenario, network adaptability is key to relocating nodes to different locations, add new nodes to cover new monitoring requirements. Additionally, the network has to be capable of being updated by itself, with no need of manual reprogramming to prevent wrong readings.
Also, tunneling sites are harsh scenarios that require robust equipment. It’s important to check water tightness, battery resistance to extreme temperatures and the possibility to work without additional casing. Units like Loadsensing, for example, are IP67 rated and tested from -40C to +80C to ensure accurate monitoring for years.
It is important that the wireless solution adopted is compatible with sensors typically present on tunneling projects (water level sensors, crackmeters, piezometers, load cells, etc.). The system should be compatible with different software suites that display data from different sources such as geodetic survey, vibration monitoring, wireless solutions, etc. For that, a system that can transfer the data to different software by different means (e.g. Modbus, API calls or FTP) would play its part.